Global Change Biology (2013), doi: 10.1111/gcb.

12194

REVIEW ARTICLE

Altered dynamics of forest recovery under a changing

climate
KRISTINA J. ANDERSON-TEIXEIRA*†, ADAM D. MILLER‡, JACQUELINE E. MOHAN§,
T A R A W . H U D I B U R G ¶ * * , B E N J A M I N D . D U V A L † † and E V A N H . D E L U C I A ‡ ¶ * *
*Conservation Ecology Center, Smithsonian Conservation Biology Institute, National Zoological Park, Front Royal, VA, USA
†Smithsonian Institution Global Earth Observatory-Center for Tropical Forest Science, Smithsonian Tropical Research Institute,
Panama, República de Panamá ‡Institute for Genomic Biology, University of Illinois, §Odum School of Ecology, University of
Georgia, Athens, GA, USA ¶Energy Biosciences Institute, University of Illinois, at Urbana-Champaign, Urbana, IL, USA **Plant
Biology, University of Illinois, at Urbana-Champaign, Urbana, IL, USA ††US Dairy Forage Research Center, USDA-ARS,
Madison, WI 53706, USA

Abstract
Forest regeneration following disturbance is a key ecological process, influencing forest structure and function, spe-
cies assemblages, and ecosystem–climate interactions. Climate change may alter forest recovery dynamics or even
prevent recovery, triggering feedbacks to the climate system, altering regional biodiversity, and affecting the ecosys-
tem services provided by forests. Multiple lines of evidence – including global-scale patterns in forest recovery
dynamics; forest responses to experimental manipulation of CO2, temperature, and precipitation; forest responses to
the climate change that has already occurred; ecological theory; and ecosystem and earth system models – all indicate
that the dynamics of forest recovery are sensitive to climate. However, synthetic understanding of how atmospheric
CO2 and climate shape trajectories of forest recovery is lacking. Here, we review these separate lines of evidence,
which together demonstrate that the dynamics of forest recovery are being impacted by increasing atmospheric CO2
and changing climate. Rates of forest recovery generally increase with CO2, temperature, and water availability.
Drought reduces growth and live biomass in forests of all ages, having a particularly strong effect on seedling recruit-
ment and survival. Responses of individual trees and whole-forest ecosystems to CO2 and climate manipulations
often vary by age, implying that forests of different ages will respond differently to climate change. Furthermore, spe-
cies within a community typically exhibit differential responses to CO2 and climate, and altered community dynam-
ics can have important consequences for ecosystem function. Age- and species-dependent responses provide a
mechanism by which climate change may push some forests past critical thresholds such that they fail to recover to
their previous state following disturbance. Altered dynamics of forest recovery will result in positive and negative
feedbacks to climate change. Future research on this topic and corresponding improvements to earth system models
will be a key to understanding the future of forests and their feedbacks to the climate system.
Keywords: biogeochemistry, climate feedback, FACE, irrigation, regime shift, succession, throughfall manipulation, warming

Received 19 December 2012 and accepted 29 January 2013

tant role in the maintenance of species diversity, as

Introduction
different species use forests of different ages as habitat
The dynamic process of forest regeneration following patches. On a regional to global level, secondary forests
disturbance is of key importance, with ramifications on are consequential for their role in climate regulation.
several scales. On a local level, forest recovery involves Forests recovering from disturbance (secondary forests)
wholesale rearrangement of vegetative structure, are strong C sinks and play an important role in the
carbon (C) and nutrient cycling, ecosystem physiology, global C cycle (Running, 2008; Pan et al., 2011). For
and community structure (Table 1). On a landscape instance, in recent years (2000–2007), regrowth of tropi-
level, disturbance–recovery dynamics play an impor- cal forests following agricultural abandonment took up
an estimated 1.7 Pg C yr 1 (Pan et al., 2011), which is
Correspondence: Kristina J. Anderson-Teixeira, Smithsonian equivalent to ca. 20% of annual global fossil fuel emis-
Conservation Biology Institute, National Zoological Park, 1500 sions. Beyond their influence on climate through their
Remount Rd., Front Royal, VA 22630 USA, tel. 1 540 635 6546, fax role in the global carbon cycle, secondary forests also
1 540 635 6506, e-mail: [email protected] influence climate through biophysical mechanisms (Liu

Published 2013. This article is a U.S. Government work and is in the public domain in the USA. 1
2 K . J . A N D E R S O N - T E I X E I R A et al.

Table 1 Typical trajectories of change in forest properties following stand-replacing disturbance

Forest property Typical trajectory References

Biomass
Biomass Rapid initial increase, peak at intermediate age followed Lichstein et al., 2009; Yang et al., 2011;
accumulation rate by slow decline to near zero in old-growth forests. Hember et al., 2012
Leaf biomass or area Rapid initial increase, relatively stable thereafter. Uhl & Jordan, 1984; Bormann & Likens,
1994; Law et al., 2003; Goulden et al.,
2011; Yang et al., 2011
Fine root biomass Rapid initial increase, relatively stable or modest Vogt et al., 1983; Claus & George, 2005;
decline thereafter. Yuan & Chen, 2010
Carbon cycle
Gross primary Rapid initial increase, relatively stable or modest Amiro et al., 2010; Goulden et al., 2011
productivity decline thereafter.
Net primary Rapid initial increase, modest decline thereafter. Gower et al., 1996; Law et al., 2003;
productivity Pregitzer & Euskirchen, 2004; Goulden
et al., 2011
Heterotrophic Relatively constant. Law et al., 2003; Pregitzer & Euskirchen,
respiration 2004; Goulden et al., 2011
Net ecosystem Initially negative (C source), increasing to maximum Law et al., 2003; Pregitzer & Euskirchen,
C balance (C sink) at intermediate ages, declining thereafter. 2004; Zhou et al., 2006; Baldocchi, 2008;
Controversy remains as to whether it declines to zero Luyssaert et al., 2008; Amiro et al., 2010;
(C neutrality). Goulden et al., 2011
Biogeochemistry
Foliar [N] Relatively constant with age, although both decreases Davidson et al., 2007; Drake et al., 2010;
(more common) and increases have been observed. Yang et al., 2011
N mineralization Mixed responses; both increases and decreases have Vitousek et al., 1989; LeDuc & Rothstein,
been observed. 2010
Hydrology
Canopy transpiration Rapid initial increase, modest decline thereafter. Roberts et al., 2001; Delzon & Loustau,
2005; Amiro et al., 2006; Drake
et al., 2011a
Hydraulic limitation Increases with age. Drake et al., 2010, 2011a
Sensitivity to variation Decreases with age. McMillan et al., 2008; Drake et al., 2010;
in water availability Voelker, 2011
Community dynamics
Species turnover Rapid initial turnover, decelerating decrease in Anderson, 2007b
turnover rate as the forest ages.
Species richness Initial increase, sometimes peaking and declining Shafi & Yarranton, 1973; Finegan, 1996;
modestly in older forests. Anderson, 2007b
Competition Increasing competitive advantage to late-successional Bazzaz, 1979; Bazzaz & Pickett, 1980;
species (e.g., shade tolerant, slow growing, higher Finegan, 1984
wood density, longer lived)
Size structure Initially, relatively even aged; competitive thinning Enquist et al., 2009
and seedling recruitment drive convergence
toward inverse square relationship between
abundance and diameter (diverse age structure)

et al., 2005; Maness et al., 2012; O’Halloran et al., 2012); 27 million km2 (Hurtt et al., 2011), and an estimated
for example, in northern regions albedo decreases with 1.2 million km2 are in use as forestry plantations
forest age and strongly shapes the net climate regula- (Kirilenko & Sedjo, 2007). In addition, natural distur-
tion services of secondary forests (Randerson et al., bances affect a significant proportion of Earth’s ecosys-
2006; Jin et al., 2012; O’Halloran et al., 2012). tems; disturbances such as fires, storms, droughts, and
A large and growing proportion of forests have been insect outbreaks affect over 100 000 km2 annually in
affected by major disturbances. Globally, secondary for- North America alone (Amiro et al., 2010). Climate change
ests recovering from anthropogenic disturbances such is generally increasing the incidence of natural distur-
as agriculture and wood harvesting cover an estimated bances (Dale et al., 2001), including fires (Westerling et al.,

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
 FOREST RECOVERY UNDER CLIMATE CHANGE 3

2006) and biotic disturbances such as insect outbreaks ent biophysical constraints on the number and size of
(Evangelista et al., 2011; Hicke et al., 2011). Theory and trees that can persist. Third, the successional pathway
models predict that future climate change may cause may be altered; for example, tree establishment may be
even more drastic changes (e.g., Westerling et al., 2011), delayed by altered climatic conditions such that pro-
depending on the future course of greenhouse gas emis- portionally more time is spent in an early-successional
sions and the resultant shifts in climate (IPCC, 2007). shrub phase. Distinguishing how climate change affects
Because secondary forests are strong carbon sinks with forests of different ages – and thereby how it alters suc-
considerable value for greenhouse gas mitigation (Ander- cessional trajectories – is critical in understanding how
son-Teixeira & DeLucia, 2011) and also represent a poten- climate change will impact both recently disturbed and
tial bioenergy source (e.g., US DOE, 2011), secondary mature forests.
forests are likely to play substantial roles in climate miti- This review considers how altered atmospheric CO2
gation initiatives and bioenergy production (Kirilenko & and climate are likely to impact trajectories of forest
Sedjo, 2007). recovery, with a particular focus on how climate
Although there is strong and abundant evidence that change may alter the rate of succession, the state of
climate change will affect forests of all ages, we lack mature forests, and successional pathways (Fig. 1). We
synthetic understanding of how climate change will consider five major lines of evidence relating forest
interact with forest age to shape the dynamics of forest structure and function to directional variation in CO2
recovery. Because forests undergo substantial reorgani- and climate (i.e., average conditions, as opposed to
zation of following major disturbance (Table 1), it is intra- or interannual variation), each of which yields
likely that climate change will have different effects on insight into how forest recovery may be altered under a
forests of different ages, thereby altering the trajectory changing climate. First, we review how climatic influ-
of succession relative to those observed for historical ences the dynamics of forest recovery across broad
climates. Climate change may alter one or more distinct climatic gradients. Second, we summarize the results of
features of successional trajectories (Fig. 1). First, the experimental studies quantifying the effects of elevated
rate at which the forest moves along the successional CO2, elevated temperature, and altered precipitation
trajectory may be altered without necessarily implying regimes on the dynamics of forest recovery. Third, we
any changes to the successional pathway or the state of review observations of altered forest recovery under
mature forests; for example, increased productivity contemporary multivariate environmental change.
may accelerate biomass accumulation without altering Fourth, we consider how altered community dynamics
the biomass of mature forests. Second, the state of may impact forest recovery. Fifth, we review model
mature forests may be altered; for example, maximum predictions. Finally, we synthesize findings from these
biomass may be altered if future climates place differ- separate lines of evidence, identify remaining uncer-
tainties, and discuss the implications for ecological
communities, biogeochemical processes, and the
climate system.

Dynamics of forest recovery across broad climatic

gradients
Although few studies have evaluated how climate
influences forest recovery across broad climatic gradi-
ents (Prach & Rehounkov
a, 2006; Anderson, 2007a),
there is clear evidence that climate exerts a strong influ-
ence on the rate of succession, the state of mature for-
ests, and their successional pathways (Fig. 1). The rate
of forest regrowth following disturbance is strongly
influenced by climate (Brown & Lugo, 1982; Johnson
Fig. 1 Schematic diagram illustrating three ways in which cli-
et al., 2000; Anderson et al., 2006). Globally, the rate of
mate change may impact the dynamics of forest recovery. Rela-
tive to the historical trajectory of change in a forest property
living biomass accumulation increases with tempera-
(here, biomass) with age, climate change may alter (1) the rate ture, being on average three to four times faster in the
of change, (2) the state to which the property converges as the tropics than in high-latitude forests (Fig. 2a; Anderson
forest matures (‘mature state’), or (3) the successional pathway et al., 2006). Likewise, biomass accumulation rate
(i.e., the sequence of states through which any given ecosystem increases with precipitation at a global scale (Fig 2b).
property passes and the relative amount of time spent in each). Within the tropics, rates of biomass accumulation are

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
4 K . J . A N D E R S O N - T E I X E I R A et al.

