Tps 057

Tree Physiology 32, 776–798
doi:10.1093/treephys/tps057
Invited review: Part of an invited issue on carbon allocation
Pulse-labelling trees to study carbon allocation dynamics: a review

of methods, current knowledge and future prospects
Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Daniel Epron1,2,12, Michael Bahn3, Delphine Derrien4, Fernando Alfredo Lattanzi5, Jukka Pumpanen6,
Arthur Gessler 7,8, Peter Högberg9, Pascale Maillard1,2, Masako Dannoura 10, Dominique Gérant1,2 and
Nina Buchmann11
1Université de Lorraine, UMR 1137, Ecologie et Ecophysiologie Forestières, Faculté des Sciences, F-54500 Vandoeuvre-les-Nancy, France; 2INRA, UMR 1137, Ecologie et
Ecophysiologie Forestières, Centre de Nancy, F-54280 Champenoux, France; 3University of Innsbruck, Institute of Ecology, Sternwartestr. 15, A-6020 Innsbruck, Austria; 4INRA,
UR 1138, Biogéochimie des Ecosystèmes Forestiers, Forestières, Centre de Nancy, F-54280 Champenoux, France; 5Technische Universität München, Lehrstuhl für
Grünlandlehre, Alte Akademie 12, D-85350, Freising-Weihenstephan, Germany; 6Department of Forest Ecology, University of Helsinki, PO Box 27, 00014 Helsinki, Finland;
7Leibniz Centre for Agricultural Landscape Research (ZALF), Institute for Landscape Biogeochemistry, D-15374 Müncheberg, Germany; 8Faculty of Agriculture and Horticulture,
Humboldt University at Berlin, D-12489 Berlin, Germany; 9Department of Forest Ecology and Management, Swedish University of Agricultural Sciences (SLU), SE-901 83 Umeå,
Sweden; 10Laboratory of Forest Utilization, Department of Forest and Biomaterial Science, Graduate School of Agriculture, Kyoto University, Kyoto 606-8502, Japan; 11ETH
Zurich, Universitaetsstr. 2, LFW C56, 8092 Zurich, Switzerland; 12Corresponding author (daniel.epron@univ-lorraine.fr)
Received February 2, 2012; accepted May 15, 2012; published online June 14, 2012; handling Editor Michael Ryan
Pulse-labelling of trees with stable or radioactive carbon (C) isotopes offers the unique opportunity to trace the fate of
labelled CO2 into the tree and its release to the soil and the atmosphere. Thus, pulse-labelling enables the quantification of
C partitioning in forests and the assessment of the role of partitioning in tree growth, resource acquisition and C sequestra-
tion. However, this is associated with challenges as regards the choice of a tracer, the methods of tracing labelled C in tree
and soil compartments and the quantitative analysis of C dynamics. Based on data from 47 studies, the rate of transfer dif-
fers between broadleaved and coniferous species and decreases as temperature and soil water content decrease. Labelled
C is rapidly transferred belowground—within a few days or less—and this transfer is slowed down by drought. Half-lives of
labelled C in phloem sap (transfer pool) and in mature leaves (source organs) are short, while those of sink organs (growing
tissues, seasonal storage) are longer. 13C measurements in respiratory efflux at high temporal resolution provide the best
estimate of the mean residence times of C in respiratory substrate pools, and the best basis for compartmental modelling.
Seasonal C dynamics and allocation patterns indicate that sink strength variations are important drivers for C fluxes. We pro-
pose a conceptual model for temperate and boreal trees, which considers the use of recently assimilated C versus stored C.
We recommend best practices for designing and analysing pulse-labelling experiments, and identify several topics which we
consider of prime importance for future research on C allocation in trees: (i) whole-tree C source–sink relations, (ii) C alloca-
tion to secondary metabolism, (iii) responses to environmental change, (iv) effects of seasonality versus phenology in and
across biomes, and (v) carbon–nitrogen interactions. Substantial progress is expected from emerging technologies, but the
largest challenge remains to carry out in situ whole-tree labelling experiments on mature trees to improve our understanding
of the environmental and physiological controls on C allocation.
Keywords: carbon isotope, forest, partitioning, residence time, transfer time.
© The Author 2012. Published by Oxford University Press. All rights reserved. For Permissions, please email: journals.permissions@oup.com
Labelling trees to study carbon allocation 777
Introduction and ecosystems (see Dawson et al. 2002), but the low signal
(a few per mil) makes resolving transfer rates and time lags
Carbon (C) allocation is an important determinant of the C difficult. Nevertheless, this approach has been used to inves-
budget of forests and its response to changing environmental tigate C transfer in trees and between forests and the atmo-
conditions. It affects tree growth (competition between sphere (Ekblad and Högberg 2001, Bowling et al. 2002,
aboveground and belowground C sinks), the acquisition of Knohl et al. 2005, Brandes et al. 2006, Keitel et al. 2006,
resources (light, nutrients and water) that often limit forest Kodama et al. 2008, Marron et al. 2009, Wingate et al. 2010).
productivity and C sequestration in both standing biomass and Correlations between the isotope composition of respired
soil organic matter (Litton et al. 2007). Carbon allocation CO2 and either climatic drivers or online measurements of
results from several processes (Cannell and Dewar 1994); and daily photosynthetic C isotope discrimination can constrain
the term ‘allocation’ is thus often used to describe many differ- transfer rates of C from the foliage to different C pools in the
ent aspects of plant and ecosystem physiology, including pat- ecosystem (Wingate et al. 2010, Brüggemann et al. 2011).

terns in live biomass, the flux of C to a particular plant Because the strength of the natural signal is weak, precise
compartment and the distribution of flux as a fraction of gross estimates of transfer rates and time lags are hampered by
photosynthesis (Litton et al. 2007). post-photosynthetic fractionation (both biological and physi-
Carbon allocation has been estimated for decades with C cal like those occurring during CO2 transport through soil
mass-balance approaches, which typically combine measure- pores), by mixing of several C sources (recently assimilated
ments of standing biomass with measurements of respiratory versus stored C) and by mixing of autotrophic and heterotro-
CO2 efflux. For example, belowground C flux, including growth phic components of soil CO2 efflux (McDowell et al. 2004,
and respiration of roots and mycorrhiza, as well as exudates, Kodama et al. 2008, Wingate et al. 2010, Salmon et al. 2011).
can be estimated as the cumulative soil CO2 efflux minus C Such mixing and fractionation effects can cause up to 12‰
input from aboveground litter plus changes in C stored in roots, short-term variation in C isotope composition of plant and of
in the forest floor and in the soil (Giardina and Ryan 2002). soil-respired CO2 (Werner and Gessler 2011), sometimes
The aboveground C flux is often inferred from annual changes causing a complete loss of the photosynthetic isotopic signal
in aboveground biomass derived from allometric relationships from the canopy to the soil (Kodama et al. 2008).
and from measurements of respiration of aboveground organs The more promising technique for understanding C allocation
that are further scaled to the stand level (Ryan et al. 1996). is to artificially alter the C isotope content of assimilated C
Alternatively, the total aboveground C flux can be computed as using stable (13C) CO2 or radioactive (14C and 11C) CO2 as
the difference between gross primary productivity inferred short pulses or over long periods. In these labelling experi-
from eddy flux measurements of net ecosystem CO2 exchange ments, partitioning of photosynthesis products to sinks can be
and belowground C flux estimates (Navarro et al. 2008). While estimated as the amount of labelled C retained in a compart-
these approaches have proven to be fruitful for quantifying the ment or lost by respiration, exudation or volatile organic com-
whole ecosystem C budget, uncertainties remain about the pound emissions, relative to the amount of labelled C
contribution of different above- and belowground C fluxes to assimilated by the plant. In addition, allocation may also refer to
ecosystem respiration, which is the major efflux of C from the the partitioning of labelled C among several C-containing com-
biosphere to the atmosphere, and to C sequestration within the pounds in a given organ, for example between structural and
ecosystem. Budget-based mass-balance approaches also pro- non-structural C (Kagawa et al. 2005, 2006b), or between
vide limited insight into the short-term dynamics of C allocation storage compounds like starch and metabolites with high turn-
that are critical for understanding the mechanisms underlying over (Vizoso et al. 2008). Because accretion of labelled C in a
the annual patterns. One of the most important unanswered compartment results from the net flux of labelled C into versus
questions is the role of environmental drivers versus phenol- out of this compartment, the term allocation refers therefore to
ogy on changes in C allocation patterns at seasonal scales and both the dynamics and the amount of labelled C retrieved in a
how these two factors are affected by climate change. Given compartment (Figure 1). Following Litton et al. (2007), we
the major importance of belowground C allocation for soil pro- here use the term partitioning as a quantitative estimate of the
cesses, we also need a quantitative mechanism for the cou- fraction of labelled C supplied to a tree that is allocated to any
pling of belowground processes with canopy C assimilation. given compartment.
Recent technological developments offer promising approaches Carbon isotope labelling experiments have been imple-
to answer these important questions by directly tracing C mented for decades on potted trees (Lippu 1994, 1998,
fluxes with high temporal resolution. Maillard et al. 1994, Andersen and Rygiewicz 1995, Rouhier
Fluctuations of photosynthetic C isotope discrimination can et al. 1996) or at the branch level for field-grown trees
be used to trace the origin and fate of C into metabolites and (Schneider and Schmitz 1989, Hansen and Beck 1990,
respired CO2 at various temporal and spatial scales in plants Lacointe et al. 2004, Kagawa et al. 2005, Nogués et al. 2006,
Tree Physiology Online at http://www.treephys.oxfordjournals.org

778 Epron et al.
Grams et al. 2011, Kuptz et al. 2011, Ritter et al. 2011, Keel
et al. 2012) and mostly done on small-sized trees. However,
only such experiments offer the opportunity to better under-
stand the dynamic changes in C allocation in trees across sea-
sons, influenced by species phenology and environmental
conditions.
Recent development of novel tools to trace labelled C in
respired CO2 by isotope ratio infrared spectroscopy (IRIS) at a
high frequency has enabled precise quantification of C resi-
dence times in short-lived storage pools and of transfer rates
among plant compartments and between plants, soil and the
atmosphere (Bahn et al. 2009, Plain et al. 2009, Barthel et al.

2011, Dannoura et al. 2011, Epron et al. 2011, Warren et al.
2012). Combining long-term tracer time courses with compart-
mental modelling techniques is promising to estimate the num-
ber and the half-life of C pools contributing to C partitioning in
trees. Although information on half-lives, transfer rates and
mixing pools is highly relevant to improve our understanding of
allocation in tree–soil systems, reliable data are still scarce.
This review aims to (i) identify the experimental, method-
ological and analytical challenges to currently available pulse-
labelling techniques and give recommendations on practices
for designing future C pulse-labelling experiments, (ii) summa-
rize results from pulse-labelling experiments on trees to syn-
thesize current knowledge and elaborate emerging research
questions and (iii) present the classical mathematical methods
commonly used for characterizing the transfer of labelled C
into different compartments and fluxes and highlight the poten-
tial of more mechanistic approaches, based on compartmental
modelling of tracer time courses.
We synthesize data reported in 47 pulse-labelling studies
on trees using 11C, 13C or 14C as tracers under laboratory
Figure 1. Schematic representation of the fate of labelled C in a tree.
The assimilated labelled C [1] fills a pool of soluble carbohydrates that conditions and in the field (Table 1), in which the tracer was
is further allocated to respiration [2], storage [3], structural growth [4] applied either to branches or to trees of different sizes by
and defence compounds [5]. The soluble carbohydrate pools are also enclosing one or several trees in chambers or using free air C
filled by remobilization from reserves [6]. The emission of organic
compounds also potentially occurs [7]. Soluble sugars are transferred
isotope enrichment systems. These studies were conducted
both basipetally [8] from leaves to the other organs or acropetally from on a limited number of species (mainly beech and different
storage tissues in perennial organs to new shoots [9]. Labelled C is species of pine) from temperate (33 studies) and boreal (14
also transferred to the rhizosphere, either indirectly via hyphae of
studies) forests, while trees from tropical ecosystems have
mycorrhiza [10] or directly via exudation [11]. It can be transferred into
the microbial biomass [12] and further recycled owing to the high turn- not yet been assessed. These studies addressed several pro-
over rate of microbes [13]. The letters l, w, r and m indicate the com- cesses, including allocation to respiration, residence time,
partment (respectively leaf, woody organs, roots and microbes). within-tree and belowground transfer of photosynthates, par-
titioning of labelled C among organs and the use of stored C
Keel et al. 2007). But until recently, field labelling experiments (Figure 1). Long-term labelling experiments can elucidate
have been restricted to short-stature, homogeneous vegetation, processes that occur over the longer term, but they are not
such as crops and grasslands (Warembourg and Paul. 1973, within the scope of this review, although we do mention
Gregory and Atwell 1991, Bélanger et al. 1992, Ostle et al. important results from long-term labelling experiments wher-
2000, Johnson et al. 2002, Bahn et al. 2009, Gamnitzer et al. ever relevant.
2009, Lattanzi et al. 2012). In situ whole-tree labelling experi- The goals of C pulse-labelling experiments are typically to
ments are still in their infancy (Carbone et al. 2007, Högberg differentiate between environmental and biological controls on
et al. 2008, 2010, Plain et al. 2009, Andersen et al. 2010, C transfer and allocation, and to elucidate the contributions of
Endrulat et al. 2010, Dannoura et al. 2011, Epron et al. 2011, old (e.g., stored) and current assimilates to C sink activities.
Tree Physiology Volume 32, 2012