(a) (b)

(c) (d)

Fig. 2 Influence of climate on forest recovery rates (a, b) and on aboveground C stocks of mature forests (c, d). The rate of aboveground
biomass accumulation in forests recovering from stand-clearing disturbance varies globally with respect to (a) mean annual tempera-
ture (MAT) and (b) precipitation (MAP). Data, which are from Anderson et al. (2006), represent natural regeneration in 68 unmanaged
forests worldwide. Solid and dashed lines represent an exponential fit and its 95% confidence interval, respectively. Similarly, above-
ground C stocks (biomass + coarse woody debris; CWD) in mature forests vary globally with respect to (c) MAT and (d) MAP. Data
from Anderson-Teixeira et al. (2011). Dashed lines represent hypothesized bioclimatic limits.

dramatically higher in moist climates (precipitation temporal patterns of N availability during secondary
1000–2500 mm yr 1) than in dry climates (precipitation succession (Vitousek et al., 1989), rates of change in soil
<1000 mm yr 1); however, the positive influence of C and N (Li et al., 2012), and plant tissue stoichiometry
precipitation appears to saturate, with rates in wet cli- (Wright et al., 2004). Thus, climate may indirectly influ-
mates (precipitation >2500 mm yr 1) less than or equal ence forest recovery through its influence on biogeo-
to those in moist climates (Brown & Lugo, 1982; Mar
ın- chemistry, as occurs in the case of forests developing
Spiotta et al., 2008). Although further research is on Hawaiian lava flows (Anderson-Teixeira et al., 2008;
required to fully understand the mechanisms through Anderson-Teixeira & Vitousek, 2012).
which temperature, water availability, and their sea- A few studies have compared rates or pathways of
sonal dynamics affect rates of biomass accumulation in secondary succession across broad climatic gradients.
secondary forests, we can say conclusively that warmer Following clear cutting in western Oregon, climate
temperatures and higher moisture availability are asso- shapes both the rate and pathway of forest succession;
ciated with higher rates of biomass accumulation. in the western Cascades region, conifer regeneration is
Growth in secondary forests is strongly linked to ele- slower and follows a longer establishment delay com-
mental cycling. Biogeochemical cycles of elements pared with the more mesic Coastal Range region (Yang
including C, nitrogen (N), and phosphorous (P) are et al., 2005). In subalpine forests of the Colorado Rock-
coupled to biomass accumulation through stoichiome- ies, the rate of succession is more than twice as rapid in
tric constraints on the elemental composition of vegeta- mesic sites as compared with xeric sites (Donnegan &
tion, such that rates at which these elements are Rebertus, 1999). Likewise, in the Medicine Bow moun-
sequestered in vegetation are grossly proportional to tains of Wyoming, succession to a mature spruce–fir
rates of biomass accumulation (Yang et al., 2011). forest is most rapid in a mesic drainage bottom, slower
Indeed, mirroring the climate dependence of rates of on a less mesic north-facing slope, and rarely occurs at
biomass accumulation (Fig. 2a–b; Anderson et al., more arid sites prior to stand-clearing fire (Romme &
2006), it has been observed that the rate of N uptake by Knight, 1981). In the Czech Republic, the rate of succes-
a regrowing tropical forest in Costa Rica is four times sion in vegetative communities (including forests and
that of a regrowing temperate forest at Hubbard Brook, nonforests) is strongly influenced by climate; mean
USA (Russell & Raich, 2012). However, the trajectory of annual change in dominant species cover during the
forest recovery is also shaped by biogeochemistry– first 12 years of succession decreases dramatically with
climate interactions. For example, climate influences increasing elevation (increasing precipitation and

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
 FOREST RECOVERY UNDER CLIMATE CHANGE 5

decreasing temperature; Prach et al., 2007). These exam- experiments. Given the logistical difficulties of elevat-
ples provide evidence that climate strongly influences ing CO2 in forests with tall canopies, the majority of
both the rate and pathway of succession. these experiments have been performed on young
In addition to its influence on the rate and trajectory forests or trees, with only one FACE experiment in a
of forest recovery, climate also shapes the types of mature forest to date (Table S1).
steady-state conditions toward which secondary forests Elevated CO2 consistently enhances photosynthesis,
can eventually converge. Globally, aboveground or GPP at the ecosystem level (Ceulemans & Mousseau,
biomass of forests is influenced by temperature and 1994; Curtis & Wang, 1998; Ainsworth & Long, 2005;
precipitation (Fig. 2c–d; Keith et al., 2009; Anderson- Hyv€ onen et al., 2007). In young forests, this results in
Teixeira et al., 2011; Larjavaara & Muller-Landau, increased NPP and biomass; at least at the onset of the
2012), and climate strongly influences most other major experiment (DeLucia et al., 1999; Norby et al., 2005).
components of ecosystem C cycles, including gross and However, whereas substantial NPP and biomass
net primary productivity (GPP and NPP, respectively; increases have occurred at the onset of experiments,
Luyssaert et al., 2007), soil and whole-ecosystem respi- this NPP stimulation has persisted in some but not all
ration, and soil organic carbon (Raich & Schlesinger, forests (Oren et al., 2001; Seiler et al., 2009; McCarthy
1992; Jobb
agy & Jackson, 2000). Similarly, species diver- et al., 2010; Norby et al., 2010). Moreover, this response
sity varies globally with respect to climate (Brown et al., becomes less consistent as forests become older, and
1998). In sum, climate can determine the state to which NPP did not increase in the only mature forest exposed
forests converge following disturbance both directly to elevated CO2 (Fig. 3; K€ orner et al., 2005; Bader et al.,
through biophysical constraints and indirectly through 2009). Similarly, leaf and fine root biomass are consis-
its influence on biogeochemistry and the surrounding tently stimulated in young forests, but may decline in
metacommunity. old forests (Fig. 3; K€orner et al., 2005; Bader et al., 2009).
The broad-scale patterns described above demon- Thus, although there is strong evidence that CO2 fertil-
strate that climate strongly influences the rate of forest ization increases the rate of biomass accrual in young
recovery, successional pathways, and the structure and forests, a question remains as to whether elevated CO2
function of mature forests. However, transient dynam- increases the biomass and productivity of mature for-
ics under a rapidly changing climate may diverge from ests (Fig. 1; K€ orner et al., 2005; Hyv€ onen et al., 2007;
expectations based on these contemporary patterns, Norby & Zak, 2011). In large part because of this uncer-
and increasing atmospheric CO2 concentrations will tainty, it remains unclear whether the net carbon
alter physiological constraints on forests. To under- balance of mature forests will increase in response to
stand the more immediate responses of forest recovery CO2 fertilization (Fig. 3).
to elevated CO2 and climate change, we turn to experi- The ability of forests to sustain increased NPP under
mental manipulations. elevated CO2 as they age – and, ultimately, the potential
for mature forests to increase C storage under elevated
CO2 – depends in large part upon biogeochemistry. One
Forest responses to experimental manipulation of
potential explanation of observed declines in NPP stimu-
CO2, temperature, and precipitation
lation under elevated CO2 as forests age is that increased
Experiments manipulating CO2, temperature, and pre- productivity immobilizes nutrients in woody tissue or soil
cipitation demonstrate that altered climatic conditions organic matter such that soil N and other nutrients
will alter ecosystem and community dynamics in sec- needed to sustain growth become depleted and may
ondary forests. The responses of terrestrial ecosystems eventually limit growth (Luo et al., 2004). Progressive N
in general to these experimental manipulations have limitation can be alleviated through a variety of mecha-
been previously reviewed (e.g., Pendall et al., 2004; De nisms: trees can increase their N use efficiency, invest
Graaff et al., 2006; Norby & Zak, 2011; Wu et al., 2011; more C in belowground nutrient acquisition, or access
Beier et al., 2012; Dieleman et al., 2012; Lu et al., 2012); deep N pools (McKinley et al., 2009; Iversen, 2010; Drake
here, we focus specifically on the responses of forests et al., 2011b; Norby & Zak, 2011). N limitation can also be
and any age dependency of their responses (Fig. 3). mitigated if greater N mineralization occurs under high
CO2, or if N2 fixation is stimulated by elevated CO2
(Zanetti et al., 1996; Hungate et al., 2004; Luo et al., 2004;
Responses to CO2 fertilization
Hoosbeek et al., 2011; Norby & Zak, 2011). Early work
Tree-dominated ecosystems – all in temperate or boreal suggested that N mineralization declines under elevated
regions – have been exposed to elevated CO2 through CO2 (Hungate et al., 1999), but more recent studies sug-
Free-Air Carbon dioxide Enrichment (FACE), Open gest that there can be a priming effect through time from
Top Chamber (OTC), and Whole Tree Chamber (WTC) increased atmospheric CO2 that stimulates soil micro-

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
6 K . J . A N D E R S O N - T E I X E I R A et al.

Fig 3 Schematic diagram illustrating typical forest successional trajectories under ambient climate (solid lines; as reviewed in Table 1)
and how these are affected by experimental CO2 fertilization, warming, and drought (increases in blue, decreases in red; color satura-
tion scales with certainty). Responses to CO2 and climate change are based on a comprehensive review of experimental studies (Tables
S1–S3). Responses are considered to have high certainty when observed in multiple sites and low certainty when observed in only one
study. *Indicates a response that is time dependent; it may change from negative to positive as increased N mineralization stimulates
biomass growth (Melillo et al., 2011).

bial activity, which in turn degrades slowly cycling uncertain whether this can continue indefinitely or
organic matter pools and release mineral N (Carney whether NPP stimulation in all forests would eventu-
et al., 2007; Langley et al., 2009; Drake et al., 2011b). ally decline given sufficient time (Hyv€onen et al., 2007;
Elevated CO2 may also increase N availability by Norby & Zak, 2011).
increasing labile C to drive the energetics of N2 The responses of tree growth to elevated CO2 are
fixation (Hungate et al., 1999). However, in a variable among species (Bazzaz, 1990; Saxe et al.,
scrub-oak system in Florida, N2 fixation was negatively 1998; Pe~ nuelas et al., 2001; K€
orner et al., 2005; Seiler
impacted by 7 years exposure to elevated CO2 (Hun- et al., 2009; Dawes et al., 2011), and differential spe-
gate et al., 2004), and N2-fixation rates have continued cies responses have commonly been observed in
to decline – perhaps due to canopy closure and light CO2-enrichment experiments (Table S1). For exam-
limitation (Duval, 2010). Thus, N deficiency may be ple, of the three codominant canopy tree species
avoided – and NPP stimulation sustained – over time (Fagus sylvatica, Quercus petraea, Carpinus betulus) in
through a variety of mechanisms; however, it remains the mature deciduous forest exposed to elevated

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
 FOREST RECOVERY UNDER CLIMATE CHANGE 7

CO2, shade-tolerant Fagus exhibited increased annual

Responses to warming
basal area increments in response to CO2 in two of
four treatment years, whereas growth of the other Over the last three decades, several tree-dominated eco-
species remained the same or declined (K€ orner et al., systems of various ages – almost all in temperate and
2005). Similarly, proportional species’ contributions boreal regions – have been exposed to experimental
to whole-ecosystem productivity shifted in a Florida warming (Table S2). These experiments have warmed
scrub–oak ecosystems exposed to elevated CO2: either aboveground vegetation or the soil (through use
dominant Quercus myrtifolia exhibited strong biomass of buried cables); there are few soil-and-air warming
growth, Q. chapmanii exhibited less of an effect, and experiments done at the scale of canopy trees (Slaney
subdominant Q. geminata showed no growth stimu- et al., 2007; Bronson & Gower, 2010).
lation (Dijkstra et al., 2002). Thus, differential species Soil warming in northern forests results in faster
growth responses consistently alter proportional spe- decomposition and microbial processing of soil C and
cies’ contributions to whole-ecosystem productivity N, which directly releases more CO2 to the atmosphere
and will likely change the composition of future because of enhanced soil respiration (Table S2; Rustad
communities. et al., 2001; Melillo et al., 2002, 2011). By increasing N
Understory vegetation can influence ecosystem func- mineralization rates, soil warming can have an indirect
tioning and future community composition (Nilsson & N fertilization effect, which generally increases above-
Wardle, 2005) and, therefore, impacts of global change ground production and lowers C allocation to fine root
on juvenile trees and influential nontree species serve biomass (Fig. 3; Zhou et al., 2011). The net ecosystem C
as a window into the forests of the future. Moreover, as balance in response to warming depends largely on the
the majority of forest biodiversity is in the understory counteracting effects of C release through increased soil
stratum, impacts on understory species as well as sym- respiration and C sequestration through increased bio-
biotic mycorrhizal fungi bear consequences for tree mass growth (Fig. 3); in a 60- to 70-year-old even-aged
recruitment, carbon cycling, forest health and biodiver- oak–maple forest in central Massachusetts subjected to
sity (Gilliam, 2007). Understory community responses 7 years of soil warming (Harvard Forest), soil C losses
to CO2 enrichment have been commonly observed were increasingly offset by stimulated growth of can-
(Table S1). At ORNL-FACE, the woody understory opy trees (after a lag of several years; Melillo et al.,
increased in importance relative to the total stand and 2011; Butler et al., 2012). Additional changes may be
to herbaceous plants, indicating a potential acceleration driven by aboveground warming; over the first few
of succession under elevated CO2 (Souza et al., 2010). years of warming in a 12-year-old black spruce (Picea
Consistent with earlier work using pots and growth mariana) plantation in Manitoba, soil respiration
chambers (Bazzaz & Miao, 1993; Kubiske & Pregitzer, increased under soil warming but decreased under
1996; Kerstiens, 1998, 2001; H€ attenschwiler & K€ orner, soil-and-air warming (Bronson et al., 2008). In this
2000), CO2 enrichment at DukeFACE tended to favor study, elevated soil and air temperatures increased
slow-growing, shade-tolerant species with low rates of spruce tree shoot growth (Bronson et al., 2009) but did
productivity in understory conditions, again suggesting not change rates of photosynthesis or autotrophic respi-
that succession may be accelerated in temperate forests ration (Bronson & Gower, 2010). Much remains to be
under future conditions, with implications for bio- learned about how warming affects whole forested eco-
sphere–atmosphere carbon feedbacks (Mohan et al., systems, particularly in subtropical and tropical forests,
2007). In addition, CO2 enrichment may favor woody where only one warming experiment has been con-
vines (lianas; e.g., Sasek & Strain, 1990). This has been ducted to date (Cheesman & Winter, 2012). Moreover,
observed in two FACE studies (Table S1); for example, although we may posit that forest age modulates
at DukeFACE, the woody vine poison ivy (Toxicoden- warming responses based on the magnitude of struc-
dron radicans) growth was disproportionately enhanced tural and functional changes associated with forest
under elevated CO2 (Mohan et al., 2006). Lianas have recovery (Table 1), there is of yet no clear evidence that
been expanding in abundance in many regions of the the direction of forest responses to warming varies by
world – often to the detriment of recruiting and mature age (Fig. 3).
trees (Dillenburg et al., 1995; Ingwell et al., 2010; Schnit- Growth responses to warming vary among tree spe-
zer & Carson, 2010; Schnitzer & Bongers, 2011) – and cies (Table S2), and this is likely to affect successional
the positive feedback of elevated CO2 for vines may dynamics and forest composition. For example,
hinder the establishment of secondary forests globally. although the large oaks at Harvard forest accounted for
Thus, increasing atmospheric CO2 may substantially the majority of C uptake and storage in woody tissue,
alter the rate and pathway of succession as well as the smaller maples exhibited a greater stimulation of
composition of mature forest communities (Fig. 1). growth in response to soil warming stimulation (Mel-

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
8 K . J . A N D E R S O N - T E I X E I R A et al.

illo et al., 2011; Butler et al., 2012; Mohan et al., unpub-

Responses to multivariate environmental manipulation
lished results). Similarly, a warming experiment in a
recently timbered oak–hickory forest in Pennsylvania Joint effects of altered CO2, temperature, and precipita-
found altered phenology (with differential responses tion are rarely purely additive (Dieleman et al., 2012),
among species) and community composition (Rollin- and understanding the interactive effects produced by
son, 2010; Rollinson & Kaye, 2011). Thus, warming is combined manipulations remains an important chal-
likely to alter species’ growth and phenology and, lenge. Experimental manipulation of more than one of
thereby, the rate and pathway of succession and ulti- these elements in a factorial design has occurred in sev-
mately the community composition of mature forests eral intact forests or experimental mesocosms (Tables
(Fig. 1). S1–S3). These studies demonstrate that tree growth and
carbon cycling in young forests are generally acceler-
ated under combined higher CO2, warmer, and wetter
Responses to altered precipitation
conditions (Tables S1–S3; e.g., Wan et al., 2004;
There have been a number of precipitation manipula- Comstedt et al., 2006; Slaney et al., 2007; Tingey et al.,
tion experiments in forests of a range of ages spanning 2007; Bauweraerts et al., 2013). In addition, different
from boreal to tropical regions (Table S3; Beier et al., species have responded differently to different ele-
2012). Across this range of climates and forest ages, ments of climate change; for example, at the alpine tree
tree growth and survival were generally increased by line in Switzerland, growth of Larix decidual responded
water addition and reduced by water removal (Fig. 3; positively to CO2 but was unresponsive to soil warm-
e.g., Hanson et al., 2001; Nepstad et al., 2002; Plaut ing, whereas Pinus cembra had a slight positive response
et al., 2012; Vasconcelos et al. (2012), as was GPP or to warming but responded minimally to CO2 (Dawes
NPP at the ecosystem level (Nepstad et al., 2002; Al- et al., 2011). The limited number of studies and the
berti et al., 2007). Soil respiration rates also tended to complexity of multifactor experiments make it prema-
increase under irrigation and decrease under drought ture to generalize about how forests of different ages
(Table S3; e.g., Conant et al., 2000; Sotta et al., 2007). will respond to interactive elements of global change.
However, water addition only accelerated forest C However, observed responses of forests to the environ-
cycling up to a point; some more mesic forests did not mental change that has already occurred – to which we
respond to precipitation manipulation (De Visser et al., turn next – reveal how secondary forests are respond-
1994; Bergh et al., 1999) or had accelerated C cycling ing to multivariate climate change to date.
under reduced precipitation (Cleveland et al., 2010).
Sensitivity to precipitation manipulation often varied
Altered forest recovery under contemporary
by size class, but results were mixed as to whether
multivariate environmental change
small or large trees were more sensitive (Hanson et al.,
2001; Nepstad et al., 2007). Whereas altered precipita- Historical reconstructions indicate that forests of all
tion had a strong effect on seedling emergence and ages have responded to the changes in CO2 and climate
survival (Richter et al., 2012; Volder et al., 2012) and at that have already occurred. By nature, these historical
times had a stronger effect on small than on large trees records do not directly separate the effects of CO2, tem-
(Hanson et al., 2001), there were also instances where perature, and precipitation from one another and from
exposed canopy trees suffered greater drought-related other potentially confounding environmental changes
stress (Nepstad et al., 2007; Schuldt et al., 2011). Thus, (e.g., atmospheric deposition, ozone, management,
within-stand relationships between tree age and altered disturbance regimes). Rather, they provide a
drought sensitivity do not necessarily mirror across- picture of how the dynamics of forest recovery are
stand relationships, where growth sensitivity to varia- responding to contemporary multivariate environmen-
tion in water availability declines with stand age tal change.
(Table 1). As with experimental manipulation of CO2 Tree-ring and observational records extending back
and temperature, differential species responses were decades to centuries have demonstrated the climate
commonly observed under precipitation manipulation dependence of forest productivity. Tree-ring records
experiments (Table S3; e.g., Yavitt & Wright, 2008), have revealed increasing growth rates in numerous
portending future changes to community composition forests including high-elevation forests in western
under altered precipitation regimes. Thus, in sum- Washington (Graumlich et al., 1989), conifers in the
mary, precipitation manipulation experiments have white mountains of California (Lamarche et al., 1984;
demonstrated that water availability affects rates of Salzer et al., 2009), Pinus ponderosa forests in the US
forest recovery, mature forest states, and probably suc- Pacific northwest (Soul
e & Knapp, 2006), aspen (Popu-
cessional pathways (Figs 1 and 3). lus tremuloides) secondary forests in Wisconsin (Cole

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
 FOREST RECOVERY UNDER CLIMATE CHANGE 9

et al., 2010), Abies and Quercus forests in France (Bec-

ker, 1989; Becker et al., 1994), and numerous other for-
ests throughout Europe (Spiecker, 1999; Babst et al.,
2013). These increased growth rates are generally
attributable to increased atmospheric CO2, tempera-
ture, or moisture (e.g., Graumlich et al., 1989; Soul
e &
Knapp, 2006; Salzer et al., 2009; Cole et al., 2010). In
contrast, tree growth rates have decreased in response
to warming or drought stress in many other forests
around the world (Allen et al., 2010), including white
spruce (Picea glauca) in interior Alaska (Barber et al.,
2000), conifers in the southwest United States
(Williams et al., 2013), and tropical forests in Panama,
Malaysia, and Costa Rica (Feeley et al., 2007b; Clark
et al., 2010). Similarly, a review documenting evidence
of altered forest productivity over the last half century Fig. 4 Evidence of increasing rates of biomass accumulation in
indicates that the productivity of many forests is coastal Douglas fir (Pseudotsuga menziesii) and western hemlock
increasing whereas that of others is declining (Boisve- (Tsuga heterophylla) forests in southwest British Columbia, Can-
nue & Running, 2006). Although powerful for under- ada. Plotted is the residual average biomass increment (ΔB) from
standing the historical influence of climate on forest 1267 permanent inventory plots after correction for factors
productivity, these records are limited in that they do including stand age, site quality, nitrogen availability, and
not characterize responses across a range of forest biomass (the five lines represent different correction methods,
ages. as detailed in Hember et al., 2012). Linear regression represents
a significant positive trend. Reprinted from Hember et al. (2012).
By comparing biomass–age relationships (deter-
mined through a chronosequence approach) with cur-
rent biomass accumulation rates in forests of various many old-growth forests throughout the world appear
ages (determined through repeated sampling), a couple to be net C sinks (Baker et al., 2004; Luyssaert et al.,
of studies have identified accelerated biomass accumu- 2007; Chave et al., 2008; Lewis et al., 2009); however,
lation in forests following stand-clearing disturbances. this effect is diminished at larger spatiotemporal scales
Specifically, accelerated biomass accumulation has been of measurement (Clark, 2002; Feeley et al., 2007a; Chave
observed in temperate deciduous forests in the eastern et al., 2008). There is strong evidence of directional
United States (McMahon et al., 2010a; see also Foster change in community composition of forests through-
et al., 2010; McMahon et al., 2010b) and in temperate- out the world; for example, long-term records from the
maritime forests in the Pacific northwest (Fig. 4; Hem- 50 ha forest dynamics plot on Barro Colorado Island,
ber et al., 2012). Likely explanations of these increases Panama, indicate increased dominance of drought-
in secondary forest biomass accumulation rates include tolerant species (Feeley et al., 2011). In addition, there
increased atmospheric CO2, increased temperature, have been general increases in forest die-back globally
increased moisture, and increased growing season – a phenomenon attributed to climate change–type
length (McMahon et al., 2010a; Hember et al., 2012). drought (Breshears et al., 2005; Allen et al., 2010;
Thus, climate change appears to be increasing the rate Williams et al., 2013).
of forest regrowth in some temperate forests; however, Thus, there is evidence of historical change in both
parallel studies have yet to be conducted in other rates of forest regrowth and the state of forests that
regions. matured under past climates. There is also some evi-
The long time frame of forest recovery precludes dence of changing successional trajectories driven by
comparison of forests that have matured under differ- altered community dynamics (reviewed below).
ent climates, which would be necessary to determine Although concurrent changes in multiple environmen-
whether climate change is altering recovery trajectories tal factors including atmospheric CO2, climate, atmo-
such that forests converge to an altered state as they spheric deposition, herbivore communities, disturbance
mature (Fig. 1). However, long-term monitoring of regimes, and management make it difficult to isolate
mature forests can provide evidence as to whether the cause of these changes, their global distribution and
climate change is affecting the state of forests that directional correlation with trends in CO2 and climate
matured under past climates. Long-term monitoring of provide strong evidence that they are at least partially
old-growth forests provides mixed evidence as to attributable to increasing atmospheric CO2 and climate
whether their total carbon storage capacity is changing; change.

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
10 K . J . A N D E R S O N - T E I X E I R A et al.

may have dramatic consequences for ecosystem func-

Community dynamics and the potential for state
tioning. For example, in Alaskan boreal forests, increas-
changes
ing fire frequency and severity have shifted
Successional pathways may be altered when elevated competitive dominance from conifers to deciduous spe-
CO2 and climate change affect community dynamics, cies, affecting biomass and soil C accumulation, albedo,
either directly through differential effects on the perfor- and energy partitioning (Beck et al., 2011).
mance of various species and size classes or indirectly Disturbance can also provide niche opportunities for
through altered disturbance regimes and consequent invaders (Shea & Chesson, 2002), and the successful
competitive outcomes. invasion can dramatically alter successional trajectories
Climate change will alter community dynamics by and also feedback to further modify disturbance
altering the physical environment in which species of regimes (Mack & D’Antonio, 1998). Although distur-
varying physiological strategies are competing. Experi- bances are commonly believed to increase invader suc-
mental climate change manipulations (reviewed above; cess, recent work suggests that it is changes to
Tables S1–S3) and decades of forestry research on the disturbance regimes, rather than disturbance events per
climate sensitivity of forest regeneration (Fowells & se that most strongly influence a communities’ suscepti-
Stark, 1965; Ferrell & Woodard, 1966; Thomas & Wein, bility to invasions (Moles et al., 2012). In this light,
1985) have demonstrated that increased CO2 and climate change is likely to change the composition of
altered climate will differentially affect growth rates of some communities by altering disturbance regimes to a
trees by size and species, thereby altering population point where invader species can become dominant.
dynamics, competitive interactions, and species compo- In some cases, climate change may push forests past
sition of both young and mature forests. In addition, critical thresholds such that, upon perturbation, they
climate change can differentially favor or inhibit com- undergo drastic changes in community composition
mon forest pathogens, providing another mechanism of and ecosystem properties (‘catastrophic shift’) and fail
impact on community structure (reviewed by Sturrock to return to their previous state (Fig. 5). In many sys-
et al., 2011). Such community changes can affect ecosys- tems, the observed state of the community is not the
tem function, altering production, C stocks, and biogeo- only possible stable state; a variety of empirical results
chemistry. For example, model predictions of climate demonstrate the existence of alternative stable states in
change effects on forest productivity can be very differ- nature (D’Antonio & Vitousek, 1992; Savage & Mast,
ent if the community is allowed to develop dynami- 2005; Schr€ oder et al., 2005; Odion et al., 2010; Scheffer
cally, compared with using parameters based on et al., 2012). Large changes in the environment can
average forest characteristics, which is a common prac- bring about large changes in ecosystems, but smooth,
tice in biogeochemical models examining the effects of gradual changes in abiotic conditions also can cause
climate change (Bolker et al., 1995). abrupt shifts in ecosystem properties and functioning
Beyond its direct effects on the dynamics of forest (Scheffer et al., 2001). Systems that are structured by
recovery through physiological mechanisms, climate disturbance and are susceptible to abiotic forcing (such
change may also impact successional pathways indi- as regenerating forests) may be more likely to display
rectly by altering the frequency, timing, severity, and alternative stable states (Didham et al., 2005). When dis-
spatial extent of disturbances including fires, droughts, turbance keeps systems in perpetual flux, as is the case
storms, floods, and herbivore or pathogen outbreaks for many forests, no true stable equilibrium (in the clas-
(e.g., Dale et al., 2001; Westerling et al., 2006; Allen sical, dynamical systems sense) is reached. Instead, for-
et al., 2010; Sturrock et al., 2011). Frequency and inten- ests undergo periodic cycles of disturbance and
sity of disturbance have been theoretically shown to regeneration, and it is these cycles that constitute the
have very different effects on community diversity ‘state’ of the system.
(Miller et al., 2011; in determining microcosm diversity Many forests are resilient (sensu Grimm & Wissel,
(Hall et al., 2012). Thus, changes in disturbance frequency 1997) to commonly experienced disturbances, but
and intensity have the potential to shift community effects of climate change, such as changes to the distur-
composition, even when species are restricted to (and bance and precipitation regimes, can change the
are still viable in) their historic ranges. Moreover, large composition and productivity of forest communities
changes to disturbance regimes are not required to (Thompson et al., 2009), forcing the system into
facilitate changes in community composition; in the different cyclical behaviors. Although different initial
annual plant model of Miller et al. (2011), changing trajectories can lead to different mature forest states,
disturbance mortality by just a few percentage points there is also the possibility that different initial trajecto-
can send a species to extinction. Changes in community ries can lead to the same mature state, or that similar
composition driven by altered disturbance regimes initial trajectories can lead to distinct mature states