Table 1. Allocation of assimilated labelled C to different components and processes in trees and soil, as studied by pulse-labelling of trees.
Numbers in square brackets refer to the flux depicted in Figure 1. IsoFACE refers to a free air fumigation system. The study reporting 14C pulse-
labelling of tall trees was done in the frame of the EBIS project (see the text for details).
Flux Methods Type of tree Species References Main finding
Respiration [2] Pulse 14C,whole Seedlings Pine Hansen et al. Respiration is a major sink of
tree (1996) labelled C assimilated in winter
Pulse 14C, Shrub, <5 m Deciduous Carbone and Three pools of labelled C with
ecosystem perennial shrubs Trumbore (2007) different half-lives contribute to
respiration
Pulse 13C, crown Tree, <15 m Beech Plain et al. (2009) Two pools of labelled C with
different half-lives contribute to
trunk and soil CO2 efflux
Pulse 14C, whole Seedling Pine, spruce and Pumpanen et al. Roots and rhizosphere respire
tree birch (2009) 9–26% of assimilated C

IsoFACE Tree, >15 m Beech, spruce Andersen et al. Recent photosynthates
(2010) contribute differently to root
respiration in the two species
IsoFACE Tree, >15 m Beech, spruce Ritter et al. (2011) O3 exposure reduces allocation
of recent assimilates to
respiration in beech
Pulse 13C, crown Tree, <15 m Beech, oak, pine Epron et al. (2011) Allocation to soil respiration
varies seasonally depending on
species
Storage [3] Pulse 13C, crown Tree, < 1 m Chestnut Mordacq et al. Storage occurs in stump in late
(1986) summer
Pulse 14C, branch Tree, <5 m Pine Hansen and Beck In autumn, recent photosyn-
(1994) thates are transiently stored as
soluble C and supply root
growth in winter
Pulse 14C, whole Seedlings Pine Hansen et al. Starch builds up in winter in root
tree (1996)
Pulse 13C and Tree, <1 m Walnut Lacointe et al. Storage is an active sink, not a
14C, branch (2004) buffer for excess C
Pulse 13C, branch Tree, >15 m Broadleaved Keel et al. (2007) Recent C mixes fast with older C
species in some species, depending on
the nature of storage
compounds
IsoFACE Tree, >15 m Beech, spruce Kuptz et al. (2011) Storage is an important sink at
the end of summer in beech
Pulse 13C, Tree, <5 m Pine Keel et al. (2012) Allocation and residence time
ecosystem are different in fine roots and
root tips
Structural growth Pulse 13C, crown Tree, <1 m Chestnut Mordacq et al. Aboveground growth is the main
[4] (1986) sink of recent C in early summer
Pulse 14C, Tree, <5 m Pine Hansen and Beck Recent photosynthates of
branch (1994) previous year needles supply
sprouting at bud break and
secondary growth in summer
Pulse 14C, Seedling Pine Lippu (1994) Carbon allocation to root is high
whole tree before and after the period of
intensive shoot growth
Pulse 14C, Tree, <5 m Poplar Horwath et al. Allocation shifts between
whole tree (1994) aboveground in early summer
and belowground in late
summer
Pulse 13C, crown Tree, <1 m Larch Kagawa et al. Allocation shifts between
(2006a) aboveground in early summer
and belowground in late summer
Pulse 13C, branch Tree, >15 m Beech, hornbeam Hoch and Keel Infructescence photosynthesis
(2006) contributes to fruit development
in hornbeam, not in beech
(continued)

780 Epron et al.
Table 1. (Continued)

Pulse 14C, whole Seedling Pine, spruce and Pumpanen et al. Root (and ectomycorrhiza)
tree birch (2009) growth uses 13–21% of recently
assimilated C
Pulse 13C, whole Tree, <1 m Fir Endrulat et al. Carbon allocation to cellulose in
tree (2010) roots is independent of fine root
diameter
Pulse 13C, branch Tree, >15 m Oak, beech, lime Keel and Schädel Expanding leaves are strong sink
tree (2010) of recent photosynthates,
especially in lime tree
Pulse 1C, whole Cuttings Birch Kasurinen et al. O3 shifts C allocation from above
tree (2012) to belowground
Remobilization Pulse 14C, branch Tree, <5 m Pine Hansen and Beck Early wood formation partly

from reserve [6] (1990) relies on stored C
Pulse 14C, whole Seedling Pine Lippu (1998) Stored C supports root
tree respiration in winter
Pulse 13C & 14C, Tree, <1 m Walnut Lacointe et al. Stored C supports sprouting in
branch 2004 spring
Pulse 13C, branch Tree, <15 m Beech Nogués et al. Stored C contributes to leaf
(2006) respiration in a proportion
varying seasonally
Pulse 13C, whole Tree, <1 m Larch Kagawa et al. Half C in new needles is derived
tree (2006a) from storage
Pulse 14C, Tree, <5 m Spruce Carbone et al. Rhizosphere respiration relies
ecosystem (2007) more on stored C than on recent
photosynthate
Pulse 13C, whole Seedling Oak Vizoso et al. 40–85% of C in new leaves
tree (2008) derives from stored C
Pulse 14C, EBIS Tree, >15 m Broadleaved Gaudinski et al. 50–60% of fine root growth is
species (2009) supported by stored C
Pulse 13C, whole Tree, <1 m Fir Endrulat et al. Stored C supports new root
tree (2010) growth in next spring
Pulse 14C, whole Seedling Acacia Schutz et al. Root starch supports shoot
tree (2009) growth after fire-induced stem
death
IsoFACE Tree, >15 m Beech, spruce Kuptz et al. (2011) Trunk respiration is mostly
supplied by stored C in spruce
all year round, but only in spring
in beech
IsoFACE Tree, >15 m Beech Grams et al. (2011) Remobilized C contributes
significantly to phloem sugars in
summer
BVOC emission [7] Pulse 13C, whole Seedlings Pine, spruce, Ghirardo et al. The fraction of BVOC emissions
tree larch, birch (2010) originating from recent photo-
synthates or from stored C
varies among species
Basipetal phloem Pulse 11C, leaf Seedling Ash, elm, spruce, Thompson et al. Transfer rate is higher in
transport [8] pine (1979) broadleaved than in coniferous
species
Pulse 11C, leaf Seedling Ash, rowan Jahnke et al. (1998) Differences in anatomy account
for difference in transfer rate
between the two species
Pulse 13C, branch Tree, <5 m Japanese cedar Kagawa et al. Sieve cell connections change
(2005) seasonally
Pulse 13C, branch Tree, <5 m Larch Kagawa et al. Spiral translocation of photosyn-
(2006b) thate occurs in this species
Pulse 13C, Tree, <5 m Pine Högberg et al. Half-life of labelled C in phloem
ecosystem (2008) sap is short (1.3 day)
Pulse 13C, whole Sapling Beech Ruehr et al. (2009) Transfer rate is reduced by
tree drought
(continued)

Table 1. (Continued)

Pulse 13C, crown Tree, <15 m Beech, oak, pine Dannoura et al. Transfer rate is higher in
(2011) broadleaved than in coniferous
species
Pulse 13C, whole Sapling Beech Barthel et al. Transfer rate is reduced by
tree (2011) drought
Pulse 13C, whole Tree, <15 m Pine Warren et al. Transfer rate is not affected by
tree (2012) shading
Acropetal phloem Pulse 14C, whole Tree, <1 m Larch Schneider and Labelled C moves acropetally
transport [9] tree Schmitz (1989) during sprouting
Transfer to soil Pulse 14C, whole Seedling Pine Andersen and O3 decreases C transfer to the
biota [10–13] tree Rygiewicz (1995) fungal symbiont

Pulse 13C and Seedling Birch and fir Simard et al. Bidirectional transfer of C occurs
14C, whole tree (1997a, 1997b) between seedlings
Pulse 14C, tree Seedling Pine Heinonsalo et al. The sink strength depends on
(2010) the root-associated mycorrhizal
fungal symbiont
Pulse 14C, tree Tree, <1 m Aspen Mikan et al. (2000) Elevated CO2 increases transfer
of labelled C to microbial
biomass
Pulse 14C, tree Seedling Pine Leake et al. (2001) Mycorrhizal mycelia allocate C to
fine roots foraging for nutrients
Pulse 13C, Tree, <5 m Pine Högberg et al. Recent C is rapidly found in
ecosystem (2008) microbial cytoplasm
Pulse 13C and Seedling, in Fir Teste et al. (2010) Transfer of C between seedlings
14C, tree the field exists but it is small and it
depends on disturbance
Pulse 13C, Tree, <5 m Pine Högberg et al. Transfer to soil microbes is
ecosystem (2010) higher in late than in early
summer and is reduced by
nitrogen fertilization
Pulse 13C, Tree, <15 m Beech, oak, pine Epron et al. Recent C is rapidly found in
crown (2011) ectomycorrhiza and microbial
cytoplasm
BVOC = biogenic volatile organic compounds.
We specifically investigate (i) how fast the assimilated C is were applied to branches, potted trees or field-grown trees of
transferred belowground, how the rate of C transfer differs different sizes (Table 1). We address whether these methods
among species and how this transfer rate responds to envi- are adequate to investigate short-term C allocation patterns
ronmental factors; (ii) how residence times of labelled C differ and based on past and current works, we recommend best
among tree compartments depending on their source/sink practices for designing future C pulse-labelling experiments.
status and on the respective season; (iii) how climate and
environmental stresses influence C allocation to different sink Carbon isotopes as tracers
organs and whether we are able to disentangle the effects of Labelling plants with isotopically enriched CO2 has been used
phenology from those mediated by environmental factors; and for decades to study the fate of C into photosynthetic products
(iv) how stored versus current assimilates contribute to in leaves. One historical example is the experiments that were
growth and respiration demands in sink organs, owing to the conducted in the 1950s by Calvin and co-workers leading to the
role of C storage in perennial species. discovery of the photosynthetic C reduction cycle (Calvin
1961). 14C (half-life of 5730 years) was initially used because
tools were available for tracing the radioactivity in the labelled
Pulse-labelling of trees
compounds (autoradiography, liquid scintillation (LS)
Pulse-labelling of trees with labelled CO2 allows C fluxes to be spectrometry and accelerator mass spectrometry (AMS)). The
traced in the tree–soil system. Different forms of labelled C precision of the LS counter technique is adequate for studies

782 Epron et al.
conducted in mesocosms in the laboratory where the β-radiation and to the soil, especially if the main interest is in the seasonal
emissions resulting from the decay of 14C in the label are high changes in C allocation dynamics (Table 1). The newly assimi-
enough to detect, but remain under the safety limits. More lated C is predominantly allocated to active metabolic sinks.
recently, the development of AMS has enabled the use of much These sinks can also attract C remobilized from storage com-
lower amounts of radioactivity opening the doors for in situ partments (Paterson et al. 2009). Depending on when the
labelling experiments (Carbone et al. 2007, Carbone and labelling is done and until when the labelled C is further traced,
Trumbore 2007), despite substantially higher costs per anal- each of these sources can be studied.
ysed sample compared with LS. Large releases of 14CO2 from a Furthermore, pulse-labelling can also estimate the mean
hazardous waste incinerator have unintentionally pulse-labelled residence time and the half-life time of C in compartments. The
a mature deciduous forest (EBIS project, http://ebis.ornl.gov/) mean residence time corresponds to the C stock to C flux ratio
and 14C was successfully used to infer residence time and turn- and it is not an intrinsic property of a compartment. Pools with
over of C in various forest ecosystem compartments, e.g., roots long residence time have either a large C stock or low exchange