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
 F O R E S T R E C O V E R Y U N D E R C L I M A T E C H A N G E 11

of Oregon and California, high-intensity fire shifts the

community from a high-biomass mixed conifer forest
to a pyrogenic low-biomass shrub–chaparral–hard-
wood community, in which state it may be maintained
by subsequent fires of any intensity (Odion et al., 2010;
Thompson & Spies, 2010). Alternative stable states may
also be driven by hydrologic, microclimatic, or biogeo-
chemical mechanisms; for example, postfire forest
resilience may be impacted by changes to soil biogeo-
chemistry and hydrological functioning (Ffolliott et al.,
2011; Smithwick, 2011).
Climate change may gradually alter the landscape of
alternative states, having minimal impact on mature
forests, but shifting conditions such that forests will be
unlikely to reestablish following disturbance (Fig. 5).
The probability of forest regeneration may be reduced
by mechanisms such as reduced probabilities of seed-
ling establishment under more arid conditions, reduced
competitive advantage of seedlings relative to grasses
Fig. 5 Schematic diagram illustrating the potential for or shrubs, or increases in disturbance frequency or
disturbance to force ecosystems from one stable state to another severity. Although ecological theory points toward the
as the climate changes. Colored shapes represent the landscape risk that some forests may unlikely to return to their
of stable ecosystem states under different climate regimes, and previous state following stand-clearing disturbance as a
balls represent states in which ecosystems can stably exist – in result of global change (Fig. 5), empirical evidence
this case, the state to which the system converges at maturity remains scant. In the southwestern United States, pon-
(which will be associated with a stable disturbance–recovery derosa pine forests meet the criteria for forests that may
regime). The plot below illustrates hysteresis, wherein alternate be vulnerable to climate change-induced catastrophic
stable states exist. As the climate changes, basins of attraction
shifts and are often failing to reestablish following fire
shift such that the stable state at maturity eventually switches
(Dore et al., 2008; Roccaforte et al., 2012); however, a cli-
from one state to the other. During the transition, however, dis-
turbance (indicated by black arrow) may hasten the shift from
mate change mechanism has not been demonstrated.
one stable state to another. Modified from Scheffer et al. (2001). Understanding the potential for climate change to dra-
matically alter or prevent postdisturbance recovery
remains an important challenge.
(Fig. 1, ‘alternative transient states’ sensu Fukami & Thus, based on empirical findings and theoretical
Nakajima, 2011). concepts, community composition and ecosystem func-
In some forested regions, the existence of alternative tion of regenerating forests under climate change are
stable states implies that forests may not return to their likely to change, both quantitatively, and in terms of
previous state following disturbance. A general mecha- stability. Given that climate change, disturbance
nism underlying such alternative stable states is that regimes, and community dynamics interact in complex
seedlings and young forests are often more vulnerable ways to shape ecosystems, correctly predicting the
to disturbances such as drought, herbivory, and fire behavior of forests over the next century will require
than their mature counterparts (Table 1; Stromayer & greater understanding of the potential for altered com-
Warren, 1997; Thompson & Spies, 2010). As a result, munity dynamics to dramatically impact carbon
conditions that support the persistence of mature cycling, biogeochemistry, and ecosystem–atmosphere
forests may not be amenable to forest regeneration. For exchanges.
example, following fire, conifer regeneration may be
delayed or prevented by drought or competitive inhibi-
Ecosystem and earth system model projections
tion by grasses or shrubs (Savage & Mast, 2005;
Roccaforte et al., 2012). There are also systems in which Ecosystem and earth system models (ESMs) provide a
postfire establishment of pyrogenic vegetation or means to project dynamically how ecosystems will be
vulnerability of young stands to crown fire reduces the impacted by multiple interacting environmental
probability of forest regeneration (D’Antonio & Vitousek, changes over spatiotemporal scales that exceed the
1992; Savage & Mast, 2005; Thompson & Spies, 2010; limits of observation and experimentation. ESMs vary
Staver et al., 2011). For example, in the Klamath region in complexity from fully coupled global circulation

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
12 K . J . A N D E R S O N - T E I X E I R A et al.

models (GCMs), which include two-way feedbacks growth vary by model, region, and climate change sce-
between the land, atmosphere, and oceans to make pre- nario; the direction of change in forest growth is
dictions about climate, to simpler models with less expected to vary regionally and to depend on the
interaction between the earth system components (e.g., course of atmospheric CO2 and climate change
one-way feedbacks to the atmosphere such as land- (Kirilenko & Sedjo, 2007). For example, in lodgepole
cover changes to net terrestrial CO2 uptake). ESMs pine (Pinus contorta) forests regenerating from fire in
include land components embedded with physiological the Yellowstone region, woody production, live
and biogeochemical mechanistic representations of the biomass, N mineralization, and total ecosystem C are
interactions between vegetation, the atmosphere, and projected to increase under two different future climate
either prognostic disturbances (i.e., fire) or prescribed scenarios, with percent increase depending on the cli-
disturbances (i.e., harvest). Vegetation is represented in mate scenario (Smithwick et al., 2009). Thus, models
terms of broadly defined plant functional types (PFTs; demonstrate likely changes in forest productivity under
e.g., temperate conifers). When coupled with a specific future climates; however, without giving specific atten-
class of ecosystem models (dynamic global vegetation tion to changes in physiology and C allocation with for-
models; DGVMs), processes are included that allow est age, they say little about the responses of
vegetation type to change based on climate conditions regenerating forests specifically.
(e.g., forest to grassland or woodland). Recent advance- Because disturbance type, severity, size, and fre-
ments to some ESMs (CESM/CLM4.0, ORCHIDEE, quency affect postdisturbance C dynamics and biogeo-
TEM) now include dynamic response variables for the chemical cycling (Amiro et al., 2010; Smithwick, 2011),
long-term physiological changes related to CO2 and/or future trajectories of forest recovery are likely to be dri-
temperature (Krinner et al., 2005; Thornton et al., 2007; ven by climate change–disturbance type interactions. In
Zaehle & Friend, 2010). The complexity of these models, most models, disturbance events are generally imple-
and the variety of factors upon which model predic- mented by altering forest biomass pools through remo-
tions and associated uncertainty depend, preclude the vals (harvest), combustion (fire), or transfer of live to
possibility of any one model incorporating all of the dead material (insect outbreaks), with the amount
known complexity of forest regeneration. However, for transferred scaled to disturbance severity. For fire and
models to make predictions about forest recovery insect outbreaks, the timing of transfer of biomass to
following disturbance, they need to be able to capture litter and forest floor components varies because tree
the interactive effects of changing environmental condi- death can occur slowly, and snag fall rates are depen-
tions and disturbance on forest recovery dynamics. dent on a variety of factors including forest type
No model pays detailed attention to the roles of for- (Campbell et al., 2007; Edburg et al., 2011). At this time,
est age and successional changes in species composition we are unaware of any model capable of representing
in shaping the dynamics of forest recovery. Rather, the specific dynamics (e.g., recruitment, altered hydrol-
regenerating forests are generally parameterized as ogy, or biogeochemistry) associated with distinct dis-
mature forests, although sometimes there are two age turbance types, severities, and sizes. Therefore, models
classes (e.g., fire BGC; Smithwick et al., 2009), and car- currently say little about how changing disturbance
bon allocation to wood may vary dynamically with age severity and size are likely to impact forests; however,
(e.g., CLM4; Hudiburg et al., 2013). We are aware of they do reveal how altered disturbance frequency is
only one model where C allocation to nonwoody likely to impact forests. For example, in the Yellow-
components or physiology changes dynamically with stone region, fire burn area and frequency are projected
age (and this improves performance in describing age to increase under a range of future climate scenarios,
trajectories of woody productivity; Davi et al., 2009). quite possibly to the extent that current forest commu-
Changes in community composition (i.e., PFTs), physi- nities will have insufficient time to recover before the
ological differences between early- and late- succes- next fire event, making the current suite of conifer spe-
sional species, and age structure within a forest cies unlikely to persist (Westerling et al., 2011). Thus,
(Table 1) generally are not incorporated (exception is models demonstrate that climate change is likely to
ED2; Medvigy et al., 2009). As a result, models have have significant impacts on forested landscapes
difficulty accurately reproducing trajectories of change through its influence on disturbance regimes.
in biomass or other components of the C cycle associ- Despite their uncertainties, ESMs have demonstrated
ated with forest age (Table 1). Nevertheless, to the that forest recovery will be substantially altered under
extent that forest responses are consistent across age future climates. Rates of recovery will change, with
classes (Fig. 3), models can predict productivity direction and magnitude varying regionally and
responses of young forests to elevated CO2 and climate depending on future courses of atmospheric CO2 and
change. Simulated climate change effects on forest climate change. Altered disturbance regimes will inter-

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
 F O R E S T R E C O V E R Y U N D E R C L I M A T E C H A N G E 13

act with altered recovery trajectories, at times driving challenge of central importance is in understanding
biome shifts (Westerling et al., 2011). In combination, how climate change responses of young forests – on
the direct and indirect effects of climate change are pre- which the majority of manipulative experiments have
dicted to have substantial impacts on regional C bal- been performed (Tables S1–S3) – relate to the ultimate
ances and forestry operations; for example, harvestable state of these ecosystems once they reach ‘maturity’
forest biomass in Canada is projected to be reduced (Fig. 1). For instance, we do not know whether
26–62% for the 21st century, depending on the model increased biomass accumulation in young forests will
assumptions of predicted growth rate, soil carbon translate to increased biomass of old forests or whether
decay rate, and area burned by fire (Metsaranta et al., these forests will simply attain maximum biomass fas-
2011). However, specific representation of physiological ter. The effect of elevated CO2 on mature forest biomass
and community changes associated with forest age and total ecosystem C remains uncertain, although
(Table 1) will be required to understand how forest decreases in either are unlikely; meanwhile, elevated
recovery trajectories will be altered by climate change CO2 is very likely to result in increased nutrient limita-
(Fig. 1). tion (Fig. 3; Table S1). Likewise, it remains unclear how
warming will affect mature forest biomass and ecosys-
tem C stocks (Fig. 3); it is likely that aboveground C
Conclusions
stocks will increase in northern climates (Fig. 2c) while
As reviewed above, there is strong evidence that soil C stocks decrease and N mineralization increases
increasing atmospheric CO2, warming, and altered (Table S2). In contrast, changes in water availability
precipitation regimes will alter trajectories of forest have predictable effects; reduced water availability will
recovery. This conclusion is supported by global patterns reduce productivity, live biomass, and total ecosystem
in both forest regrowth rates and biomass of mature C stocks (Figs 2d and 3; Table S3). In all cases, altered
forests (Fig. 2); responses of forests of various ages to community composition is very likely (Table S3).
CO2, temperature, and precipitation manipulation Responses of mature forest states to combined changes
(Fig. 3; Tables S1–S3); observations of altered forest in CO2, temperature, and precipitation will vary region-
recovery under contemporary multivariate environ- ally, and understanding how the states toward which
mental change (Fig. 4); our understanding of succes- future forests will converge as they recovery from dis-
sional community dynamics and alternative stable turbance (Fig. 1) remains an important challenge.
states (Fig. 5); and models. Because forests undergo Climate change is also likely to impact pathways of
major structural, physiological, biogeochemical, and forest recovery (Fig. 1; Table 2), which may occur
compositional changes as they age (Table 1), it is logical through a variety of mechanisms including altered bio-
that responses to climate change vary as a function of geochemistry (e.g., decreased N limitation during early
forest age (Fig. 3). Depending on differential responses stages due to increased N mineralization), changing
of forests of different ages, climate change can impact biophysical constraints (e.g., reduced frequency of
rates of forest recovery, states of mature forests, and/or years with enough precipitation to support seedling
recovery pathways (Fig. 1, Table 2), and understanding establishment), or altered community dynamics. As
the impact of climate change on forests therefore reviewed above, different species within the same com-
requires attention to the role of forest age (Fig. 1). munity commonly have substantially different
Through its influence on young forests, climate responses to altered CO2 or climate (K€ orner et al., 2005;
change will impact rates of forest recovery (Fig. 1, Mohan et al., 2006, 2007; Seiler et al., 2009; Dawes et al.,
Table 2). Multiple lines of evidence point to accelerated 2011), and consequent changes to community structure
regrowth in mesic northern forests under future climates may impact ecosystem functioning in ways that cannot
(Figs 2–4; Tables S1–S2); however, responses of tropical be predicted based solely on characteristic physiologi-
forest regeneration rates to elevated CO2 and increasing cal responses of dominant taxa (Bolker et al., 1995). For
temperature remain uncertain. For forests globally, there example, increased liana biomass under future climates
is strong evidence that biomass accumulation rates will could meaningfully reduce forest biomass (Phillips
decrease under more arid conditions (Figs 2b and 3; et al., 2002; Mohan et al., 2006; Ingwell et al., 2010). Dif-
Table S3) – sometimes to the point where forests may ferential responses are likely to be most influential early
never recover (Fig. 5). Changes to rates of nutrient in succession, when species turnover rate is highest and
accumulation in biomass, biogeochemical cycling, and trees are most sensitive to environmental variation
community change are likely to parallel responses of (Table 1), and may have an enduring influence on com-
biomass accumulation rate (Table 2). munity composition and ecosystem function (D’Anto-
Climate change will also impact the state toward nio & Vitousek, 1992; Bunker et al., 2005; Beck et al.,
which forests converge as they age (Fig. 1; Table 2). A 2011; Hooper et al., 2012).