(Joslin et al. 2006, Gaudinski et al. 2009). Another radioactive rates (or both). Long-term labelling approaches are more
C isotope, 11C, with very short half-life (about 20.4 min), has appropriate to estimate residence times in pools that turnover
proven to be a useful tracer for studying the response of trans- slowly.
location of photosynthates in small plants to fast changes in Long-term labelling experiments have so far been restricted
environmental conditions or to source–sink manipulations. The to young, small potted trees because of the high cost of con-
advantage of 11C is that it allows multiple labelling of the same tinuously flushing the labelling chamber with isotopically
plants at very short time intervals (Roeb and Britz 1991, Thorpe enriched CO2 (Maillard et al. 1994, Dyckmans and Flessa
et al. 2011). However, 11C has been rarely used and only for 2001, Dijkstra and Cheng 2007, Guérard et al. 2007,
tree seedlings because the need to produce 11CO2 and detect Esperschütz et al. 2009b, Palacio et al. 2011). The free air CO2
11C restricts its use to specifically equipped laboratories enrichment (FACE) studies, which were designed to study the
(Thompson et al. 1979, Jahnke et al. 1998, 2009). effects of elevated CO2 on ecosystems (Palmroth et al. 2006,
For pulse-labelling trees in the field, one advantage of 14CO2 Millard et al. 2007) gave opportunities to follow the fate of
compared with 13CO2 is that very small amounts of 14CO2 can assimilated C over longer times because industrial CO2 is 13C
be supplied to the foliage because of the high sensitivity of depleted compared with ambient CO2 (Keel et al. 2006, von
AMS and the very weak environmental background for 14C com- Felten et al. 2007, Bader et al. 2009). However, elevated CO2
pared with 13C in the atmosphere and in the plant–soil system. may shift C partitioning towards belowground biomass com-
In contrast, when trees are pulse-labelled with 13CO2, either the partments, and thus bias estimates of C allocation. Such CO2
CO2 concentration has to be raised above the ambient level or effects are still unclear and may depend on the genetically
the CO2 concentration in the labelling chamber has to be determined development and growth of the species, the devel-
decreased prior to adding the label. Nevertheless, because of opmental stage of the tree and the availability of resources
the high cost of sample preparation and of AMS analyses of 14C, other than C at the site (Körner 2006, Dieleman et al. 2010). In
using 13CO2 for pulse-labelling is much more cost effective than addition, labelling starts when the CO2 treatment begins, there-
using 14CO2, and laws restrict the field use of the radioactive fore any dynamics revealed by the ‘tracer’ corresponds to a
isotopes in many countries. The use of 13CO2 for studying the transient system that is adapting its growth, allocation and
fate of C in the plant–soil system built upon the lack of restric- storage patterns in response to elevated CO2, and thus limiting
tions of its use, and the increased availability of stable isotope inferences regarding mechanisms for ambient, elevated or
ratio mass spectrometry since the early 1990s. slowly changing CO2 concentrations.
Dual labelling with both 13C and 14C is a powerful approach for Free air systems (IsoFACE) have been specifically devel-
studying C flux and turnover in pools that differ in turnover rates: oped for labelling small trees in the field (Talhelm et al. 2007)
the low cost of 13C analysis allows frequent samplings for pools but also tall trees in situ (Grams et al. 2011) for several days or
with high turnover rates and 14C can be used for infrequent sam- weeks. Although the labelling duration is shorter than typically
plings of pools with slower turnover rates (Lacointe et al. 2004, used in long-term labelling experiments, trees are however
Carbone and Trumbore 2007). Dual labelling with 13C and 14C exposed to above ambient CO2 concentrations for longer peri-
can also be used to study bi-directional transfer of C, e.g., ods than in classical pulse-labelling approaches. While this lon-
between tree seedlings (Simard et al. 1997b, Teste et al. 2010). ger labelling period may be useful for estimating half-lives in
pools that turn over slowly and bypasses the effects of diurnal
Potential and limitations of the pulse-labelling allocation changes, it is of limited value for studying the rapid
approaches dynamics of C transfer among pools or compartments.
The pulse-labelling approach is appropriate for studying how An alternative pulse-labelling approach based on the injec-
fast and where C fixed in photosynthesis moves into the tree tion of dissolved 13C-carbonate into the xylem was recently

applied on small cedar trees (Powers and Marshall 2011). This In IsoFACE systems, trees are not enclosed in chambers.
technique might become a cheap and easy option for pulse- However, even using 99% 13CO2, the enrichment of the plant
labelling big trees in remote areas. Xylem-delivered carbonates that can be achieved is relatively low, while the pulse duration
are thought to provide CO2 to the leaves where it will be fixed needs to be quite long (5 days in Talhelm et al. 2007). Thus,
by photosynthesis and transported into the other plant com- the costs for labelled 13CO2 are clearly restricting the use of
partments. However, labelled organic C from non-photosyn- this approach for taller trees. Although costs could be signifi-
thetic, anaplerotic CO2 fixation in the trunk might lead to cantly reduced using 13C-depleted CO2 from fossil fuels, the
misinterpretations of assimilate transport and transport labelling signal would be relatively small compared with that of
velocities. enriched 13CO2 and again require long labelling periods (Grams
et al. 2011).
Implementing setups for pulse-labelling
Because of their simplicity, closed systems have been used in Whole-tree pulse-labelling

most pulse-labelling experiments with trees. A suitable label- For the practical reasons mentioned above, pulse-labelling
ling chamber, whatever its size, should be airtight to limit the experiments have largely been restricted to seedlings and
amount of labelled CO2 lost and not taken up by the foliage. small trees ex situ until recently (Table 1). However, C alloca-
Since CO2 and water vapour concentrations inside the cham- tion rules are thought to change with tree age and to be influ-
ber will change rapidly over time because of leaf photosyn- enced by neighbouring trees (Litton et al. 2007, Poorter et al.
thesis and transpiration, they should be monitored. 2012), requiring in situ labelling of trees. For several decades,
Furthermore, it is strongly advised to regulate both air tem- in situ labelling was restricted to individual branches, ever
perature and air humidity inside the chamber using air-condi- since the pioneering work on mature white oak trees
tioning devices (Högberg et al. 2008, Plain et al. 2009), (McLaughlin et al. 1979) and orchard trees (Hansen 1970).
since high air temperatures, potentially damaging the photo- The main limitation of branch labelling is the mixing of labelled
synthetic apparatus, are expected in closed chambers under carbohydrates with a large amount of unlabelled carbohy-
high solar radiation. This is one of the reasons why the label- drates coming from the remaining branches of the crown.
ling duration should be as short as possible and, if the cool- Recovery of labelled C at the whole-tree level or in the soil is
ing capacity is insufficient, should be done quite early in the therefore compromised. Despite its limitation, branch labelling
morning. However, allocation patterns to different foliar C was useful to study the transition from heterotrophy to autot-
pools (e.g., transient starch storage, respiration, VOC emis- rophy after leaf emergence (Keel and Schädel 2010), branch
sion) may change over the course of the day and thus need autonomy (Lacointe et al. 2004) and the mixing of old and
to be considered for data interpretation. new C pools in the branch tissues (Keel et al. 2007), and to
While the injection of 13CO2 tracer can be regulated by mass reveal spiral translocation paths in the stem of larch saplings
flow controllers to maintain CO2 concentrations within the (Kagawa et al. 2006b).
chamber at ambient level, additional unlabelled CO2 should be Until recently, whole-tree labelling in the field was restricted
injected into the chamber when using 14CO2 to compensate for to short coniferous trees in the boreal forest (Kagawa et al.
photosynthetic uptake, maintaining CO2 concentrations at the 2006a, Carbone et al. 2007, Högberg et al. 2008, 2010, Keel
ambient level. In both cases, mixing of air is needed to enable et al. 2012) or to woody shrubs in semi-arid ecosystems
a uniform distribution of labelled CO2 inside the chamber (Carbone and Trumbore 2007). More recently, the crowns of
(Carbone et al. 2007). 10 m tall trees (Plain et al. 2009, Dannoura et al. 2011, Epron
Using open systems could solve many of the problems et al. 2011) were successfully pulse-labelled. Two options have
described above: ambient CO2 concentrations would not been proposed for labelling whole trees in the field: large can-
change, enrichment would be constant and chambers would opy chambers covering 10–50 m2 of soil including a small ‘for-
not need to be air-tight, although the consumption of labelled est ecosystem patch’ with several short trees (Carbone et al.
CO2 would be significantly larger. At optimized air flow, changes 2007, Högberg et al. 2008) and chambers in which the entire
in air humidity and temperature inside labelling chambers crown of a single tree is enclosed while excluding the soil sur-
would be minimal, even under high irradiance. But until now, face (Plain et al. 2009). Crown chambers allow labelling of
open systems have been used only in grasslands (Lattanzi trees only one by one and require manpower when moving
et al. 2012) or for long-term labelling of potted, small trees them from tree to tree, limiting the number of replicates.
(Dyckmans and Flessa 2001, Esperschütz et al. 2009b, Moreover, trees are often labelled on different days under dif-
Palacio et al. 2011), although in principle, they are suitable for ferent environmental conditions (Dannoura et al. 2011, Epron
short-term labelling of several large chambers operating simul- et al. 2011). In contrast, canopy chambers allow labelling of
taneously on large trees, depending only on the size of the several trees at the same time (which are in fact pseudo-repli-
air-generating unit. cates), although being restricted to short trees. A disadvantage

784 Epron et al.
of canopy chambers is back-diffusion, i.e., labelled CO2 enters by a large pool of unlabelled C. This is more challenging with
the soil pores during the labelling and diffuses back out several 13CO
2 than with 14CO2 because of their respective atmospheric
hours later, potentially confounding recovery of labelled CO2 in backgrounds and detection limits (isotope ratio mass spec-
belowground respiration (Subke et al. 2009). However, back trometry (IRMS) versus AMS). Thus, a trade-off exists between
diffusion of labelled CO2 can be monitored on root-free collars the length of the labelling period and the level of enrichment
inside the chamber. Furthermore, labelled CO2 entering soil used, since exposing trees to CO2 concentrations well above
pores may be assimilated by the soil microbes (Miltner et al. ambient levels should be avoided. Increasing the amount of
2004) or even by the roots through the anaplerotic pathway labelled CO2 without increasing the total chamber CO2 concen-
(Ford et al. 2007, Gessler et al. 2009). A similar artefact could trations could be done by reducing chamber CO2 concentra-
also be expected for bark, into which labelled CO2 might diffuse tions prior to labelling either by removing air inside the chamber
during labelling and there be assimilated by bark photosynthe- (Keel et al. 2007), scrubbing CO2 with soda lime (Ruehr et al.
sis or anaplerotic reactions. 2009) or waiting for reductions in CO2 concentrations through

The duration of the pulse should be as short as possible for photosynthesis (Plain et al. 2009).
a precise determination of time lags between the start of the
labelling and the first appearance of labelled compounds in Monitoring the amount of labelled C delivered to the tree
sink compartments or respiratory efflux (Figure 2). The shape While the isotope composition of CO2 within the labelling
of the curve is affected by the duration of the labelling, and chamber can be known a posteriori by taking air samples from
this impact is amplified by the radial velocity profile in the the headspace frequently and analysing these samples, moni-
sieve tubes of the phloem due to friction along the tube wall toring total CO2 concentration during the labelling is more chal-
(Dannoura et al. 2011) and by the mixing of labelled and unla- lenging because the new generation of infrared gas analysers
belled C. For this reason, we strongly recommend using a (IRGAs) is quite specific to 12CO2 and rather insensitive to
constant labelling duration among treatments and replicates, if 13CO and 14CO (Tohjima et al. 2009). While this is not an
2 2
the main objective is to study the rate of transfer and resi- issue for 14CO2 labelling because the amount of 14CO2 remains
dence times of C in labile pools. On the other hand, if the main low compared with the amount of 12CO2, the problem is more
objective is to study allocation among plant compartments, serious for 13CO2 labelling, since 13CO2 is typically the major
using the same amount of labelled substrate for each tree is isotopologue of CO2 in the chamber during labelling. One
defensible, even if it may change the duration of the labelling option is to use an old generation IRGA and accurately mea-
depending on the photosynthetic activity of the tree (Epron sure total CO2 concentrations. Another option is using two
et al. 2011). IRGAs with different 13CO2 sensitivities. A further option is to
The amount of label should be high enough, especially in the use a dual IRGA that measures both 12CO2 and 13CO2 concen-
field, to be detected in compartments where it will be diluted trations within the same concentration range (Plain et al.
2009). Isotope ratio infrared spectrometers are not necessarily
suitable for measuring high 13CO2 concentration in the labelling
chamber, unless the air stream coming from the chamber is
strongly diluted with 13CO2-free air before entering the IRIS.
Tracing the labelled C

The goal of pulse-labelling experiments is to trace the labelled
C within the tree–soil system. Technologies are developing fast
but sampling issues remain the Achilles heel of these experi-
ments, especially under field conditions.
Expression of isotope composition

While the isotope composition at natural abundance levels is
often expressed as a ratio relative to an international reference
Figure 2. Schematic diagram showing the exponential, first-order
standard (δ13C, Δ14C), the isotope composition of an enriched
kinetics of a tracer (13C) in any plant compartment with a mean resi-
dence time of 2 (t1/2 = 1.4, continuous line), 15 (t1/2 = 10.4, dotted compartment can be better expressed as percent atom excess
line) or 40 (t1/2 = 27.6, dashed line) time units, as typically observed in (Dawson et al. 2002). This value is defined as the relative
pulse-labelling experiments. The time unit on the x-axis is arbitrary. abundance of the heavier isotope in a labelled compartment
The labelled CO2 is introduced at time 0 (vertical dashed bar) as a
pulse. No time lag is considered between labelling and tracer exceeding the natural isotope abundance in the same unla-
recovery. belled compartment.

Percent atom excess = (AbL − AbUL ) × 100 isotope ratios but the mixing ratios of individual isotopo-
logues (e.g., 12CO2 and 13CO2). The working standards should
therefore bracket the expected mixing ratios of both
H
C isotopologues.
with Ab = 12
C+HC
Labelled C recovered in trees
Ab is the relative abundance of the heavy isotope in the In order to properly trace the labelled C into the bulk material
labelled (L) and unlabelled (UL) compartments and HC is either of trees the sampling scheme must allow capturing both the
13C or 14C. For 13C, Ab can be calculated directly from the con- label peak and the long lasting tail of the recovery kinetics.
centration of both isotopologues (e.g., measured by IRIS) or Since an exponential decay of the isotope signal might be
derived from the δ values (IRMS) and the isotope ratio of the expected, it is advisable to take samples frequently at the
international reference standard (Rref): beginning of the experiment—starting immediately after

labelling—and to reduce the sampling rate progressively fol-
Ab =
((δ13C)/(1000) + 1) × Rref lowing a geometric law. Studying allocation will require sam-
((δ13C)/(1000) + 1) × Rref  + 1 pling of foliage, twigs, branches, trunk cores as well as coarse
  and fine roots. Soil samples are needed for the analysis of
labelled C in microbial biomass and soil organisms, in group-
specific membrane-bound fatty acids or putatively in nucleic
Labelled C in the CO2 efflux acids.
Several approaches are commonly used for tracing labelled Special care should be taken into account for the spatial
CO2 in the respiratory efflux: IRMS or IRIS are used for 13C heterogeneity, especially for fine roots retrieved from soil
and LS or AMS for 14C. When using laboratory IRMS facilities, cores (Endrulat et al. 2010). Two main points need to be con-
CO2 efflux is monitored using closed-loop chamber systems sidered when sampling fine roots in the field: dead fine roots
and the air is sampled at different times after closing the should be disregarded and the living roots sampled should
chamber (i.e., at different CO2 concentration in the chamber). only belong to the labelled trees. The labelled trees can be
The C isotope composition of the respiratory CO2 can be isolated for at least several months before labelling by dig-
deduced from the Keeling-plot intercept (Högberg et al. ging a trench around the tree. The trench needs to be lined
2008, Ruehr et al. 2009). The main limitation of this with a polyethylene foil and refilled with soil. This approach
approach is the low frequency of sampling and the analysis ensures that all roots and root exudates within the known soil
costs. Mobile IRMS facilities and flow through chambers volume originate from the isolated tree and that all active
allow continuous measurement of δ13C of CO2 in the respira- roots of the labelled tree are contained inside this trenched
tory efflux on site, thus increasing the measurement fre- area (e.g., Plain et al. 2009). However, regrowth of roots to
quency (Schnyder et al. 2004, Subke et al. 2009). More re-establish the root-to-shoot ratio might take much longer
recently, IRIS has offered the opportunity to sharply increase depending on the intensity of the disturbance. The size of the
the measurement frequency at reduced costs (Bahn et al. delimited area should be large enough to avoid any water
2009, Marron et al. 2009, Plain et al. 2009). The frequency stress resulting from a reduction of the size of the root sys-
of the measurements depends on the number of chambers tem. This may not be possible with species that spread their
connected to the device, but typically 10 chambers can be roots far from the trunk or with deep-rooting species. The
measured within 1 h or less. Several IRIS technologies are problem of sampling roots of unlabelled trees is also avoided
now available including tuneable diode laser systems (e.g., when using large chambers that cover e.g., 50 m2 or more of
TGA 100 A, Campbell Scientific, Logan, UT, USA), quantum soil, instead of single crown chambers. Högberg et al. (2008)
cascade laser systems (e.g., CO2 Isotope Trace Gas Monitor, assessed the influence of unlabelled trees by labelling the soil
Aerodyne, Billerica, MA, USA), cavity ring down spectros- in the studied central area of the plots with 15N, and traced
copy (e.g., G2131, Picarro, Santa Clara, CA, USA), off axis the 15N label in the tree canopies inside and just outside the
integrated cavity output spectroscopy (e.g., CCIA-36, Los plots, assuming that the transfer of N in acropetal direction
Gatos Research, Mountain View, CA, USA) or Fourier trans- was most likely associated with the transfer of C in the
form infrared spectroscopy (e.g., Vector 22, Bruker Optics, reverse direction. A more laborious option is to genetically
Ettlingen, Germany), all presenting advantages and disad- identify the target tree species from root fragments (Endrulat
vantages that are not within the scope of this review. et al. 2010).
Calibration is an important issue whatever technology is Plant organs are composed of various C pools of different
used because a wide range of C isotope compositions is ages and with different residence times, and it is worthwhile
expected. In contrast to IRMS, IRIS does not measure the to study each pool individually to assess the dynamics of each

786 Epron et al.
pool and the rates of exchange between the pools. It is at kinetically distinct pools in a metabolic network. Robust
least interesting to isolate the soluble fraction that contains a hypotheses accounting for the effects of environmental factors
mixture of sugars, organic acids and amino acids (Brandes on—and species differences in—the rate of transfer have
et al. 2006, Ruehr et al. 2009), since starch and structural C already emerged. An increase in the measurement frequency
have different turnover rates. It is also an option to collect of labelled C in the CO2 efflux combined with compartmental
phloem sap samples either directly by making shallow incision analysis of tracer time courses is promising for characterizing
into the bark (bleeding techniques) in suitable species, by the half-life and the relative importance of the different sub-
exudation of soluble compounds from small pieces of bark tis- strates that fuel respiration.
sues punched from the trunk or by using aphids as sap collec-
tors (Yoneyama et al. 1997, Gessler et al. 2004, Dannoura The pattern of recovery: what a shape can tell
et al. 2011). The shape of the recovery of labelled C in a compartment or in
the respiratory efflux has a few distinct phases (Figure 2).
Labelled C recovered in soil microorganisms

Tracer appearance is typically delayed relative to the labelling
Because recent photosynthates are rapidly transferred to the pulse (time lag). Then, the amount of labelled C increases until
soil microbial compartments (Högberg and Read 2006, a maximum is reached (peak), and subsequently decreases
Högberg et al. 2010), either directly through the mycorrhizal more or less exponentially until eventually no tracer is left (new
network or through root exudation, labelled C is expected to isotopic equilibrium).
be retrieved in soil soluble organic matter and in the microbial The time lag can vary from a few minutes to several hours or
biomass. Soluble organic C can be extracted with a weak saline days. This, of course, depends on the distance between the
solution (typically K2SO4) from fresh soil samples. Microbial C source leaves and the observed sink. It also depends on the
can also be extracted from fresh soil samples that have previ- rate of transport of C within the plant and into the soil. Recent
ously been fumigated with chloroform to disrupt the microbial studies show that transport rates vary depending on species,
membranes (Vance et al. 1987). The difference in labelled C phenology and environmental conditions (Ruehr et al. 2009,
between fumigated and non-fumigated soil samples is sup- Dannoura et al. 2011, Epron et al. 2011).
posed to be the microbial C released from microorganisms that The shape of tracer time courses reflects both (i) the extent
have been destroyed by the fumigation (Rangel-Castro et al. to which labelled C has mixed with unlabelled C pools before
2005b, Ruehr et al. 2009, Epron et al. 2011). This approach arriving in the sampled compartment (i.e., the number and
does not discriminate between fungi, bacteria and Archaea. turnover rate of those mixing pools), and (ii) the turnover rate
The 13C transferred to the mycorrhizal hyphae network can be of the observed compartment itself. Thus, time courses char-
retrieved in soil cores surrounded by an appropriate nylon acterized by sharp and rapid increases and decreases are
mesh (Johnson et al. 2002, Epron et al. 2011). Molecular indicative of pools closely associated to (mainly fed by) current
approaches to trace the 13C or 14C in nucleic acids (Ostle et al. assimilates. As the label is transferred along the translocation
2003, Rangel-Castro et al. 2005a), amino-sugars or fatty path, for instance, from the top of the tree to belowground
acids (Treonis et al. 2004, Amelung et al. 2008, Denef et al. sinks (Dannoura et al. 2011), the shape of the tracer time
2009, Esperschütz et al. 2009b, Högberg et al. 2010) are course is changing. The often observed slower rates of
promising to identify the microbial groups that strongly respond increase, the lower and broader peaks, and the slower rates of
to tree photosynthesis. They are, however, challenging in the decrease reflect the fact that the tracer is mixing with more
field because of the dilution of the tracer within the large soil C and more pools as it travels downwards.
pool (Griffiths et al. 2004). While pulse-labelling experiments A clear record of the shape of the peak requires a high
are normally sufficient for tracing 13C from plants to phospho- measurement frequency, in particular before and after the
lipid fatty acids of microbial groups, longer labelling periods maximum when rates of change are fast. For compartments
might be required for probing 13C in microbial DNA and RNA close to C assimilation, such as soluble carbohydrates in pho-
for the identification of functional or taxonomic groups (Neufeld tosynthetic leaves, this implies that sampling should ideally
et al. 2007) in forest soil. start during labelling. It also requires a sustained sampling
effort throughout the exponential decrease, so that all acting
pools can be detected. For tissues with a slow turnover rate,
Dynamics of photosynthates in tree and soil
such as the structural biomass of roots, sampling should
compartments
thus last for several months (although not necessarily at high
Tracing labelled C after pulse-labelling a tree allows quantifying frequency).
at least three important aspects of whole-plant C metabolism: Of course, the actual values of tracer content are strongly
the rate of transfer of C substrates between compartments, the affected by the duration and enrichment of the labelling pulse.
residence time of C in these compartments and the number of Therefore, direct comparisons of tracer contents among

treatments, or even among individuals within a given treat-

ment, are often meaningless. Yet, the parameters that can be
derived from the shape of the tracer timecourse—i.e., the
time lag, the time to peak, the rate of exponential decrease
and the time to new isotopic equilibrium—would be
invariant.
Variations in the rate of C transfer among species

The rate of C transfer (flux [8] in Figure 1) is often calculated
from the time lag between C uptake (i.e., labelling time) and
the recovery of the label in the CO2 efflux in the soil, or from
differences in time lags between the start of the labelling and

the appearance of 13C in CO2 efflux measured at different posi-
tions along the trunk or along a coarse root (Dannoura et al.
2011). The rate of C transfer in the tree as well as between the
canopy and the soil varies greatly among different labelling
experiments and numerous factors may account for this large
range of variation.
One source of variation that explains differences in time
lags is undoubtedly the time resolution of measurements. The
time lag between uptake and recovery also depends on the
size of the tree, i.e., on the distance between the source and
Figure 3. Time lag of peak 13CO2 efflux from soil (or coarse roots).
the sink, and on the species (Figure 3a). There are indeed (a) For different tree species and sizes. The size of the box reflects
clear species differences with coniferous species exhibiting both the seasonal variability and uncertainties due to low measure-
on average 10 times lower rates of label transfer than broad- ment frequency (data from Horwath et al. 1994, Carbone et al. 2007,
Carbone and Trumbore 2007, Högberg et al. 2008, 2010, Barthel
leaved species (Thompson et al. 1979, Schneider and Schmitz et al. 2011, Epron et al. 2011, Kuptz et al. 2011, Warren et al. 2012).
1989, Jahnke et al. 1998, Högberg et al. 2008, Plain et al. (b) Relationships between the lag before labelled C peaks in soil CO2
2009, Dannoura et al. 2011, Table 2). Pulse label experiments efflux (Epron et al. 2011) and the rate of C transfer in the trunk
(Dannoura et al. 2011) in 10 m tall beech (closed circles), oak (open
thus confirm observations based on the transfer of weather- diamonds) and pine (closed triangles) trees pulse-labelled at different
induced change in photosynthetic isotope discrimination sig- times of the year.
nals (Kuzyakov and Gavrichkova 2010, Mencuccini and Hölttä
2010) and most likely reflect differences in phloem anatomy
(Jensen et al. 2012). When the rate of transfer to the trunk is transport also varies within a tree, with higher speed at the top
estimated with the same method for trees of similar size, the of the trunk than at the bottom.
relation between the rate of C transfer and the time lag of The rate of transfer is also strongly related to environmental
peak 13CO2 efflux from the soil differs markedly between conditions and is decreasing with lower temperature and soil
broadleaved species (oak and beech) and pine (Figure 3b). water content (Ruehr et al. 2009, Barthel et al. 2011, Dannoura
Differences in soil macroporosity were unlikely to explain the et al. 2011). In contrast, shade did not affect the rate of C trans-
differences in the rate of transfer in this study. The results fer in pine (Warren et al. 2012). Changes in phloem sap viscosity
suggests a longer retention time in the pine foliage or the and changes in the turgor pressure gradients between source
roots that may be due to strong mixing of new, labelled C with and sink organs are thought to account for these changes. The
old unlabelled C. changes in turgor gradients are mediated by tree transpiration,
xylem water potential and by the activity of source and sink
Responses of the rate of transfer to environmental factors organs where phloem loading and unloading occur. However,
The velocity itself might also be size dependent, with higher the processes that control the sink activity and the feedbacks on
rates of transfer in tall compared with small trees. For instance, source activity, and therefore the long distance transport of C in
13C label in soil microbial biomass was detected one day after the tree, are still poorly understood at the whole-plant level.
labelling in beech saplings of 0.5 m height (Ruehr et al. 2009)
as well as in 10 m tall beech trees (Epron et al. 2011). This Fast belowground C transfer
finding is substantiated by the observation that in beech the Whatever the species differences in transport velocity are, all
velocity of C transfer in the trunk is positively related to the results highlight a rather fast transfer of recent assimilates to
diameter of the tree (Dannoura et al. 2011). The velocity of C the soil biota (fluxes [10-13] in Figure 1). The labelled C is

788 Epron et al.
Table 2. Rates of C transfer in broadleaved and coniferous tree species, as estimated from pulse-labelling experiments.
Species Methods References Velocity (m h−1)
Broadleaved Ash, elm 11C (seedlings) Thompson et al. (1979) 0.3–6.0