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
 Table 2 Probable climate change impacts on trajectories of several forest properties following disturbance (sensu Fig. 1)

Expected response to climate change

Recovery
Forest property trajectory Elevated CO2 Elevated temperature* Altered water availability Multivariate change

Biomass Rate of Very likely increase. Very likely increase in temperate Very likely increase with water Likely increase in temperate and
change and boreal forests; Uncertain availability/decrease with boreal forests (absent moisture
response in tropics. drought stress. stress); Likely decrease under
drought stress; Uncertain response
in tropics.
Mature Uncertain (likely Possible changes in some Very likely increase with water Likely changes (region and time
state increase or no change) regions (e.g., increase in cold availability/decrease with frame specific).
regions) drought stress.
Pathway Likely changes driven Likely changes driven by shifts Likely changes driven by shifts Likely changes driven by shifts in
by shifts in community in community dynamics, in community dynamics, community dynamics,
dynamics, biogeochemistry, or biophysics. biogeochemistry, or biophysics; biogeochemistry, or biophysics.
14 K . J . A N D E R S O N - T E I X E I R A et al.

biogeochemistry, or High potential for catastrophic

biophysics. shift to low-biomass (nonforest)
state under severe drought
conditions.
Total C stock Rate of Very likely increases. Likely increase in temperate and Very likely increase with water Likely increase in temperate and
change boreal forests; Uncertain availability/decrease with boreal forests (absent moisture
response in tropics. drought stress. stress); Likely decrease under
drought stress; Uncertain response
in tropics.
Mature Uncertain (likely increase Very likely decrease in soil Very likely increase with water Likely changes (region and time
state or no change). organic matter, possible availability/decrease with frame specific).
increase in biomass carbon drought stress.
(higher latitudes). Net balance
uncertain and likely region
specific.
Pathway Likely changes driven by Likely concurrent reductions in Likely changes driven by shifts Likely changes driven by shifts in
shifts in community soil organic C and increases in in community dynamics, community dynamics,
dynamics, biomass C. biogeochemistry, or biophysics; biogeochemistry, or biophysics
biogeochemistry, or High potential for catastrophic (region- and time frame specific).
biophysics. shift to low C (nonforest) state
under severe drought
conditions.
Biogeochemistry Rate of Likely acceleration of Likely acceleration of nutrient Likely acceleration of nutrient Likely acceleration of nutrient
change nutrient accumulation in accumulation in vegetation in accumulation in vegetation accumulation in vegetation in
vegetation; Likely northern forests; Likely with increased water mesic northern forests; Likely rate
acceleration of C & N acceleration of C & N cycling in availability/rate decrease with decrease under drought stress;
temperate and boreal forests; drought stress. Uncertain response in tropics.

gcb.12194
Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
 Table 2 (continued)

Expected response to climate change

Recovery

gcb.12194
Forest property trajectory Elevated CO2 Elevated temperature* Altered water availability Multivariate change

cycling in temperate and Uncertain response in the

boreal forests. tropics.
Mature Likely acceleration of C & Likely acceleration of C & N Likely decrease in nutrient Likely acceleration of C & N cycling
state N cycling in temperate cycling in temperate and boreal limitation under drought stress. in temperate and boreal forests;
and boreal forests; Likely forests; Likely decrease in soil N Likely decrease in soil N pool;
decrease in soil N pool; pool; Uncertain response in the Possible increase in total nutrient
Uncertainty changes tropics. storage in vegetation.
total nutrient storage in
vegetation.
Pathway Likely increase in nutrient Likely concurrent reductions in Likely changes driven by shifts Likely increase in nutrient
limitation. soil N and increases in biomass N. in community dynamics. limitation in mesic forests.
Community Rate of Likely acceleration of Likely acceleration of Uncertain (drought may reduce Likely acceleration in temperate
composition change community change. community change; Uncertain the rate of community change and boreal forests (absent moisture
response in the tropics. through retarded plant growth stress); Uncertain response in
or enhance it by increasing tropics.
mortality).
Mature Very likely alteration of Very likely alteration of mature Very likely alteration of mature Very likely alteration of mature
State mature community community composition driven community composition driven community composition driven by
composition driven by by differential species by differential species differential species responses;
differential species responses. responses. Likely increase in non-native
responses; Likely species.
increase in liana
abundance.
Pathway Very likely alteration Very likely alteration driven by Very likely alteration driven by Very likely alteration driven by
driven by differential differential size- and species differential size- and species differential size- and species
size- and species responses, sometimes causing responses, sometimes causing responses, sometimes causing shift
responses, sometimes shift to alternative state. shift to alternative state. to alternative state.
causing shift to
alternative state.

*Elevated temperature responses assume no change in moisture stress; responses to changes in water availability are listed in ‘Altered water availability’ column.
‘Rate of change’ refers to the rate at which the forest approaches its mature state. ‘Mature state’ refers to the state to which forests converge as they age. ‘Pathway’ refers to the
sequence of states through which any given ecosystem property passes and the relative amount of time spent in each.
F O R E S T R E C O V E R Y U N D E R C L I M A T E C H A N G E 15

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
16 K . J . A N D E R S O N - T E I X E I R A et al.

In the most dramatic cases, altered successional long-term monitoring of mature forests (e.g., Chave
pathways may result in catastrophic shifts to an et al., 2008), we are aware of only one study manipulat-
alternate stable state (e.g., a forest to grassland tran- ing CO2 or temperature at the whole-tree level in a field
sition; Fig. 5). There are documented instances setting in the tropics (Cheesman & Winter, 2012). This
where, following disturbance, young forests fail to constrains our ability to predict climate change
establish or persist under conditions that are amena- responses of tropical forests. Understanding how
ble to persistence of mature forests (Thompson & climate change will affect tropical forest regeneration is
Spies, 2010; Roccaforte et al., 2012). When these con- particularly important given the widespread use of
ditions are linked to climate, as they often are (e.g., slash-and-burn agriculture in the tropics and the signif-
sufficient moisture for seedling establishment, fire icant role of tropical forest regrowth in the global C
regime), climate change is likely to force a transition cycle (Pan et al., 2011).
to an alternate stable state (Fig. 5). As a result, An additional challenge lies in improved representa-
directional changes to forest ecosystems that would tion of forest recovery dynamics in ESMs, which are
happen gradually in the absence of disturbance may currently simplistic in their treatment of forest recovery
be greatly accelerated by disturbance (Fig. 5). dynamics. Although detailed representation of forest
There remain several important unanswered ques- recovery dynamics in global models is infeasible, we
tions regarding the impact of climate change on the believe that two advances will be important to improv-
dynamics of forest recovery: ing the treatment of forest regeneration. First, the most
(1) How does forest age modulate responses to climate important stand age-dependent physiology and alloca-
change? Forests of different ages have responded differ- tion strategies (driven by aging of dominant species
ently to climate manipulations (Fig. 3; Tables S1–S3); and changes in species composition) should be identi-
however, at this point climate manipulation experi- fied and incorporated. This will allow improved repre-
ments provide only circumstantial evidence of age dif- sentation of the dynamics of forest recovery in current
ferences in climate change response. Systematic and future climates. Importantly, this will help to iden-
comparison of responses of forests of different ages to tify situations where young forests fail to establish
experimental CO2 or climate manipulation and to natu- despite the persistence of their mature counterparts,
ral climate variability will be crucial to understanding suggesting climate change-driven regime shifts (Fig. 5).
and modeling climate change impacts on forests of all Second, although modeling individual species in ESMs
ages. is infeasible, it will be necessary to represent the conse-
(2) How will successional trajectories differ under future quences of demonstrated variability in species
climates? Beyond understanding how age modulates responses to climate change and inevitable resultant
forest responses to climate change, we face the chal- shifts in community composition and ecosystem pro-
lenge of understanding how climate change will impact cesses. With changing community composition, the net
entire trajectories of forest recovery (Fig. 1). It is impor- ecosystem response may differ significantly from that
tant to note that, because the climate history under which would be predicted based on mean characteris-
which a stand has developed affects its current state tics of the original community (Bolker et al., 1995). In
and future trajectory, changes to entire trajectories can- the most dramatic cases, altered competitive interac-
not be understood simply by integrating across tions may result in a regime shift from forest to a grass-
responses of forests different ages. Rather, it will be or shrub-dominated state (Fig. 5). Predicting regime
important to understand how altered biogeochemical shifts in ESMs will be particularly important, as these
dynamics and community composition shape succes- imply feedbacks to the climate system through altered
sional pathways and the states toward which forests C storage, albedo, and hydrology.
converge as they mature. Changes in the dynamics of forest recovery following
(3) Where and when will state changes occur? Climate disturbance will result in potentially significant climate
change-driven regime shifts (Fig. 5) will have dramatic feedbacks. Altered disturbance–recovery dynamics may
consequences, yet they remain difficult to document impact the C cycle enough to reverse the sign of a regio-
and predict. There is a need for experimental, observa- nal C cycle feedback (Kurz et al., 2008; Running, 2008;
tional, and modeling studies to identify the conditions Metsaranta et al., 2011). Moreover, albedo and evapo-
under which such shifts are likely and the mechanisms transpiration are important components of the climate
through which they may occur. regulation services of ecosystems (Anderson-Teixeira
(4) How will tropical forest regeneration respond to cli- et al., 2012), change systematically over the course of for-
mate change? Although tropical forests are well repre- est recovery (Randerson et al., 2006; Kirschbaum et al.,
sented in global-scale comparisons (Fig. 2), 2011; Jin et al., 2012; O’Halloran et al., 2012), and may
precipitation manipulation experiments (Table S3), and shift substantially in response to climate change – partic-

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
 F O R E S T R E C O V E R Y U N D E R C L I M A T E C H A N G E 17

ularly if the new community differs dramatically from Anderson KJ (2007b) Temporal patterns in rates of community change during succes-
sion. American Naturalist, 169, 780–793.
the old (Beck et al., 2011). Altered forest recovery Anderson KJ, Allen AP, Gillooly JF, Brown JH (2006) Temperature-dependence of bio-
dynamics will result in particularly strong feedbacks to mass accumulation rates during secondary succession. Ecology Letters, 9, 673–682.
climate change when a critical threshold is passed such Anderson-Teixeira KJ, DeLucia EH (2011) The greenhouse gas value of ecosystems.
Global Change Biology, 17, 425–438.
that forests fail to recover (Fig. 5), resulting in dramatic Anderson-Teixeira KJ, Vitousek PM (2012) Ecosystems. In: Metabolic Ecology: a Scaling
reductions in C storage and altered biophysical proper- Approach (eds. Sibley RM, Brown JH, Kodric-Brown A), pp. 99–111. Wiley-Black-
ties. For example, in semiarid regions such as the US well, Chichester.
Anderson-Teixeira KJ, Vitousek PM, Brown JH (2008) Amplified temperature depen-
southwest, current forest communities may not be sup- dence in ecosystems developing on the lava flows of Mauna Loa, Hawai’i. PNAS,
ported under future more arid conditions and may not 105, 228–233.
re-establish following disturbance, resulting in a posi- Anderson-Teixeira KJ, Delong JP, Fox AM, Brese DA, Litvak ME (2011) Differential
responses of production and respiration to temperature and moisture drive the car-
tive C cycle feedback (Breshears et al., 2005; Williams bon balance across a climatic gradient in New Mexico. Global Change Biology, 17, 410.
et al., 2010, 2013; Anderson-Teixeira et al., 2011; Rocca- Anderson-Teixeira KJ, Snyder PK, Twine TE, Cuadra SV, Costa MH, DeLucia EH
forte et al., 2012). Thus, recently disturbed forests may (2012) Climate-regulation services of natural and agricultural ecoregions of the
Americas. Nature Climate Change, 2, 177–181.
play a key role in shaping terrestrial feedbacks to climate Babst F, Poulter B, Trouet V et al. (2013) Site- and species-specific responses of forest
change. growth to climate across the European continent. Global Ecology and Biogeography,
This review has demonstrated that the dynamics of in press.
Bader M, Hiltbrunner E, K€ orner C (2009) Fine root responses of mature deciduous
forest recovery are likely to be significantly impacted forest trees to free air carbon dioxide enrichment (FACE). Functional Ecology, 23,
by rising atmospheric CO2 and climate change. This 913–921.
will have repercussions for biodiversity, climate, and Baker TR, Phillips OL, Malhi Y et al. (2004) Increasing biomass in Amazonian forest
plots. Philosophical Transactions of the Royal Society of London Series B-Biological Sci-
even economics, as the forestry industry and emerging ences, 359, 353–365.
woody bioenergy industry stand to be affected by Baldocchi D (2008) “Breathing” of the terrestrial biosphere: lessons learned from a
altered forest regeneration rates (Kirilenko & Sedjo, global network of carbon dioxide flux measurement systems. Australian Journal of
Botany, 56, 1–26.
2007; Metsaranta et al., 2011; Hanewinkel et al., 2013). Barber VA, Juday GP, Finney BP (2000) Reduced growth of Alaskan white spruce in
Because the course of forest recovery shapes forest the twentieth century from temperature-induced drought stress. Nature, 405,
structure and function for decades or centuries, climate 668–673.
Bauweraerts I, Wertin TM, Ameye M, McGuire MA, Teskey RO, Steppe K (2013) The
change impacts on secondary forests will have a lasting effect of heat waves, elevated [CO2] and low soil water availability on northern
legacy. Although the proportion of recently disturbed red oak (Quercus rubra L.) seedlings. Global Change Biology, 19, 517–528.
forests is relatively small at any given time, disturbance Bazzaz F (1979) Physiological ecology of plant succession. Annual Review of Ecology
and Systematics, 10, 351–371.
eventually affects all forests, and the proportion of for- Bazzaz F (1990) The response of natural ecosystems to the rising global CO2 levels.
ests that have regenerated under altered climate condi- Annual Review of Ecology and Systematics, 21, 167–196.
tions will steadily grow. In these ways, climate change Bazzaz FA, Miao SL (1993) Successional status, seed size, and responses of tree seed-
lings to CO2, light, and nutrients. Ecology, 74, 104–112.
will broadly impact forested regions through its influ- Bazzaz FA, Pickett STA (1980) Physiological ecology of tropical succession: a compar-
ence on forest recovery dynamics. ative review. Annual Review of Ecology and Systematics, 11, 287–310.
Beck PSA, Goetz SJ, Mack MC, Alexander HD, Jin Y, Randerson JT, Loranty MM
(2011) The impacts and implications of an intensifying fire regime on Alaskan bor-
Acknowledgements eal forest composition and albedo. Global Change Biology, 17, 2853–2866.
Becker M (1989) The role of climate on present and past vitality of silver fir forests in
Thanks to Jonathan Thompson and Sean McMahon for helpful
the Vosges mountains of northeastern France. Canadian journal of forest research, 19,
discussion and to three anonymous reviewers for helpful com- 1110–1117.
ments. This research was supported by DOE grant Becker M, Nieminen T, G
er
emia F (1994) Short-term variations and long-term
#DE-SC0008085 to KJAT, EHD, and BDD. changes in oak productivity in northeastern France. The role of climate and atmo-
spheric CO2. Annales des Sciences Forestieres, 51, 477–492.
Beier C, Beierkuhnlein C, Wohlgemuth T et al. (2012) Precipitation manipulation
References experiments – challenges and recommendations for the future. Ecology Letters, 15,
Ainsworth EA, Long SP (2005) What have we learned from 15 years of free-air 899–911.
CO2 enrichment (FACE)? A meta-analytic review of the responses of photo- Bergh J, Linder S, Lundmark T, Elfving B (1999) The effect of water and nutrient
synthesis, canopy properties and plant production to rising CO2. New Phytolo- availability on the productivity of Norway spruce in northern and southern Swe-
gist, 165, 351. den. Forest Ecology and Management, 119, 51–62.
Alberti G, Inglima I, Arriga N et al. (2007) Changes in rainfall patterns in Mediterra- Boisvenue C, Running SW (2006) Impacts of climate change on natural forest produc-
nean ecosystems: the MIND project. Forest@ - Rivista di Selvicoltura ed Ecologia Fore- tivity – evidence since the middle of the 20th century. Global Change Biology, 12,
stale,4, 460–468. 862–882.
Allen CD, Macalady AK, Chenchouni H et al. (2010) A global overview of drought Bolker BM, Pacala SW, Bazzaz FA, Canham CD, Levin SA (1995) Species diversity
and heat-induced tree mortality reveals emerging climate change risks for forests. and ecosystem response to carbon dioxide fertilization: conclusions from a tem-
Forest Ecology and Management, 259, 660–684. perate forest model. Global Change Biology, 1, 373–381.
Amiro BD, Orchansky AL, Barr AG et al. (2006) The effect of post-fire stand age on Bormann FJ, Likens GE (1994) Pattern and Process in a Forested Ecosystem: disturbance,
the boreal forest energy balance. Agricultural and Forest Meteorology, 140, 41–50. Development, and the Steady State Based on the Hubbard Brook Ecosystem Study.
Amiro BD, Barr AG, Barr JG et al. (2010) Ecosystem carbon dioxide fluxes after distur- Springer-Verlag, Berlin.
bance in forests of North America. Journal of Geophysical Research, 115, G00K02. Breshears DD, Cobb NS, Rich PM et al. (2005) Regional vegetation die-off in response
Anderson KJ (2007a) Rates of change in ecosystem and community properties during succes- to global-change-type drought. Proceedings of the National Academy of Sciences of the
sion. PhD Dissertation; University of New Mexico, Albuquerque, NM. United States of America, 102, 15144–15148.