Ash, rowan 11C (seedlings) Jahnke et al. (1998) 0.3–1.3
Beech 13C (10 m tall trees) Plain et al. (2009) 1.0 (late summer)
Beech, oak 13C (10 m tall trees) Dannoura et al. (2011) 0.2–1.2 (over the growing season)
Beech 13C (saplings) Barthel et al. (2011) 0.4 (control)–0.1 (drought)
Coniferous Spruce, pine 11C (seedlings) Thompson et al. (1979) 0.1
Larch, boreal 14C (saplings) Schneider and Schmitz (1989) 0.1–0.2 (over the growing season)
Pine (boreal) 13C (2 m tall trees) Högberg et al. (2008) 0.1 (late summer)
Pine (temperate) 13C (10 m tall trees) Dannoura et al. (2011) 0.1–0.2 (over the year)

observed after a few days or less and detected almost at the acting, and that tracer disappearance is only governed by
same time in roots and in the microbial biomass (Högberg tracer efflux, the mean residence time (τ, the average time
et al. 2008, Ruehr et al. 2009, Epron et al. 2011). The likely which C atoms reside in a reservoir) and the half-life (t1/2, the
mechanism for this to happen is that labelled C is rapidly cap- time required to exchange 50% of the C atoms in a reservoir)
tured by symbionts and/or exuded by the mycorrhizal roots can be calculated (Derrien and Amelung 2011) as
and taken up by soil microorganisms.
Among soil biota, fungi (especially ectomycorrhizal fungi) 1 ln(2)
τ= and t1/2 =
and Collembola were rapidly and strongly labelled, compared k k
with bacteria, Acari and Enchytraeidae (Esperschütz et al.
2009b, Heinonsalo et al. 2010, Högberg et al. 2010). A fast Mean residence time and half-life are both important attri-
transfer of labelled C was observed in litter patches, where butes of the C dynamics. It is therefore important that the
growing mycorrhizal mycelia were an active C sink for pine caveats for their estimation are explicitly stated. For most stud-
seedlings (Leake et al. 2001). Labelled C transferred to the ies, the assumption that the system is in steady state and fol-
mycorrhizal network can further be shared among competing lows first-order kinetics is only a good approximation. Moreover,
seedlings (Simard et al. 1997a, 1997b, Teste et al. 2010). the condition of a pool which only loses tracer is not realized for
The challenge now is to quantify the importance of this flux belowground compartments (roots, microbial biomass, soil), or
for the metabolism of both plants and microorganisms (Grimoldi for structural compounds, which typically exhibit bell-shaped
et al. 2006). Most probably, the transfer of labelled C from the patterns of label recovery (Endrulat et al. 2010, Epron et al.
plant is most relevant for microorganisms living in close prox- 2011). Furthermore, analysing a single physical compartment
imity to root tips, and less important for those feeding on older does not assure that the tracer is present in a single kinetic
pools of soil organic matter. Most interestingly, drought stress pool. For instance, leaves contain (at least) three kinetically
decreased (even cancelled) C transfer to soil microorganisms contrasting pools: labile, transiently stored and structural com-
(Gorissen et al. 2004, Ruehr et al. 2009, Barthel et al. 2011). pounds (Figure 1). Therefore, it is important that fairly complete
To what extent this finding is due to the lower C assimilation records of tracer content are available (i.e., from the beginning
rate, to changes in C partitioning, or to changes in microorgan- of the pulse and for a chase period of adequate duration) so
ism activity is not yet clear. that the presence of slow pools can be detected and accounted
for when estimating mean residence time and half-life.
The residence time in tree and soil
The rate of decrease of the amount or the fraction of labelled C Differences in half-lives between tree compartments
recovered in a compartment or a flux after the peak can often The half-life of labelled C in the phloem sap is similar between
be described by an exponential function: deciduous broadleaved species and evergreen coniferous spe-
cies (Table 3). In pine, t1/2 values in winter are more than twice
C(t ) = C0 exp(− kt ) those estimated for the growing season. The half-life in phloem
sap is typically short, indicating its transport function from the
where t is the time after the peak, C0 the quantity or fraction of canopy to sink organs (Högberg et al. 2008, Dannoura et al.
labelled C at peak time, k is the rate constant of tracer loss and 2011, Warren et al. 2012). For similar reasons, the half-life of
C(t) the quantity or fraction of labelled C at time t (Figure 2). labelled C in the soil solution is short (Esperschütz et al. 2009a)
Assuming that the system is in a steady state (no change in and the labelled C is sometimes undetectable due to a rapid
size), that the process follows first-order kinetics (fluxes are absorption by the soil microorganisms (Epron et al. 2011). The
directly proportional to pool sizes), that only one kinetic pool is amount of labelled C recovered in the mature foliage decreases

Table 3. Half-life of labelled C in mature leaves, phloem sap and respiratory substrate pools contributing to soil CO2 efflux (assuming either one or
two pools), as estimated from pulse-labelling experiments.
Species Methods References Half-life (days)
Phloem Pine (boreal) 13C (2 m tall trees) Högberg et al. (2008) 1.3 (late summer)
Pine (temperate) 13C (7 m tall trees) Warren et al. (2012) 4.7 (late summer)
Pine (temperate) 13C (10 m tall trees) Dannoura et al. (2011) 0.4–3.4 (over the year)
Beech 13C (10 m tall trees) Dannoura et al. (2011) 0.4–1.7 (over the growing season)
Oak 13C (10 m tall trees) Dannoura et al. (2011) 0.9–1.3 (over the growing season)
Leaves Pine (boreal) 13C (2 m tall trees) Högberg et al. (2008) 1.3 (late summer)
Pine (temperate) 13C (7 m tall trees) Warren et al. (2012) 0.5 (late summer)
Pine (temperate) 13C (10 m tall trees) Unpublished1 1.4–3.6 (over the year)
Beech 13C (10 m tall trees) Unpublished1 1.3 (over the growing season)
Beech 13C (saplings) Ruehr et al. (2009) 1.7 (control)–2.3 (drought),
Soil CO2 efflux—1 pool Poplar 14C (3 m tall trees) Horwath et al. (1994) 2.9–4.1 (over the growing season)

Spruce (boreal) 14C (<4 m tall trees) Carbone et al. 2007 10 (late summer)
Pine (boreal) 13C (2 m tall trees) Högberg et al. (2008) 1.5 (late summer)
Beech 13C (saplings) Barthel et al. (2011) 0.8 (control)–1.7 (drought)
Soil CO2 efflux—2 pools Shrubs (semi-arid) 14C (shrubs) Carbone and Trumbore 2007 2.6–3.1 (over the growing season)
Beech 13C (10 m tall trees) Epron et al. (2011) 4.8–13.1 (over the growing season)
Oak 13C (10 m tall trees) Epron et al. (2011) 6.0–7.7 (over the growing season)
Pine (temperate) 13C (10 m tall trees) Epron et al. (2011) 4.9–33.6 (over the year)
1Details about this experiment can be found in Dannoura et al. (2011).
rapidly after labelling, with a half-life of <2 days during the in mind that several differently labelled substrates can con-
growing season (Högberg et al. 2008, Warren et al. 2012, tribute to respiration depending on the time of the day. The
Table 3). The half-life of labelled C was higher in drought- overall pattern of tracer dynamics in respired CO2 is often
stressed beech saplings (Ruehr et al. 2009) or during winter in best approximated by using two (or more) exponential func-
pine (Table 3), due to lower export rates. tions. For beech, recovery patterns of labelled C in trunk and
Longer half-lives of labelled C are expected in growing sinks soil CO2 efflux have been well described with two exponential
where the labelled C is not only used to fuel respiration or to functions with half-lives of 3.5 to 5 days for the first exponen-
be exported, but is allocated to structural and storage com- tial function and of 16–18 days for the second (Plain et al.
pounds. Young, immature pine needles exhibited a different 2009). This pattern highlights the existence of a rapidly
pattern of recovery compared with mature needles with an cycling, metabolically active C pool and much slower cycling
increase in labelled C several days after labelling and a longer pools of stored C (Carbone and Trumbore 2007, Mortazavi
half-life. This pattern highlights the fact that these immature et al. 2009). These fast cycling C pools were found to con-
needles are strong C sinks and that they accumulate labelled C tribute to 71% of soil CO2 efflux in a pine plantation (Taneva
into structural biomass (unpublished). Labelled C in fine roots et al. 2006).
of fir had a half-life of several months independent of diameter Furthermore, estimates of the half-life of labelled C in the
class (Endrulat et al. 2010). In pine, labelled C in root tips whole plant–soil system before being respired are higher
exhibited shorter half-life than in fine roots (6 versus 15 when based on two exponential fits as compared with a single
months, Keel et al. 2012). A half-life of several years was even exponential fit (Table 3). Half-life of 1.5 days or 3–4 days was
found in fine roots in a mixed hardwood forest after an acci- reported for boreal pines (Högberg et al. 2008) and temper-
dental ecosystem labelling with 14CO2 (Joslin et al. 2006). ate poplars (Horwath et al. 1994), respectively, when a single
When the tracer is continuously or repeatedly supplied to the exponential decay function was fitted to the data, whereas it
tree, it can be traced over long time scales. This approach is ranged from 5 to 13 days for beech, oak and pine during the
more suitable than pulse-labelling for estimating the mean resi- growing season when a two-compartment model was used
dence time of C in long-lived pools of organic matter both in (Epron et al. 2011). Fitting a simple exponential function in
trees (e.g., in tree rings, Palacio et al. 2011) and in the soil such cases inevitably underestimates the half-life of C in these
(Andrews et al. 1999, Allen et al. 2000, Paterson et al. 2009, compartments. This highlights the need for more mechanistic
Esperschütz et al. 2009b). approaches to assess residence times in complex systems.
Half-life of labelled C in respiratory substrate pools Compartmental modelling of respiratory C pools

The decrease in labelled CO2 in the respiratory flux may be The availability of detailed time courses of both tracer content
used as a proxy for respiratory substrates. It should be kept and CO2 efflux rate allows the application of more refined

790 Epron et al.
mathematical analyses. For instance, compartmental analyses of the root system it is difficult to obtain representative sam-
have proven to be an effective and elegant tool to analyse the ples, (ii) because the tracer tends to concentrate in the finer,
system supplying C substrates to respiration in ryegrass easier-to-lose, root tips and (ii) because of difficulties with
(Lehmeier et al. 2008, 2010a, 2010b). In this approach, sorting live and dead fine roots (Ruehr et al. 2009, Endrulat
conceptual models including fast turnover, metabolically active et al. 2010). Scaling the amount of labelled C recovered in
pools and storage pools (Figure 4, adapted from Epron et al. respiration chambers to whole trees is also prone to large
2011) are translated into a set of differential equations that uncertainties. Trunk respiration can vary with diameter and
are fitted to the kinetics of labelled C to estimate the half-life position in the canopy (Ryan et al. 1996, Damesin et al. 2002).
of pools and their relative contribution to respiration. Also soil respiration is typically characterized by high spatial
Compartmental modelling was used to account for the mixing variability (Søe and Buchmann 2005, Ngao et al. 2012). Thus,
of stored C and newly assimilated C after pulse-labelling of C partitioning calculations that have been reported so far are
trees with 13C (Epron et al. 2011). However, the situation likely associated with large uncertainties, but nevertheless pro-

encountered by trees in the field under fluctuating climatic vide important qualitative information.
conditions is rather complex and the limited numbers of pools
(usually two or three) may not fully reflect physiological Partitioning the total amount of labelled C assimilated
processes. Because C partitioning refers to the fraction of labelled C in
one compartment, the total amount of labelled C assimilated by
the tree should be known. While it is possible to estimate the
Carbon partitioning
total amount of labelled C delivered to the tree (see above), it
For better understanding of processes underlying C allocation is less straightforward to quantify the amount that has been
dynamics it is important to disentangle the effects of phenol- actually assimilated (flux [1] in Figure 1).
ogy from those mediated by environmental factors, which both One option is to sum up the labelled C recovered in all com-
affect sink and source activities. Because trees are perennial partments. This can be simplified by collecting leaves immedi-
species, pulse-labelling experiments also provide insights into ately after labelling when most of the tracer has not yet
the contribution of old and current assimilates to growth and transported to other organs. The amount of labelled C in any
respiration in sink organs. compartment will thus be expressed relative to the initial
amount recovered in the foliage (Keel et al. 2007, Plain et al.
Scaling labelled contents to trees 2009, Endrulat et al. 2010). In order to account for the hetero-
The amount of labelled C recovered in one compartment is geneity of leaf photosynthetic activity in the crown, a stratified
obtained by multiplying the percent atom excess of the heavi- sampling strategy is needed. Typically, the crown can be virtu-
est isotope by the C content of the biomass of this compart- ally separated into bottom, middle and upper parts, for which
ment. Knowing the amount of labelled C allocated to fine roots foliage can be sampled and analysed separately. In the case of
is challenging: (i) because of the lateral and vertical extension evergreen species, the sampling design should also account
for the different needle cohorts. The proportion of foliage mass
belonging to each part of the crown should be known to
upscale the C isotope composition to the crown level. In addi-
tion, a good mixing of air in the labelling chamber is a prereq-
uisite to ensure that the heterogeneity in foliar C isotope
composition reflects the heterogeneity of leaf photosynthetic
activity. This approach provides a conservative estimate of the
amount of labelled C taken up by the crown because respira-
tory losses and export of carbohydrates take place already
during the labelling. Nonetheless, for herbaceous vegetation,
Lattanzi et al. (2012) estimated that for labelling periods
Figure 4. Typical compartmental model describing the kinetics of the shorter than 6 h the underestimation is <10% of assimilated C.
labelled C recovered in respiration (R). The labelled C enters a meta- This is because the tracer mixes rapidly with several metabolic
bolic pool (M) by photosynthesis (P), and a quantity of label (Q) is
pools, and consequently is not respired or allocated immedi-
transferred and back-transferred to a storage pool (S) at rates follow-
ing first-order kinetics (with rate constants kM− S and kS− M respec- ately after its assimilation.
tively). The label in the metabolic pool fuels respiration at a rate which Another option to estimate the total amount of labelled C
also follows a first-order kinetics (kR). A lag, depending on the velocity assimilated by the tree is to calculate uptake from photosyn-
of the phloem sap and on the path length, can be introduced between
the uptake of the labelled C by the canopy and its arrival in the meta- thesis measured with a canopy chamber before and after
bolic pool in the sink organ (not included in the equation). labelling, or from using data from a nearby eddy flux tower.