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
18 K . J . A N D E R S O N - T E I X E I R A et al.

Bronson DR, Gower ST (2010) Ecosystem warming does not affect photosynthesis or DeLucia EH, Hamilton JG, Naidu SL et al. (1999) Net primary production of a forest
aboveground autotrophic respiration for boreal black spruce. Tree Physiology, 30, ecosystem with experimental CO2 enrichment. Science, 284, 1177–1179.
441–449. Delzon S, Loustau D (2005) Age-related decline in stand water use: sap flow and tran-
Bronson DR, Gower ST, Tanner M, Linder S, Van Herk I (2008) Response of soil sur- spiration in a pine forest chronosequence. Agricultural and Forest Meteorology, 129,
face CO2 flux in a boreal forest to ecosystem warming. Global Change Biology, 14, 105–119.
856–867. Didham RK, Watts CH, Norton DA (2005) Are systems with strong underlying abi-
Bronson DR, Gower ST, Tanner M, Van Herk I (2009) Effect of ecosystem warming on otic regimes more likely to exhibit alternative stable states? Oikos, 110, 409–416.
boreal black spruce bud burst and shoot growth. Global Change Biology, 15, Dieleman WIJ, Vicca S, Dijkstra FA et al. (2012) Simple additive effects are rare: a
1534–1543. quantitative review of plant biomass and soil process responses to combined
Brown S, Lugo A (1982) The storage and production of organic matter in tropical for- manipulations of CO2 and temperature. Global Change Biology, 18, 2681–2693.
ests and their role in the global carbon cycle. Biotropica, 14, 161–187. Dijkstra P, Hymus G, Colavito D et al. (2002) Elevated atmospheric CO2 stimulates
Brown JH, Brown TE, Lomolino M (1998) Biogeography. 2 Sub. Sinauer Associates, aboveground biomass in a fire-regenerated scrub-oak ecosystem. Global Change
Sunderland, MA. Biology, 8, 90–103.
Bunker DE, DeClerck F, Bradford JC, Colwell RK, Perfecto I, Phillips OL et al. (2005) Dillenburg LR, Teramura AH, Forseth IN, Whigham DF (1995) Photosynthetic and
Species loss and aboveground carbon storage in a tropical forest. Science, 310, biomass allocation responses of Liquidambar styraciflua (Hamamelidaceae) to vine
1029–1031. competition. American journal of botany, 82, 454–461.
Butler S, Melillo J, Johnson J et al. (2012) Soil warming alters nitrogen cycling in a Donnegan JA, Rebertus AJ (1999) Rates and mechanisms of subalpine forest succes-
New England forest: implications for ecosystem function and structure. Oecologia, sion along an environmental gradient. Ecology, 80, 1370–1384.
168, 819–828. Dore S, Kolb TE, Montes-Helu M et al. (2008) Long-term impact of a stand-replacing
Campbell J, Donato D, Azuma D, Law B (2007) Pyrogenic carbon emission from a fire on ecosystem CO2 exchange of a ponderosa pine forest. Global Change Biology,
large wildfire in Oregon United States. Journal of Geophysical Research, 112, G04014. 14, 1801–1820.
Carney KM, Hungate BA, Drake BG, Megonigal JP (2007) Altered soil microbial com- Drake JE, Raetz LM, Davis SC, DeLucia EH (2010) Hydraulic limitation not declining
munity at elevated CO2 leads to loss of soil carbon. PNAS, 104, 4990–4995. nitrogen availability causes the age-related photosynthetic decline in loblolly pine
Ceulemans R, Mousseau M (1994) Tansley Review No. 71 Effects of elevated atmo- (Pinus taeda L.). Plant, Cell and Environment, 33, 1756–1766.
spheric CO2 on woody plants. New Phytologist, 127, 425–446. Drake JE, Davis SC, Raetz LM, DeLucia EH (2011a) Mechanisms of age-related
Chave J, Condit R, Muller-Landau HC et al. (2008) Assessing evidence for a pervasive changes in forest production: the influence of physiological and successional
alteration in tropical tree communities. PLoS Biology, 6, e45. changes. Global Change Biology, 17, 1522–1535.
Cheesman AW, Winter K (2013) Elevated night-time temperatures increase growth in Drake JE, Gallet-Budynek A, Hofmockel KS et al. (2011b) Increases in the flux of car-
seedlings of two tropical pioneer tree species. New Phytologist, 197, 1185–1192. bon belowground stimulate nitrogen uptake and sustain the long-term enhance-
Clark DA (2002) Are tropical forests an important carbon sink? Reanalysis of the ment of forest productivity under elevated CO2. Ecology Letters, 14, 349.
long-term plot data. Ecological Applications, 12, 3–7. Duval BD (2010) The Impact of Elevated CO2 on N2 Fixation and Ecosystem Level Element
Clark DB, Clark DA, Oberbauer SF (2010) Annual wood production in a tropical rain Cycling. PhD; Northern Arizona University, Flagstaff, AZ.
forest in NE Costa Rica linked to climatic variation but not to increasing CO2. Glo- Edburg SL, Hicke JA, Lawrence DM, Thornton PE (2011) Simulating coupled carbon
bal Change Biology, 16, 747–759. and nitrogen dynamics following mountain pine beetle outbreaks in the western
Claus A, George E (2005) Effect of stand age on fine-root biomass and biomass distri- United States. Journal of Geophysical Research, 116, G04033.
bution in three European forest chronosequences. Canadian Journal of Forest Enquist BJ, West GB, Brown JH (2009) Extensions and evaluations of a general quanti-
Research, 35, 1617–1625. tative theory of forest structure and dynamics. Proceedings of the National Academy
Cleveland CC, Wieder WR, Reed SC, Townsend AR (2010) Experimental drought in a of Sciences, 106, 7046–7051.
tropical rain forest increases soil carbon dioxide losses to the atmosphere. Ecology, Evangelista PH, Kumar S, Stohlgren TJ, Young NE (2011) Assessing forest vulnerability
91, 2313–2323. and the potential distribution of pine beetles under current and future climate sce-
Cole CT, Anderson JE, Lindroth RL, Waller DM (2010) Rising concentrations of atmo- narios in the Interior West of the US. Forest Ecology and Management, 262, 307–316.
spheric CO2 have increased growth in natural stands of quaking aspen (Populus Feeley KJ, Davies SJ, Ashton PS, Bunyavejchewin S, Supardi MNN, Kassim AR et al.
tremuloides). Global Change Biology, 16, 2186–2197. (2007a) The role of gap phase processes in the biomass dynamics of tropical for-
Comstedt D, Bostr€ om B, Marshall J, Holm A, Slaney M, Linder S, Ekblad A (2006) ests. Proceedings of the Royal Society B, 274, 2857–2864.
Effects of elevated atmospheric carbon dioxide and temperature on soil respiration Feeley KJ, Joseph Wright S, Nur Supardi MN, Kassim AR, Davies SJ (2007b) Deceler-
in a boreal forest using 13C as a labeling tool. Ecosystems, 9, 1266–1277. ating growth in tropical forest trees. Ecology Letters, 10, 461–469.
Conant RT, Klopatek JM, Klopatek CC (2000) Environmental factors controlling soil res- Feeley KJ, Davies SJ, Perez R, Hubbell SP, Foster RB (2011) Directional changes in the
piration in three semiarid ecosystems. Soil Science Society of America Journal, 64, 383. species composition of a tropical forest. Ecology, 92, 871–882.
Curtis PS, Wang X (1998) A meta-analysis of elevated CO2; effects on woody plant Ferrell WK, Woodard ES (1966) Effects of Seed Origin on Drought Resistance of
mass, form, and physiology. Oecologia, 113, 299–313. Douglas-Fir (Pseudotsuga Menziesii) (Mirb.) Franco. Ecology, 47, 499–503.
Dale VH, Joyce LA, McNulty S et al. (2001) Climate change and forest disturbances. Ffolliott PF, Stropki CL, Chen H, Neary DG (2011) The 2002 Rodeo-Chediski Wildfire’s
BioScience, 51, 723–734. Impacts on Southwestern Ponderosa Pine Ecosystems, Hydrology, and Fuels. U.S.
D’Antonio CM, Vitousek PM (1992) Biological invasions by exotic grasses, the Department of Agriculture, Forest Service, Rocky Mountain Research Station, Fort
grass/fire cycle, and global change. Annual Review of Ecology and Systematics, Collins, CO.
23, 63–87. Finegan B (1984) Forest succession. Nature, UK, 312, 109–114.
Davi H, Barbaroux C, Francois C, Dufr^ene E (2009) The fundamental role of reserves Finegan B (1996) Pattern and process in neotropical secondary rain forests: the first
and hydraulic constraints in predicting LAI and carbon allocation in forests. Agri- 100 years of succession. Trends in Ecology and Evolution, 11, 119–124.
cultural and Forest Meteorology, 149, 349–361. Foster JR, Burton JI, Forrester JA et al. (2010) Evidence for a recent increase in forest
Davidson EA, de Carvalho CJR, Figueira AM et al. (2007) Recuperation of nitrogen growth is questionable. PNAS, 107, E86–E87.
cycling in Amazonian forests following agricultural abandonment. Nature, 447, Fowells HA, Stark NB (1965) . Natural regeneration in relation to environment in the
995–998. mixed conifer forest type of California. Pacific Southwest Forest and Range Experi-
Dawes MA, H€attenschwiler S, Bebi P, Hagedorn F, Handa IT, K€ orner C, Rixen C ment Station, Forest Service, US Department of Agriculture, 1965.
(2011) Species-specific tree growth responses to 9 years of CO2 enrichment at the Fukami T, Nakajima M (2011) Community assembly: alternative stable states or alter-
alpine treeline. Journal of Ecology, 99, 383–394. native transient states? Ecology Letters, 14, 973–984.
De Graaff M-A, Van Groenigen K-J, Six J, Hungate B, Van Kessel C (2006) Interactions Gilliam FS (2007) The ecological significance of the herbaceous layer in temperate for-
between plant growth and soil nutrient cycling under elevated CO2: a meta-analy- est ecosystems. BioScience, 57, 845–858.
sis. Global Change Biology, 12, 2077. Goulden ML, McMillan AMS, Winston GC, Rocha AV, Manies KL, Harden JW,
de Visser PHB, Beier C, Rasmussen L, Kreutzer K, Steinberg N, Bredemeier M et al. Bond-Lamberty BP (2011) Patterns of NPP, GPP, respiration, and NEP during bor-
(1994) Biological response of five forest ecosystems in the EXMAN project to input eal forest succession. Global Change Biology, 17, 855–871.
changes of water, nutrients and atmospheric loads. Forest Ecology and Management, Gower ST, McMurtrie RE, Murty D (1996) Aboveground net primary production
68, 15–29. decline with stand age: potential causes. Trends in Ecology and Evolution, 11, 378.