This approach should account for the 13CO2/12CO2 mixing Keel 2006). Production of sporocarps by mycorrhizal fungi is
ratio during labelling and requires that that during labelling also an important C sink (flux [10] in Figure 1) at the end of
leakage is negligible. the growing season, which diverts recent assimilates in pine
(Högberg et al. 2010). In boreal species, there is strong evi-
Environmental controls of C allocation dence that C allocation shifts from aboveground to below-
The functional-balance concept predicts that C partitioning is ground from the beginning to the end of the growing season
tuned to maintain an optimal internal resource status, like a (Lippu 1994, Kagawa et al. 2006a, Högberg et al. 2010),
constant biomass C : N ratio (Reynolds and Chen 1996, probably related to the delayed summer warming of the soil
Franklin et al. 2012). As expected from this concept, the compared with the air.
amount of labelled C allocated belowground after a 1-year At the whole-tree level, C allocation patterns are highly
chase period following pulse-labelling was 60% lower in nitro- dynamic and change with the seasons (Figure 5), indicating
gen (N)-fertilized plots compared with unfertilized plots in a that priorities among sink organs for recent assimilates exist at

boreal pine forest (Högberg et al. 2010). In contrast, N fertil- this time scale. Seasonal shifts in C allocation may also be
ization did not affect labelled C partitioning in potted young expected in tropical species encountering a transition between
beech saplings (Dyckmans and Flessa 2001), but affected the wet and dry seasons. We are not aware of any labelling experi-
shift in C allocation to belowground compartments that is ments addressing this question so far.
observed in response to fumigation with elevated CO2 concen-
tration (Dyckmans and Flessa 2002). Elevated CO2 concentra- Seasonal changes in allocation to storage
tion increased the amount of labelled C recovered in the An important flux of labelled C into storage pools (flux [3] in
microbial biomass in trembling aspen, suggesting that exuda- Figure 1) typically occurs at the end of the growing season
tion was stimulated (Mikan et al. 2000). (Hansen et al. 1996, Lacointe et al. 2004, Kuptz et al. 2011). It
Ozone was found to favour C allocation belowground at the is still unclear as to whether C reserves are a passive pool or
expense of leaves in potted silver birch trees (Kasurinen et al. an actively regulated sink (Sala et al. 2012). For example, 80%
2012) and allocation to the fungal symbiont in mycorrhizal pine of the newly assimilated C was allocated to aboveground
seedlings (Andersen and Rygiewicz 1995). However, no clear growth in July (leaves and stems) while the same proportion
patterns were observed in 60-year-old European beech and was stored in the stump when coppiced chestnuts were
Norway spruce trees (Andersen et al. 2010). Allocation of pho- labelled in October (Mordacq et al. 1986). A similar seasonal
tosynthates to the stem respiration (flux [2w] in Figure 1) shift in C allocation was reported for young poplar trees
decreased in beech exposed to O3 while it increased in spruce (Horwath et al. 1994). In contrast, results obtained in the field
(Ritter et al. 2011). on 20-year-old beech showed that C allocation to soil CO2
efflux, and to respiration and biomass of fine roots and
Disentangling the influence of seasonal variations in microbes, peaked in summer (Epron et al. 2011) and thus did
weather and species phenology not compete with C allocation to storage that occurred at the
When interpreting seasonal patterns of C allocation it is impor- end of the summer (Kuptz et al. 2011). Seasonal shifts in C
tant to disentangle the effects of phenology, as primarily driven allocation belowground were not observed in spruce, sessile
by biological controls, and those directly caused by environ- oak and maritime pine (Epron et al. 2011, Kuptz et al. 2011).
mental factors. This is especially the case for seasonally vari- The discrepancy among species may reflect differences in the
able parameters like temperature and soil water content. contribution of stored carbohydrates that might buffer the sea-
Variations in sink strength are important drivers of the seasonal changes in the partitioning of newly fixed C in some tem-
sonal dynamics of C allocation patterns (Table 1, Lippu perate species (Epron et al. 2011, Figure 5).
1994). Bud break and sprouting (flux [4l] in Figure 1) in Besides seasonal changes in allocation to storage com-
spring divert an important fraction of C assimilated by older partments in perennial organs, transient storage also occurs
foliage in evergreen species. Indeed, acropetal transport of at a daily time scale, especially in the foliage (fluxes [3l and
labelled C (flux [9] in Figure 1) to the growing needles and 6l] in Figure 1). For example, it was shown that respiration of
shoots was reported for larch (Schneider and Schmitz 1989), beech leaves was supplied by a mixture of current and stored
and it was demonstrated that bud break and sprouting in carbohydrates (Nogués et al. 2006). In some species, the
spring was exclusively supplied by the recent photosynthates emission of biogenic volatile organic compounds (BVOC, flux
of previous year’s needles in pine (Hansen and Beck 1994, [7] in Figure 1) is an additional sink of recently assimilated C,
Lippu 1998). Fruiting is another seasonally fluctuating C sink. which is mostly active in summer when light and temperature
This was nicely exemplified in beech and hornbeam for which are high (Loreto et al. 1996). However, the fraction of BVOC
almost all the labelled C assimilated by the fruit-bearing emissions originating from recent photosynthates or from
branches was allocated to their infructescences (Hoch and stored C may vary among species (Ghirardo et al. 2010). In

792 Epron et al.
experiments were thus useful to disentangle the contribution

of new (flux [1] in Figure 1) versus old/stored C [6] to growth
[4] or to respiration [2]. While the contribution of recently
assimilated and stored C to trunk and coarse root respiration
was almost constant in spruce and was dominated by old car-
bohydrates, it changed seasonally in 60-year-old beech trees
(Kuptz et al. 2011). Trunk and coarse root respiration are domi-
nated by stored C when the C demand to sustain shoot growth
uses most of the recently assimilated C (Figure 5).
Daily oscillations of the isotope signal after pulse-labelling in
trunk CO2 efflux were ascribed to diurnal changes in the C
sources fuelling respiration, i.e., remobilization of labelled

starch during the night versus use of unlabelled newly assimi-
lated C during daytime (Plain et al. 2009, Barthel et al. 2011).
In trees, however, diurnal changes in C sources could be con-
founded by internal CO2 transfer if, for instance, CO2 produced
belowground affects the composition of trunk CO2 efflux.
Departures of root-produced CO2 in the transpiration stream
that is contributing to trunk CO2 efflux have indeed been
reported on several occasions (Teskey and McGuire 2007,
Aubrey and Teskey 2009, Grossiord et al. 2012).
Sprouting in deciduous broadleaved species relies more or
less on C that has been assimilated during the previous year
and stored overwinter, questioning the concept of C autonomy
of branches in spring (Lacointe et al. 2004, Keel et al. 2007).
The leaves of 3-year-old potted beech were composed of
50–70% newly assimilated C, highlighting a significant contri-
bution of stored C, especially for the initial synthesis of acid
detergent-fibre lignin (Dyckmans et al. 2002). This was also
Figure 5. Conceptual representation of seasonal changes in C alloca-
tion in temperate and boreal trees, accounting for the use of recently observed in a deciduous coniferous species (larch) for which
assimilated C versus stored C. Black arrows indicate fluxes of recently about half the C in new needles was derived from stored C
assimilated C, white arrows fluxes of C remobilized from storage and (Kagawa et al. 2006a), and the contrast between evergreen
grey arrows a mixture of both. The width of arrows indicates sink or
and deciduous coniferous species was demonstrated by com-
source strength. During spring, sprouting (foliage growth and foliage
respiration) is an important sink for new C in evergreen species paring larch and pine in a FACE system (von Felten et al.
(a) while storage remobilization largely contributes to sprouting in 2007). Labelled C that was stored over winter was also found
deciduous species (b). Stem growth, root growth, stem respiration and in new fine roots that grew in the spring (Endrulat et al. 2010).
root (and associated symbionts) respiration are mainly sustained by
storage remobilization. In summer, foliage growth almost stops in many Similarly, it was shown in a FACE experiment that fine root and
species. Stem and root growth and respiration are major sinks of C and mycorrhiza growth in scrub oaks were partly supported by C
are either (c) mainly fuelled by new C (e.g., beech) or (d) by a mixture stored in rhizomes (Langley et al. 2002), in agreement with a
of old and new C (e.g., spruce, oak). In autumn, storage build-up is the
major sink of recent photosynthate (e). During winter (f), photosynthe- large-scale 14C labelling showing that stored C accounted for
sis and growth are either low (evergreen species) or zero (deciduous 55% of new root growth in a mixed hardwood forest (Gaudinski
species), and respiration is fully sustained by storage remobilization. et al. 2009).
Wood formation results also from a mixture of recent and
view of the diurnal changes of the use of new C for storage stored C in a proportion that varies among species, the rela-
in starch, respiration or exportation, future studies could tive contribution of stored C being greater in ring-porous than
address the question of how the timing of labelling may influ- in diffuse-porous species (Keel et al. 2006, Palacio et al.
ence the fate of C. 2011). This difference is thought to be related to the differ-
ence in their ability to restore hydraulic conductivity after
Contribution of stored C to growth and respiration winter embolism (Barbaroux and Bréda 2002). A different
When a pulse of labelled C is provided to a tree, the recently contribution of old and current assimilates to early wood and
assimilated C has an isotopic composition that differs from the late wood formation was found in larch, early wood being
isotope signature of the unlabelled stored C. Pulse-labelling more dependent on stored C than late wood (Kagawa et al.

2006b). Diametric growth in summer might also compete impact on C allocation within the tree and on soil biota, and
with shoot growth in polycyclic species like birch, relying thus can strongly affect the rate of C translocation.
more on stored C than shoot elongation (Palacio et al. 2011). This review has identified several open research questions
Scots pine seedlings used preferentially stored C to synthe- which should be of high priority on the research agenda. (i)
size secondary metabolites in stems when inoculated with a Source–sink relations: shading tree crowns to manipulate
bark-beetle-associated fungus (Guérard et al. 2007). There source activity (Warren et al. 2012) and varying fruit load
are large sets of intra-annual records of the natural abun- (Palmer 1992) or tapping intensity (Chantuma et al. 2009) to
dance of C isotope composition in tree rings (Barbour et al. manipulate sink demand, combined with labelling experiments,
2002, Helle and Schleser 2004, Offermann et al. 2011), will offer promising perspectives to better understand biologi-
which allow qualitative assessment of the contribution of cal controls on C allocation to storage organs and potentially
stored starch-derived versus recently assimilated C during differentiate between sink- versus source-driven controls. (ii) C
the growing season. Combining 13C labelling with such natu- allocation to secondary metabolism: insect and pathogen

ral abundance approaches might give additional insights into attacks play a major role in drought and heat wave-induced
the seasonal variation of storage and remobilization and C tree mortality (McDowell et al. 2011); however, patterns of C
resource use in trees. allocation to defence compounds (flux [5] in Figure 1) have not
yet been considered. (iii) Response of C allocation to environ-
Sources of ‘new’ versus ‘old’ C mental change: combining labelling and ecosystem manipula-
In most of the above-mentioned studies, labelled C is referred tion approaches like fertilization, water exclusion or warming
to as ‘new C’ and interpreted as derived from ‘current photo- experiments will be useful for disentangling the effects of envi-
synthesis’ or ‘recent assimilates’ while unlabelled C is referred ronmental factors versus phenology, both being confounded
as ‘old C’ and interpreted as derived from ‘stores’ or ‘reserves’. when addressing seasonal variations of allocation patterns.
This is convenient but not strictly valid, as there is no invariant (iv) Seasonality versus phenology effects in and across different
relationship between labelled and unlabelled substrates and biomes: there is a particular need for pulse-labelling experi-
the sources supplying a sink (Farrar and Gunn 1998, Lattanzi ments on tropical trees. They play a major role in the global C
et al. 2005). After its assimilation, the tracer starts to mix cycle (Malhi 2010) and have a distinct phenology compared
within the plant metabolic network, moving through several with temperate and boreal tree species. (v) C–N interactions:
physical and biochemical compartments. Thus, while part of dual C and N labelling experiments will not only deepen our
the tracer goes directly to the sink (flux [8] in Figure 1), part of understanding of the coupling of C and N dynamics in trees
it is incorporated in transient stores (e.g., chloroplastic starch, (Millard and Grelet 2010), but will also improve the parameter-
flux [3l]), amino acids, proteins, and longer-term stores [3w ization of coupled biogeochemical cycles in forest C balance
and 3r]. When mobilized [6], these pools supply labelled C. models (Chapin et al. 2009).
This also holds for unlabelled C, which is not completely New tools like compound-specific isotope analysis (CSIA;
derived from stores but is also present in labile pools, which Meier-Augenstein 1999, Godin et al. 2007), proton-transfer
are closely related to current photosynthesis. Therefore, infer- reaction mass spectrometry (PTR-MS; Ghirardo et al. 2010) or
ences based on a simplified definition of ‘new’ versus ‘old’ C nano-scale secondary ion mass spectrometry (nano-SIMS;
require caution and, ideally, need to be informed by additional Herrmann et al. 2007, Hatton et al. 2012), which have become
knowledge of the studied system. available only relatively recently, will permit tracing isotopes into
specific C compounds and will enable micro-scale localization of
labelled C in tree tissues and in the soil. This will facilitate the
Conclusion
study of a range of topics, including the within-tree translocation
Pulse-labelling experiments have provided unique and highly of C, the exudation of organic C compounds by the rhizosphere
valuable information on the transfer and allocation of C in trees. or their emission to the atmosphere. Besides pulse-labelling,
They have revealed a fast transfer of assimilated C from the long-term labelling of trees in the field will improve our knowl-
foliage to belowground, with species-specific velocities, which edge on residence times and turnover rates of C in long-lived
are likely due to phloem anatomy. Experiments have also high- pools of organic matter in perennial tree organs and in soils,
lighted the seasonality of allocation patterns as affected by both crucial for C sequestration (Körner 2006).
environmental and endogenous controls on transfer rates and Finally it should be noted that much of the available knowl-
activities of sink organs. Growth and respiration rely both on edge has been gained in pulse-labelling experiments using
recent photosynthates and on stored C, in a proportion that relatively small trees, while in situ whole-tree labelling experi-
depends on the season and that varies among species. ments are still scarce although urgently needed to improve our
Environmental constraints (ozone, drought) and differences in understanding of environmental and physiological controls on
resource availability (soil water, soil fertility and light) have an C allocation.

794 Epron et al.
Acknowledgments Barbour, M.M., A.S. Walcroft and G.D. Farquhar. 2002. Seasonal varia-
tion in δ13C and δ18O of cellulose from growth rings of Pinus radiata.
Plant Cell Environ. 25:1483–1499.
We are particularly grateful to M.G. Ryan for helpful advice Barthel, M., A. Hammerle, P. Sturm, T. Baur, L. Gentsch and A. Knohl.
when revising the manuscript. 2011. The diel imprint of leaf metabolism on the δ13C signal of soil
respiration under control and drought conditions. New Phytol.
192:925–938.
Conflict of interest Bélanger, G., F. Gastal and F.R. Warembourg. 1992. The effects of
nitrogen fertilization and the growing season on carbon partitioning
None declared. in a sward of tall fescue (Festuca arundinacea Schreb). Ann. Bot.
70:239–244.
Bowling, D.R., N.G. McDowell, B.J. Bond, B.E. Law and J.R. Ehleringer.
Funding 2002. 13C content of ecosystem respiration is linked to precipitation
and vapor pressure deficit. Oecologia 131:113–124.
This idea for this review emerged during a workshop on Brandes, E., N. Kodama, K. Whittaker, C. Weston, H. Rennenberg, C.

‘Analysing post labelling experiments’ held in Nancy, France, in Keitel, M.A. Adams and A. Gessler. 2006. Short-term variation in the
March 2010, which was supported by the COST Action isotopic composition of organic matter allocated from the leaves to the
stem of Pinus sylvestris: effects of photosynthetic and postphotosyn-
ES0806 SIBAE (Stable Isotopes in Biosphere-Atmosphere-
thetic carbon isotope fractionation. Glob. Change Biol. 12:1922–1939.
Earth System Research, http://www.sibae.ethz.ch/cost-sibae/). Brüggemann, N., A. Gessler, Z. et al. 2011. Carbon allocation and car-
M.D. came to Nancy within the frame of a bilateral JSPS-INRA bon isotope fluxes in the plant-soil-atmosphere continuum: a review.
project. Research by F.A.L. has been funded by the DFG Biogeosciences 8:3457–3489.
Calvin, M. 1961. The path of carbon in photosynthesis (Nobel Prize
(LA:2390/1-1). Research of J.P. was funded by Academy of
Lecture). Ernest Orlando Lawrence Berkeley National Laboratory,
Finland project number 218094. N.B. and M.B. acknowledge University of California Radiation Laboratory-Berkeley, United States
the support from the European Commission for the project Department of Energy, Berkeley, CA, 47 p.
Carbo-Extreme (FP7-ENV-2008-1-226701). Cannell, G.R. and R.C. Dewar. 1994. Carbon allocation in trees: a
review of concepts for modelling. Adv. Ecol. Res. 25:59–104.
Carbone, M.S. and S.E. Trumbore. 2007. Contribution of new photo-
synthetic assimilates to respiration by perennial grasses and shrubs:
References residence times and allocation patterns. New Phytol. 176:124–135.
Allen, A.S., J.A. Andrews, A.C. Finzi, R. Matamala, D.D. Richter and W.H. Carbone, M.S., C.I. Czimczik, K.E. McDuffee and S.E. Trumbore. 2007.
Schlesinger. 2000. Effects of free-air CO2 enrichment (FACE) on Allocation and residence time of photosynthetic products in a boreal
belowground processes in a Pinus taeda forest. Ecol. Appl. forest using a low-level 14C pulse-chase labeling technique. Glob.
10:437–448. Change Biol. 13:466–477.
Amelung, W., S. Brodowski, A. Sandhage-Hofmann and R. Bol. 2008. Chantuma, P., A. Lacointe, P. Kasemsap, S. Thanisawanyangkura, E.
Combining biomarker with stable isotope analyses for assessing the Gohet, A. Clement, A. Guilliot, T. Ameglio and P. Thaler. 2009.
transformation and turnover of soil organic matter. In Advances in Carbohydrate storage in wood and bark of rubber trees submitted to
Agronomy. Ed. D. Sparks. Academic Press, Burlington, pp different level of C demand induced by latex tapping. Tree Physiol.
155–250. 29:1021–1031.
Andersen, C.P. and P.T. Rygiewicz. 1995. Allocation of carbon in mycor- Chapin, I., F. Stuart, J. McFarland, A.D. McGuire, E.S. Euskirchen, R.W.
rhizal Pinus ponderosa seedlings exposed to ozone. New Phytol. Ruess and K. Kielland. 2009. The changing global carbon cycle:
131:471–480. linking plant-soil carbon dynamics to global consequences. J. Ecol.
Andersen, C.P., W. Ritter, J. Gregg, R. Matyssek and T.E.E. Grams. 97:840–850.
2010. Below-ground carbon allocation in mature beech and spruce Damesin, C., E. Ceschia, N. Le Goff, J.-M. Ottorini and E. Dufrêne. 2002.
trees following long-term, experimentally enhanced O3 exposure in Stem and branch respiration of beech: from tree measurements to
southern Germany. Environ. Pollut. 158:2604–2609. estimations at the stand level. New Phytol. 153:159–172.
Andrews, J.A., K.G. Harrison, R. Matamala and W.H. Schlesinger. 1999. Dannoura, M., P. Maillard, C. et al. 2011. In situ assessment of the
Separation of root respiration from total soil respiration using car- velocity of carbon transfer by tracing 13C in trunk CO2 efflux after
bon-13 labelling during free-air carbon dioxide enrichment (FACE). pulse labelling: variations among tree species and season. New
Soil Sci. Soc. Am. J. 63:1429–1435. Phytol. 190:181–192.
Aubrey, D.P. and R.O. Teskey. 2009. Root-derived CO2 efflux via xylem Dawson, T.E., S. Mambelli, A.H. Plamboeck, P.H. Templer and K.P. Tu.
stream rivals soil CO2 efflux. New Phytol. 184:35–40. 2002. Stable isotopes in plant ecology. Annu. Rev. Ecol. Syst.
Bader, M., E. Hiltbrunner and C. Körner. 2009. Fine root responses of 33:507–559.
mature deciduous forest trees to free air carbon dioxide enrichment Denef, K., D. Roobroeck, M.C.W. Manimel Wadu, P. Lootens and P.
(FACE). Funct. Ecol. 23:913–921. Boeckx. 2009. Microbial community composition and rhizodeposit-
Bahn, M., M. Schmitt, R. Siegwolf, A. Richter and N. Brüggemann. carbon assimilation in differently managed temperate grassland
2009. Does photosynthesis affect grassland soil-respired CO2 and soils. Soil Biol. Biochem. 41:144–153.
its carbon isotope composition on a diurnal timescale? New Phytol. Derrien, D. and W. Amelung. 2011. Computing the mean residence
182:451–460. time of soil carbon fractions using stable isotopes: impacts of the
Barbaroux, C. and N. Bréda. 2002. Contrasting distribution and sea- model framework. Eur. J. Soil Sci. 62:237–252.
sonal dynamics of carbohydrate reserves in stem wood of adult Dieleman, W.I.J., S. Luyssaert, A. Rey, et al. 2010. Soil [N] modulates
ring-porous sessile oak and diffuse-porous beech trees. Tree soil C cycling in CO2-fumigated tree stands: a meta-analysis. Plant
Physiol. 22:1201–1210. Cell Environ. 33:2001–2011.

Dijkstra, F.A. and W. Cheng. 2007. Interactions between soil and tree Giardina, C.P. and M.G. Ryan. 2002. Total belowground carbon alloca-
roots accelerate long-term soil carbon decomposition. Ecol. Lett. tion in a fast-growing Eucalyptus plantation estimated using a car-
10:1046–1053. bon balance approach. Ecosystems 5:487–499.
Dyckmans, J. and H. Flessa. 2001. Influence of tree internal N status Godin, J.-P., L.-B. Fay and G. Hopfgartner. 2007. Liquid chromatogra-
on uptake and translocation of C and N in beech: a dual 13C and 15N phy combined with mass spectrometry for C-13 isotopic analysis in
labeling approach. Tree Physiol. 21:395–401. life science research. Mass Spectrom. Rev. 26:751–774.
Dyckmans, J. and H. Flessa. 2002. Influence of tree internal nitrogen Gorissen, A., A. Tietema, N.N. Joosten, M. Estiarte, J. Peñuelas, A.
reserves on the response of beech (Fagus sylvatica) trees to elevated Sowerby, B.A. Emmett and C. Beier. 2004. Climate change affects
atmospheric carbon dioxide concentration. Tree Physiol. 22:41–49. carbon allocation to the soil in shrublands. Ecosystems
Dyckmans, J., H. Flessa, K. Brinkmann, C. Mai and A. Polle. 2002. 7:650–661.
Carbon and nitrogen dynamics in acid detergent fibre lignins of Grams, T.E.E., H. Werner, D. Kuptz, W. Ritter, F. Fleischmann, C.P.
beech (Fagus sylvatica L.) during the growth phase. Plant Cell Andersen and R. Matyssek. 2011. A free-air system for long-term
Environ. 25:469–478. stable carbon isotope labeling of adult forest trees. Trees
Ekblad, A. and P. Högberg. 2001. Natural abundance of 13C in CO2 25:187–198.
respired from forest soils reveals speed of link between tree photo- Gregory, P.J. and B.J. Atwell. 1991. The fate of carbon in pulse-labelled