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
 F O R E S T R E C O V E R Y U N D E R C L I M A T E C H A N G E 19

Graumlich LJ, Brubaker LB, Grier CC (1989) Long-term trends in forest net primary Krinner G, Viovy N, Noblet-Ducoudr
e N et al. (2005) A dynamic global vegetation
productivity: cascade mountains, Washington. Ecology, 70, 405–410. model for studies of the coupled atmosphere-biosphere system. Global Biogeochemi-
Grimm V, Wissel C (1997) Babel, or the ecological stability discussions: an inventory cal Cycles, 19, GB1015.
and analysis of terminology and a guide for avoiding confusion. Oecologia, 109, Kubiske ME, Pregitzer KS (1996) Effects of elevated CO2 and light availability on the
323–334. photosynthetic light response of trees of contrasting shade tolerance. Tree Physiol-
Hall AR, Miller AD, Leggett HC, Roxburgh SH, Buckling A, Shea K (2012) Diversity–dis- ogy, 16, 351–358.
turbance relationships: frequency and intensity interact. Biology Letters, 8, 768–771. Kurz WA, Dymond CC, Stinson G et al. (2008) Mountain pine beetle and forest car-
Hanewinkel M, Cullmann DA, Schelhaas M-J, Nabuurs G-J, Zimmermann NE (2013) bon feedback to climate change. Nature, 452, 987.
Climate change may cause severe loss in the economic value of European forest Lamarche VC, Graybill DA, Fritts HC, Rose MR (1984) Increasing atmospheric carbon
land. Nature Climate Change, 3, 203–207. dioxide: tree ring evidence for growth enhancement in natural vegetation. Science,
Hanson PJ, Todd DE, Amthor JS (2001) A six-year study of sapling and large-tree 225, 1019–1021.
growth and mortality responses to natural and induced variability in precipitation Langley JA, McKinley DC, Wolf AA, Hungate BA, Drake BG, Megonigal JP (2009)
and throughfall. Tree Physiology, 21, 345–358. Priming depletes soil carbon and releases nitrogen in a scrub-oak ecosystem
orner CH (2000) Tree seedling responses to in situ CO2-enrich-
H€attenschwiler ST, K€ exposed to elevated CO2. Soil Biology and Biochemistry, 41, 54–60.
ment differ among species and depend on understorey light availability. Global Larjavaara M, Muller-Landau HC (2012) Temperature explains global variation in
Change Biology, 6, 213–226. biomass among humid old-growth forests. Global Ecology and Biogeography, 21,
Hember RA, Kurz WA, Metsaranta JM, Black TA, Guy RD, Coops NC (2012) Acceler- 998–1006.
ating regrowth of temperate-maritime forests due to environmental change. Global Law BE, Sun OJ, Campbell J, Tuyl SV, Thornton PE (2003) Changes in carbon storage
Change Biology, 18, 2026–2040. and fluxes in a chronosequence of ponderosa pine. Global Change Biology, 9, 510–524.
Hicke JA, Allen CD, Desai AR et al. (2011) Effects of biotic disturbances on forest car- LeDuc SD, Rothstein DE (2010) Plant-available organic and mineral nitrogen shift in
bon cycling in the United States and Canada. Global Change Biology, 18, 7–34. dominance with forest stand age. Ecology, 91, 708–720.
Hooper DU, Adair EC, Cardinale BJ et al. (2012) A global synthesis reveals biodiver- Lewis SL, Lopez-Gonzalez G, Sonke B et al. (2009) Increasing carbon storage in intact
sity loss as a major driver of ecosystem change. Nature, 486, 105–108. African tropical forests. Nature, 457, 1003.
Hoosbeek MR, Lukac M, Velthorst E, Smith AR, Godbold DL (2011) Free atmospheric Li D, Niu S, Luo Y (2012) Global Patterns of the Dynamics of Soil Carbon and Nitrogen
CO2 enrichment increased above ground biomass but did not affect symbiotic Stocks Following Afforestation: a Meta-Analysis. New Phytologist.
N2-fixation and soil carbon dynamics in a mixed deciduous stand in Wales. Bio- Lichstein JW, Wirth C, Horn HS, Pacala SW (2009) Biomass chronosequences of Uni-
geosciences, 8, 353–364. ted States forests: implications for carbon storage and forest management. In: Old-
Hudiburg TW, Law BE, Thornton PE (2013) Evaluation and improvement of the Growth Forests, Ecological Studies (eds Wirth C, Gleixner G, Heimann M), pp.
Community Land Model (CLM 4.0) in Oregon forests. Biogeosciences, 10, 453–470. 301–341. Springer, Berlin, Heidelberg.
Hungate BA, Dijkstra P, Johnson DW, Hinkle CR, Drake BG (1999) Elevated CO2 Liu HP, Randerson JT, Lindfors J, Chapin FS (2005) Changes in the surface energy
increases nitrogen fixation and decreases soil nitrogen mineralization in Florida budget after fire in boreal ecosystems of interior Alaska: an annual perspective.
scrub oak. Global Change Biology, 5, 781–789. Journal of Geophysical Research-Atmospheres, 110, D13101.
Hungate BA, Stiling PD, Dijkstra P et al. (2004) CO2 elicits long-term decline in nitro- Lu M, Zhou X, Yang Q et al. (2012) Responses of ecosystem carbon cycle to experi-
gen fixation. Science, 304, 1291–1291. mental warming: a meta-analysis. Ecology, in press.
Hurtt G, Chini L, Frolking S et al. (2011) Harmonization of land-use scenarios for the Luo Y, Su B, Currie WS et al. (2004) Progressive nitrogen limitation of ecosystem
period 1500–2100: 600 years of global gridded annual land-use transitions, wood responses to rising atmospheric carbon dioxide. BioScience, 54, 731–739.
harvest, and resulting secondary lands. Climatic Change, 109, 117–161. Luyssaert S, Inglima I, Jung M et al. (2007) CO2 balance of boreal, temperate, and
Hyv€ 
onen R, Agren GI, Linder S et al. (2007) The likely impact of elevated [CO2], nitro- tropical forests derived from a global database. Global Change Biology, 13,
gen deposition, increased temperature and management on carbon sequestration 2509–2537.
in temperate and boreal forest ecosystems: a literature review. New Phytologist, Luyssaert S, Schulze ED, Borner A et al. (2008) Old-growth forests as global carbon
173, 463–480. sinks. Nature, 455, 213.
Ingwell LL, Joseph Wright S, Becklund KK, Hubbell SP, Schnitzer SA (2010) The Mack MC, D’Antonio CM (1998) Impacts of biological invasions on disturbance
impact of lianas on 10 years of tree growth and mortality on Barro Colorado regimes. Trends in Ecology and Evolution, 13, 195–198.
Island, Panama. Journal of Ecology, 98, 879–887. Maness H, Kushner PJ, Fung I (2012) Summertime climate response to mountain pine
IPCC (2007) Climate Change 2007: the Physical Science Basis. Contribution of working beetle disturbance in British Columbia. Nature Geoscience, 6, 65–70.
group I to the fourth assessment report of the intergovernmental panel on climate Mar
ın-Spiotta E, Cusack D, Ostertag R, Silver W(2008) Trends in above and below-
change. Cambridge University Press, Cambridge. ground carbon with forest regrowth after agricultural abandonment in the neo-
Iversen CM (2010) Digging deeper: fine-root responses to rising atmospheric CO2 tropics. In: Post-Agricultural Succession in the Neotropics (ed. Myster RW), pp 22–72.
concentration in forested ecosystems. New Phytologist, 186, 346–357. Springer, NewYork, NY.
Jin Y, Randerson JT, Goulden ML, Goetz SJ (2012) Post-fire changes in net shortwave McCarthy HR, Oren R, Johnsen KH et al. (2010) Re-assessment of plant carbon
radiation along a latitudinal gradient in boreal North America. Geophysical Research dynamics at the Duke free-air CO2 enrichment site: interactions of atmospheric
Letters, 39, L13403. [CO2] with nitrogen and water availability over stand development. New Phytolo-
Jobb
agy EG, Jackson RB (2000) The vertical distribution of soil organic carbon and its gist, 185, 514–528.
relation to climate and vegetation. Ecological Applications, 10, 423–436. McKinley DC, Romero JC, Hungate BA, Drake BG, Megonigal JP (2009) Does deep
Johnson C, Zarin D, Johnson A (2000) Post-disturbance aboveground biomass accu- soil N availability sustain long-term ecosystem responses to elevated CO2? Global
mulation in global secondary forests. Ecology, 81, 1395–1401. Change Biology, 15, 2035–2048.
Keith H, Mackey BG, Lindenmayer DB (2009) Re-evaluation of forest biomass carbon McMahon SM, Parker GG, Miller DR (2010a) Evidence for a recent increase in forest
stocks and lessons from the world’s most carbon-dense forests. Proceedings of the growth. Proceedings of the National Academy of Sciences, 107, 3611–3615.
National Academy of Sciences, 106, 11635–11640. McMahon SM, Parker GG, Miller DR (2010b) Reply to Foster et al.: using a forest to
Kerstiens G (1998) Shade-tolerance as a predictor of responses to elevated CO2 in measure trees: determining which vital rates are responding to climate change.
trees. Physiologia Plantarum, 102, 472–480. PNAS, 107, E88–E89.
Kerstiens G (2001) Meta-analysis of the interaction between shade-tolerance, light McMillan AMS, Winston GC, Goulden ML (2008) Age-dependent response of boreal
environment and growth response of woody species to elevated CO2. Acta Oeco- forest to temperature and rainfall variability. Global Change Biology, 14, 1904–1916.
logica, 22, 61–69. Medvigy D, Wofsy SC, Munger JW, Hollinger DY, Moorcroft PR (2009) Mechanistic
Kirilenko AP, Sedjo RA (2007) Climate change impacts on forestry. PNAS, 104, scaling of ecosystem function and dynamics in space and time: ecosystem demog-
19697–19702. raphy model version 2. Journal of Geophysical Research, 114, G01002.
Kirschbaum MUF, Whitehead D, Dean SM, Beets PN, Shepherd JD, Ausseil A-GE Melillo JM, Steudler PA, Aber JD et al. (2002) Soil warming and carbon-cycle feed-
(2011) Implications of albedo changes following afforestation on the benefits of for- backs to the climate system. Science, 298, 2173–2176.
ests as carbon sinks. Biogeosciences, 8, 3687–3696. Melillo JM, Butler S, Johnson J et al. (2011) Soil warming, carbon-nitrogen interac-
orner C, Asshoff R, Bignucolo O et al. (2005) Carbon flux and growth in mature
K€ tions, and forest carbon budgets. Proceedings of the National Academy of Sciences,
deciduous forest trees exposed to elevated CO2. Science, 309, 1360–1362. 108, 9508–9512.

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
20 K . J . A N D E R S O N - T E I X E I R A et al.

Metsaranta JM, Dymond CC, Kurz WA, Spittlehouse DL (2011) Uncertainty of 21st Rollinson CR, Kaye MW (2011) Experimental warming alters spring phenology of
century growing stocks and GHG balance of forests in British Columbia, Canada certain plant functional groups in an early-successional forest community. Global
resulting from potential climate change impacts on ecosystem processes. Forest Change Biology, 18, 1108–1116.
Ecology and Management, 262, 827–837. Romme WH, Knight DH (1981) Fire frequency and subalpine forest succession along
Miller AD, Roxburgh SH, Shea K (2011) How frequency and intensity shape diver- a topographic gradient in wyoming. Ecology, 62, 319–326.
sity–disturbance relationships. PNAS, 108, 5643–5648. Running SW (2008) Ecosystem disturbance, carbon, and climate. Science, 321, 652–653.
Mohan JE, Ziska LH, Schlesinger WH, Thomas RB, Sicher RC, George K, Clark JS Russell AE, Raich JW (2012) Rapidly growing tropical trees mobilize remarkable
(2006) Biomass and toxicity responses of poison ivy (Toxicodendron radicans) to amounts of nitrogen, in ways that differ surprisingly among species. PNAS, 109,
elevated atmospheric CO2. PNAS, 103, 9086–9089. 10398–10402.
Mohan JE, Clark JS, Schlesinger WH (2007) Long-term CO2 enrichment of a forest Rustad L, Campbell J, Marion G et al. (2001) A meta-analysis of the response of soil
ecosystem: implications for forest regeneration and succession. Ecological Applica- respiration, net nitrogen mineralization, and aboveground plant growth to experi-
tions, 17, 1198–1212. mental ecosystem warming. Oecologia, 126, 543–562.
Moles AT, Flores-Moreno H, Bonser SP et al. (2012) Invasions: the trail behind, the Salzer MW, Hughes MK, Bunn AG, Kipfmueller KF (2009) Recent unprecedented
path ahead, and a test of a disturbing idea. Journal of Ecology, 100, 116–127. tree-ring growth in bristlecone pine at the highest elevations and possible causes.
Nepstad DC, Moutinho P, Dias-Filho MB et al. (2002) The effects of partial throughfall PNAS, 106, 20348–20353.
exclusion on canopy processes, aboveground production, and biogeochemistry of Sasek TW, Strain BR (1990) Implications of atmospheric CO2 enrichment and climatic
an Amazon forest. Journal of Geophysical Research, 107, 8085. change for the geographical distribution of two introduced vines in the U.S.A. Cli-
Nepstad DC, Tohver IM, Ray D, Moutinho P, Cardinot G (2007) Mortality of large matic Change, 16, 31–51.
trees and lianas following experimental drought in an amazon forest. Ecology, 88, Savage M, Mast JN (2005) How resilient are southwestern ponderosa pine forests
2259–2269. after crown fires? Canadian Journal of Forest Research, 35, 967–977.
Nilsson M-C, Wardle DA (2005) Understory vegetation as a forest ecosystem driver: Saxe H, Ellsworth DS, Heath J (1998) Tree and forest functioning in an enriched CO2
evidence from the northern Swedish boreal forest. Frontiers in Ecology and the Envi- atmosphere. New Phytologist, 139, 395–436.
ronment, 3, 421–428. Scheffer M, Carpenter S, Foley JA, Folke C, Walker B (2001) Catastrophic shifts in eco-
Norby RJ, Zak DR (2011) Ecological lessons from free-air CO2 enrichment (face) systems. Nature, 413, 591–596.
experiments. Annual Review of Ecology, Evolution, and Systematics, 42, 181–203. Scheffer M, Hirota M, Holmgren M, Nes EHV, Chapin FS (2012) Thresholds for bor-
Norby RJ, DeLucia EH, Gielen B et al. (2005) Forest response to elevated CO2 is con- eal biome transitions. PNAS, 109, 21384–21389.
served across a broad range of productivity. Proceedings of the National Academy of Schnitzer SA, Bongers F (2011) Increasing liana abundance and biomass in tropical
Sciences of the United States of America, 102, 18052–18056. forests: emerging patterns and putative mechanisms. Ecology Letters, 14, 397–406.
Norby RJ, Warren JM, Iversen CM, Medlyn BE, McMurtrie RE (2010) CO2 enhance- Schnitzer SA, Carson WP (2010) Lianas suppress tree regeneration and diversity in
ment of forest productivity constrained by limited nitrogen availability. Proceed- treefall gaps. Ecology Letters, 13, 849–857.
ings of the National Academy of Sciences, 107, 19368–19373. Schr€oder A, Persson L, De Roos AM (2005) Direct experimental evidence for alterna-
Odion DC, Moritz MA, DellaSala DA (2010) Alternative community states main- tive stable states: a review. Oikos, 110, 3–19.
tained by fire in the Klamath Mountains, USA. Journal of Ecology, 98, 96–105. Schuldt B, Leuschner C, Horna V, Moser G, K€ ohler M, van Straaten O, Barus H (2011)
O’Halloran TL, Law BE, Goulden ML et al. (2012) Radiative forcing of natural forest Change in hydraulic properties and leaf traits in a tall rainforest tree species sub-
disturbances. Global Change Biology, 18, 555–565. jected to long-term throughfall exclusion in the perhumid tropics. Biogeosciences, 8,
Oren R, Ellsworth DS, Johnsen KH et al. (2001) Soil fertility limits carbon sequestra- 2179–2194.
tion by forest ecosystems in a CO2-enriched atmosphere. Nature, 411, 469–472. Seiler TJ, Rasse DP, Li J et al. (2009) Disturbance, rainfall and contrasting species
Pan Y, Birdsey RA, Fang J et al. (2011) A large and persistent carbon sink in the responses mediated aboveground biomass response to 11 years of CO2 enrich-
world’s forests. Science, 333, 988–993. ment in a Florida scrub-oak ecosystem. Global Change Biology, 15, 356–367.
Pendall E, Bridgham S, Hanson PJ et al. (2004) Below-ground process responses to Shafi MI, Yarranton GA (1973) Diversity, floristic richness, and species evenness dur-
elevated CO2 and temperature: a discussion of observations, measurement meth- ing a secondary (post-fire) succession. Ecology, 54, 897–902.
ods, and models. New Phytologist, 162, 311–322. Shea K, Chesson P (2002) Community ecology theory as a framework for biological
nuelas J, Filella I, Tognetti R (2001) Leaf mineral concentrations of Erica arborea,
Pe~ invasions. Trends in Ecology and Evolution, 17, 170–176.
Juniperus communis and Myrtus communis growing in the proximity of a natural Slaney M, Wallin G, Medhurst J, Linder S (2007) Impact of elevated carbon dioxide
CO2 spring. Global Change Biology, 7, 291–301. concentration and temperature on bud burst and shoot growth of boreal norway
Phillips OL, V
asquez Mart
ınez R, Arroyo L et al. (2002) Increasing dominance of large spruce. Tree Physiology, 27, 301–312.
lianas in Amazonian forests. Nature, 418, 770–774. Smithwick EAH (2011) Pyrogeography and biogeochemical resilience. In: The Land-
Plaut JA, Yepez EA, Hill J, Pangle R, Sperry JS, Pockman WT, McDowell N(2012) scape Ecology of Fire, Ecological Studies (eds McKenzie D, Miller C, Falk DA, Cald-
Hydraulic limits preceding mortality in a pi~ non–juniper woodland under experi- well MM, Heldmaier G, Jackson RB et al.), pp. 143–163. Springer, Netherlands.
mental drought. Plant, Cell and Environment, 35, 1601–1617. Smithwick EAH, Ryan MG, Kashian DM, Romme WH, Tinker DB, Turner MG (2009)
Prach K, Rehounkov
a K (2006) Vegetation succession over broad geographical scales: Modeling the effects of fire and climate change on carbon and nitrogen storage in
which factors determine the patterns? Preslia, 78, 469–480. lodgepole pine (Pinus contorta) stands. Global Change Biology, 15, 535–548.
Prach K, Pysek P, Jarosik V (2007) Climate and pH as determinants of vegetation suc- Sotta ED, Veldkamp E, Schwendenmann L et al. (2007) Effects of an induced drought
cession in Central European man-made habitats. Journal of Vegetation Science, 18, on soil carbon dioxide (CO2) efflux and soil CO2 production in an Eastern Amazo-
701–710. nian rainforest, Brazil. Global Change Biology, 13, 2218–2229.
Pregitzer KS, Euskirchen ES (2004) Carbon cycling and storage in world forests: Soul
e PT, Knapp PA (2006) Radial growth rate increases in naturally occurring pon-
biome patterns related to forest age. Global Change Biology, 10, 2052–2077. derosa pine trees: a late-20th century CO2 fertilization effect? New Phytologist, 171,
Raich JW, Schlesinger WH (1992) The global carbon dioxide flux in soil respiration 379–390.
and its relationship to vegetation and climate. Tellus B, 44, 81–99. Souza L, Belote RT, Kardol P, Weltzin JF, Norby RJ (2010) CO2 enrichment accelerates
Randerson JT, Liu H, Flanner MG et al. (2006) The impact of boreal forest fire on cli- successional development of an understory plant community. Journal of Plant Ecol-
mate warming. Science, 314, 1130–1132. ogy, 3, 33–39.
Richter S, Kipfer T, Wohlgemuth T, Calder
on Guerrero C, Ghazoul J, Moser B (2012) Spiecker H (1999) Overview of recent growth trends in European forests. Water, Air,
Phenotypic plasticity facilitates resistance to climate change in a highly variable and Soil Pollution, 116, 33–46.
environment. Oecologia, 169, 269–279. Staver AC, Archibald S, Levin S (2011) Tree cover in sub-Saharan Africa: rainfall and
Roberts S, Vertessy R, Grayson R (2001) Transpiration from Eucalyptus sieberi (L. John- fire constrain forest and savanna as alternative stable states. Ecology, 92,
son) forests of different age. Forest Ecology and Management, 143, 153–161. 1063–1072.
Roccaforte JP, Ful
e PZ, Chancellor WW, Laughlin DC (2012) Woody debris and tree Stromayer KAK, Warren RJ (1997) Are overabundant deer herds in the eastern united
regeneration dynamics following severe wildfires in Arizona ponderosa pine for- states creating alternate stable states in forest plant communities? Wildlife Society
ests. Canadian Journal of Forest Research, 42, 593–604. Bulletin, 25, 227–234.
Rollinson CR (2010) Simulated Climate Change Alters Post-Clear Cut Forest Vegetation Sturrock RN, Frankel SJ, Brown AV et al. (2011) Climate change and forest diseases.
Communities. M.S.; Penn State University, State College, PA. Plant Pathology, 60, 133–149.

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194
 F O R E S T R E C O V E R Y U N D E R C L I M A T E C H A N G E 21

Thomas PA, Wein RW (1985) Water availability and the comparative emergence of Wright IJ, Reich PB, Westoby M et al. (2004) The worldwide leaf economics spectrum.
four conifer species. Canadian Journal of Botany, 63, 1740–1746. Nature, 428, 821–827.
Thompson J, Spies T (2010) Factors associated with crown damage following recur- Wu Z, Dijkstra P, Koch GW, Pe~ nuelas J, Hungate BA (2011) Responses of terrestrial
ring mixed-severity wildfires and post-fire management in southwestern Oregon. ecosystems to temperature and precipitation change: a meta-analysis of experi-
Landscape Ecology, 25, 775–789. mental manipulation. Global Change Biology, 17, 927–942.
Thompson I, Mackey B, McNulty S, Mosseler A(2009) Forest Resilience, Biodiversity, Yang Z, Cohen WB, Harmon ME (2005) Modeling early forest succession following
and Climate Change. Secretariat of the Convention on Biological Diversity, Mon- clear-cutting in western Oregon. Canadian Journal of Forest Research, 35, 1889–1900.
treal. Technical Series. Vol. 43. 2009 Yang Y, Luo Y, Finzi AC (2011) Carbon and nitrogen dynamics during forest stand
Thornton PE, Lamarque J-F, Rosenbloom NA, Mahowald NM (2007) Influence of car- development: a global synthesis. New Phytologist, 190, 977–989.
bon-nitrogen cycle coupling on land model response to CO2 fertilization and Yavitt JB, Wright SJ (2008) Seedling growth responses to water and nutrient augmen-
climate variability. Global Biogeochemical Cycles, 21, GB4018. tation in the understorey of a lowland moist forest, Panama. Journal of Tropical
Tingey DT, Lee EH, Phillips DL, Rygiewicz PT, Waschmann RS, Johnson MG, Olszyk Ecology, 24, 19–26.
DM (2007) Elevated CO2 and temperature alter net ecosystem C exchange in a Yuan ZY, Chen HYH (2010) Fine root biomass, production, turnover rates, and nutri-
young Douglas fir mesocosm experiment. Plant, Cell and Environment, 30, ent contents in boreal forest ecosystems in relation to species, climate, fertility, and
1400–1410. stand age: literature review and meta-analyses. Critical Reviews in Plant Sciences,
Uhl C, Jordan CF (1984) Succession and nutrient dynamics following forest cutting 29, 204–221.
and burning in Amazonia. Ecology, 65, 1476–1490. Zaehle S, Friend AD (2010) Carbon and nitrogen cycle dynamics in the O-CN land
US DOE (2011). US Billion-Ton Update. Biomass Supply for a Bioenergy and Bioproducts surface model: 1 Model description, site-scale evaluation, and sensitivity to param-
Industry. Oak Ridge National Laboratory, Oak Ridge, TN. eter estimates. Global Biogeochemical Cycles, 24, GB1005.
Vasconcelos SS, Zarin DJ, Ara
ujo MM, de Miranda I, S. (2012) Aboveground net pri- Zanetti S, Hartwig UA, Luscher A et al. (1996) Stimulation of Symbiotic N2 Fixation
mary productivity in tropical forest regrowth increases following wetter dry- in Trifolium repens L. under Elevated Atmospheric pCO2 in a Grassland Ecosystem.
seasons. Forest Ecology and Management, 276, 82–87. Plant Physiology, 112, 575–583.
Vitousek P, Matson P, Cleve K (1989) Nitrogen availability and nitrification during Zhou G, Liu S, Li Z et al. (2006) Old-growth forests can accumulate carbon in soils.
succession: primary, secondary, and old-field seres. Plant and Soil, 115, 229–239. Science, 314, 1417.
Voelker SL (2011) Age-dependent changes in environmental influences on tree Zhou Y, Tang J, Melillo JM, Butler S, Mohan JE (2011) Root standing crop and chemis-
growth and their implications for forest responses to climate change. In: Size- and try after 6 years of soil warming in a temperate forest. Tree Physiology, 31, 707–717.
Age-Related Changes in Tree Structure and Function, Tree Physiology (eds Meinzer
FCC, Lachenbruch B, Dawson TEE, Meinzer FC, Niinemets U), € pp. 455–479.
Springer, Netherlands.
Vogt KA, Moore EE, Vogt DJ, Redlin MJ, Edmonds RL (1983) Conifer fine root and
mycorrhizal root biomass within the forest floors of Douglas-fir stands of different
Supporting Information
ages and site productivities. Canadian Journal of Forest Research, 13, 429–437.
Volder A, Briske DD, Tjoelker MG (2012) Climate warming and precipitation redistri- Additional Supporting Information may be found in the
bution modify tree-grass interactions and tree species establishment in a warm- online version of this article:
temperate savanna. Global Change Biology, 19, 843–857.
Wan S, Norby RJ, Pregitzer KS, Ledford J, O’Neill EG (2004) CO2 enrichment and
Table S1. Summary of experimental manipulations of CO2
warming of the atmosphere enhance both productivity and mortality of maple tree
in tree-dominated ecosystems through Free-Air Carbon
fine roots. New Phytologist, 162, 437–446.
Westerling AL, Hidalgo HG, Cayan DR, Swetnam TW (2006) Warming and earlier
dioxide Enrichment (FACE), Open Top Chamber (OTC), or
spring increase western US forest wildfire activity. Science, 313, 940–943. Whole Tree Chamber (WTC; in situ only) methodology.
Westerling AL, Turner MG, Smithwick EAH, Romme WH, Ryan MG (2011) Continued Table S2. Summary of experimental warming in tree-domi-
Warming Could Transform Greater Yellowstone Fire Regimes by Mid-21st Century. Pro- nated ecosystems (listed in order of forest age).
ceedings of the National Academy of Sciences. Table S3. Summary of experimental manipulations of pre-
Williams AP, Allen CD, Millar CI, Swetnam TW, Michaelsen J, Still CJ, Leavitt SW cipitation (PPT) in tree-dominated ecosystems (listed in
(2010) Forest responses to increasing aridity and warmth in the southwestern Uni- order of forest age).
ted States. PNAS, 107, 21289–21294.
Williams AP, Allen CD, Macalady AK et al. (2013) Temperature as a potent driver of
regional forest drought stress and tree mortality. Nature Climate Change, 3,
292–297.

Published 2013. This article is a U.S. Government work and is in the public domain in the USA., Global Change Biology, doi: 10.1111/
gcb.12194