synthesis and root respiration. Oecologia 127:305–308. crops of barley and wheat. Plant Soil. 136:205–213.
Endrulat, T., M. Saurer, N. Buchmann and I. Brunner. 2010. Incorporation Griffiths, R.I., M. Manefield, N. Ostle, N. McNamara, A.G. O’Donnell,
and remobilization of 13C within the fine-root systems of individual M.J. Bailey and A.S. Whiteley. 2004. 13CO2 pulse labelling of plants
Abies alba trees in a temperate coniferous stand. Tree Physiol. in tandem with stable isotope probing: methodological consider-
30:1515–1527. ations for examining microbial function in the rhizosphere. J.
Epron, D., J. Ngao, M. Dannoura, M.R. Bakker, et al. 2011. Seasonal Microbiol. Methods 58:119–129.
variations of belowground carbon transfer assessed by in situ 13CO2 Grimoldi, A.A., M. Kavanová, F.A. Lattanzi, R. Schäufele and H. Schnyder.
pulse labelling of trees. Biogeosciences 8:1153–1168. 2006. Arbuscular mycorrhizal colonization on carbon economy in
Esperschütz, J., F. Buegger, J.B. Winkler, J.C. Munch, M. Schloter and A. perennial ryegrass: quantification by 13CO2/12CO2 steady-state label-
Gattinger. 2009a. Microbial response to exudates in the rhizosphere ling and gas exchange. New Phytol. 172:544–553.
of young beech trees (Fagus sylvatica L.) after dormancy. Soil Biol. Grossiord, C., L. Mareschal and D. Epron. 2012. Transpiration alters
Biochem. 41:1976–1985. the contribution of autotrophic and heterotrophic components of soil
Esperschütz, J., A. Gattinger, F. Buegger, H. Lang, J. Munch, M. Schloter CO2 efflux. New Phytol. 194:647–653.
and J. Winkler. 2009b. A continuous labelling approach to recover Guérard, N., P. Maillard, C. Bréchet, F. Lieutier and E. Dreyer. 2007. Do
photosynthetically fixed carbon in plant tissue and rhizosphere trees use reserve or newly assimilated carbon for their defense
organisms of young beech trees (Fagus sylvatica L.) using 13C reactions? A 13C labeling approach with young Scots pines inocu-
depleted CO2. Plant Soil. 323:21–29. lated with a bark-beetle-associated fungus (Ophiostoma brunneo
Farrar, J.F. and S. Gunn. 1998. Allocation: allometry, acclimation-and ciliatum). Ann. For. Sci. 64:601–608.
alchemy? In Inherent Variation in Plant Growth. Physiological Hansen, J. and E. Beck. 1990. The fate and path of assimilation prod-
Mechanisms and Ecological Consequences. Eds. H. Lambers, H. ucts in the stem of 8-year-old Scots pine (Pinus sylvestris L.) trees.
Poorter and M. Van Vuuren. Backhuys Publishers, Leiden, pp Trees Struct. Funct. 4:16–21.
183–198. Hansen, J. and E. Beck. 1994. Seasonal changes in the utilization and
Ford, C.R., N. Wurzburger, R.L. Hendrick and R.O. Teskey. 2007. Soil DIC turnover of assimilation products in 8-year-old Scots pine (Pinus syl-
uptake and fixation in Pinus taeda seedlings and its C contribution to vestris L.) trees. Trees Struct. Funct. 8:172–182.
plant tissues and ectomycorrhizal fungi. Tree Physiol. 27:375–383. Hansen, J., G. Vogg and E. Beck. 1996. Assimilation, allocation and
Franklin, O., J. Johansson, R.C. Dewar, U. Dieckmann, R.E. McMurtrie, utilization of carbon by 3-year-old Scots pine (Pinus sylvestris) trees
A. Brännström and R. Dybzinski. 2012. Modeling carbon allocation during winter and early spring. Trees Struct. Funct. 11:83–90.
in trees: a search for principles. Tree Physiol. 32:648–666. Hansen, P. 1970. 14C-studies on apple trees. VI. The influence of the
Gamnitzer, U., R. Schäufele and H. Schnyder. 2009. Observing 13C fruit on the photosynthesis of the leaves, and the relative photosyn-
labelling kinetics in CO2 respired by a temperate grassland ecosys- thetic yields of fruits and leaves. Physiol. Plant. 23:805–810.
tem. New Phytol. 184:376–386. Hatton, P., L. Remusat, B. Zeller and D. D. 2012. A multi-scale approach
Gaudinski, J.B., M.S. Torn, W.J. Riley, C. Swanston, S.E. Trumbore, J.D. to determine accurate elemental and isotopic ratios by nano-scale
Joslin, H. Majdi, T.E. Dawson and P.J. Hanson. 2009. Use of stored secondary ion mass spectrometry imaging. Rapid Commun. Mass
carbon reserves in growth of temperate tree roots and leaf buds: Spectrom. in press.
analyses using radiocarbon measurements and modeling. Glob. Heinonsalo, J., J. Pumpanen, T. Rasilo, K.-R. Hurme and H. Ilvesniemi.
Change Biol. 15:992–1014. 2010. Carbon partitioning in ectomycorrhizal Scots pine seedlings.
Gessler, A., H. Rennenberg and C. Keitel. 2004. Stable isotope compo- Soil Biol. Biochem. 42:1614–1623.
sition of organic compounds transported in the phloem of European Helle, G. and G. Schleser. 2004. Beyond CO2 fixation by Rubisco—an
beech—evaluation of different methods of phloem sap collection interpretation of 13C/12C variations in tree rings from novel intra-
and assessment of gradients in carbon isotope composition during seasonal studies on broad-leaf trees. Plant Cell Environ.
leaf-to-stem transport. Plant Biol. 6:1–10. 27:367–380.
Gessler, A., G. Tcherkez, O. Karyanto, C. Keitel, J.P. Ferrio, J. Ghashghaie, Herrmann, A.M., K. Ritz, N. Nunan, P.L. Clode, J. Pett-Ridge, M.R.
J. Kreuzwieser and G.D. Farquhar. 2009. On the metabolic origin of Kilburn, D.V. Murphy, A.G. O’Donnell and E.A. Stockdale. 2007.
the carbon isotope composition of CO2 evolved from darkened light- Nano-scale secondary ion mass spectrometry—A new analytical
acclimated leaves in Ricinus communis. New Phytol. 181:374–386. tool in biogeochemistry and soil ecology: a review article. Soil Biol.
Ghirardo, A., K. Koch, R. Taipale, I. Zimmer, J.-P. Schnitzler and J. Rinne. Biochem. 39:1835–1850.
2010. Determination of de novo and pool emissions of terpenes Hoch, G. and S.G. Keel. 2006. 13C labelling reveals different contribu-
from four common boreal/alpine trees by 13CO2 labelling and tions of photoassimilates from infructescences for fruiting in two
PTR-MS analysis. Plant Cell Environ. 33:781–792. temperate forest tree species. Plant Biol. 8:606–614.

796 Epron et al.
Högberg, M.N., M.J.I. Briones, S.G. et al. 2010. Quantification of effects newly assimilated organic carbon to respired carbon dioxide.
of season and nitrogen supply on tree below-ground carbon transfer Oecologia 156:737–750.
to ectomycorrhizal fungi and other soil organisms in a boreal pine Körner, C. 2006. Plant CO2 responses: an issue of definition, time and
forest. New Phytol. 187:485–493. resource supply. New Phytol. 172:393–411.
Högberg, P. and D.J. Read. 2006. Towards a more plant physiological Kuptz, D., F. Fleischmann, R. Matyssek and T.E.E. Grams. 2011.
perspective on soil ecology. Trends Ecol. Evol. 21:549–554. Seasonal patterns of carbon allocation to respiratory pools in 60-yr-
Högberg, P., M.N. Högberg, S.G. et al. 2008. High temporal resolution old deciduous (Fagus sylvatica) and evergreen (Picea abies) trees
tracing of photosynthate carbon from the tree canopy to forest soil assessed via whole-tree stable carbon isotope labeling. New Phytol.
microorganisms. New Phytol. 177:220–228. 191:160–172.
Horwath, W.R., K.S. Pregitzer and E.A. Paul. 1994. 14C allocation in Kuzyakov, Y. and O. Gavrichkova. 2010. Time lag between photosyn-
tree-soil systems. Tree Physiol. 14:1163–1176. thesis and CO2 efflux from soil: A review of mechanisms and con-
Jahnke, S., U. Schlesinger, G.B. Feige and E.J. Knust. 1998. Transport trols. Glob. Change Biol. 16:3386–3406.
of photoassimilates in young trees of Fraxinus and Sorbus: measure- Lacointe, A., E. Deleens, T. Ameglio, B. Saint-Joanis, C. Lelarge, M.
ment of translocation in vivo. Bot. Acta 111:307–315. Vandame, G.C. Song and F.A. Daudet. 2004. Testing the branch
Jahnke, S., M.I. Menzel, D. van Dusschoten, G.W. et al. 2009. Combined autonomy theory: a 13C/14C double-labelling experiment on differen-

MRI-PET dissects dynamic changes in plant structures and func- tially shaded branches. Plant Cell Environ. 27:1159–1168.
tions. Plant J. 59:634–644. Langley, J.A., B.G. Drake and B.A. Hungate. 2002. Extensive below-
Jensen, K.H., J. Liesche, T. Bohr and A. Schulz. 2012. Universality of ground carbon storage supports roots and mycorrhizae in regener-
phloem transport in seed plants. Plant Cell Environ. 35:1065–1076. ating scrub oaks. Oecologia 131:542–548.
Johnson, D., J.R. Leake, N. Ostle, P. Ineson and D.J. Read. 2002. In situ Lattanzi, F.A., H. Schnyder and B. Thornton. 2005. The sources of
13CO pulse-labelling of upland grassland demonstrates a rapid
2 carbon and nitrogen supplying leaf growth. Assessment of the role
pathway of carbon flux from arbuscular mycorrhizal mycelia to the of stores with compartmental models. Plant Physiol. 137:383–395.
soil. New Phytol. 153:327–334. Lattanzi, F.A., G.D. Berone, W. Feneis and H. Schnyder. 2012.
Joslin, J.D., J.B. Gaudinski, M.S. Tom, W.J. Riley and P.J. Hanson. 2006. 13C-labeling shows the effect of hierarchy on the carbon gain of
Fine-root turnover patterns and their relationship to root diameter individuals and functional groups in dense field stands Ecology. doi:
and soil depth in a 14C-labeled hardwood forest. New Phytol. 10.1890/11–1166.1.
172:523–535. Leake, J.R., D.P. Donnelly, E.M. Saunders, L. Boddy and D.J. Read.
Kagawa, A., A. Sugimoto, K. Yamashita and H. Abe. 2005. Temporal 2001. Rates and quantities of carbon flux to ectomycorrhizal myce-
photosynthetic carbon isotope signatures revealed in a tree ring lium following 14C pulse labeling of Pinus sylvestris seedlings: effects
through 13CO2 pulse-labelling. Plant Cell Environ. 28:906–915. of litter patches and interaction with a wood-decomposer fungus.
Kagawa, A., A. Sugimoto and T.C. Maximov. 2006a. Seasonal course Tree Physiol. 21:71–82.
of translocation, storage and remobilization of 13C pulse-labeled Lehmeier, C.A., F.A. Lattanzi, R. Schaufele, M. Wild and H. Schnyder.
photoassimilate in naturally growing Larix gmelinii saplings. New 2008. Root and shoot respiration of perennial ryegrass are supplied
Phytol. 171:793–804. by the same substrate pools: assessment by dynamic 13C labeling
Kagawa, A., A. Sugimoto and T. Maximox. 2006b. 13CO2 pulse-label- and compartmental analysis of tracer kinetics. Plant Physiol.
ling of photoassimilates reveals carbon allocation within and 148:1148–1158.
between tree rings. Plant Cell Environ. 29:1571–1584. Lehmeier, C.A., F.A. Lattanzi, U. Gamnitzer, R. Schäufele and H.
Kasurinen, A., C. Biasi, T. Holopainen, M. Rousi, M. Mäenpää and E. Schnyder. 2010a. Day-length effects on carbon stores for respira-
Oksanen. 2012. Interactive effects of elevated ozone and tempera- tion of perennial ryegrass. New Phytol. 188:719–725.
ture on carbon allocation of silver birch (Betula pendula) genotypes Lehmeier, C.A., F.A. Lattanzi, R. Schäufele and H. Schnyder. 2010b.
in an open-air field exposure. Tree Physiol. 32:737–751. Nitrogen deficiency increases the residence time of respiratory car-
Keel, S.G. and C. Schädel. 2010. Expanding leaves of mature decidu- bon in the respiratory substrate supply system of perennial rye-
ous forest trees rapidly become autotrophic. Tree Physiol. grass. Plant Cell Environ. 33:76–87.
30:1253–1259. Lippu, J. 1994. Patterns of dry matter partitioning and 14C-photosynthate
Keel, S.G., R.T.W. Siegwolf and C. Korner. 2006. Canopy CO2 enrich- allocation in 1.5-year-old Scots pine seedlings. Silva Fenn.
ment permits tracing the fate of recently assimilated carbon in a 28:145–153.
mature deciduous forest. New Phytol. 172:319–329. Lippu, J. 1998. Redistribution of 14C-labelled reserve carbon in Pinus
Keel, S.G., R.T.W. Siegwolf, M. Jaggi and C. Körner. 2007. Rapid mixing sylvestris seedlings during shoot elongation. Silva Fenn. 32:3–10.
between old and new C pools in the canopy of mature forest trees. Litton, C.M., J.W. Raich and M.G. Ryan. 2007. Carbon allocation in for-
Plant Cell Environ. 30:963–972. est ecosystems. Glob. Change Biol. 13:2089–2109.
Keel, S.G., C.D. Campbell, M.N. Högberg, A. Richter, B. Wild, X. Zhou, V. Loreto, F., P.Ciccioli, A. Cecinato, E. Brancaleoni, M. Frattoni, C. Fabozzi
Hurry, S. Linder, T. Näsholm and P. Högberg. 2012. Allocation of car- and D. Tricoli. 1996. Evidence of the photosynthetic origin of mono-
bon to fine root compounds and their residence times in a boreal for- terpenes emitted by Quercus ilex L. leaves by 13C labeling. Plant
est depend on root size class and season. New Phytol. 194:972–981. Physiol. 110:1317–1322.
Keitel, C., A. Matzarakis, H. Rennenberg and A. Gessler. 2006. Carbon Maillard, P., E. Deléens, F.A. Daudet, A. Lacointe and J.S. Frossard.
isotopic composition and oxygen isotopic enrichment in phloem and 1994. Carbon economy in walnut seedlings during the acquisition of
total leaf organic matter of European beech (Fagus sylvatica L.) autotrophy studied by long-term labelling with 13CO2. Physiol. Plant.
along a climate gradient. Plant Cell Environ. 29:1492–1507. 91:359–368.
Knohl, A., R.A. Werner, W.A. Brand and N. Buchmann. 2005. Short- Malhi, Y. 2010. The carbon balance of tropical forest regions, 1990–
term variations in δ13C of ecosystem respiration reveals link between 2005. Curr. Opin. Environ. Sustainability 2:237–244.
assimilation and respiration in a deciduous forest. Oecologia Marron, N., C. Plain, B. Longdoz and D. Epron. 2009. Seasonal and
142:70–82. daily time course of the 13C composition in soil CO2 efflux recorded
Kodama, N., R. Barnard, Y. Salmon, et al. 2008. Temporal dynamics of with a tunable diode laser spectrophotometer (TDLS). Plant Soil.
the carbon isotope composition in a Pinus sylvestris stand: from 318:137–151.

Meier-Augenstein, W. 1999. Applied gas chromatography coupled to Palmer, J.W. 1992. Effects of varying crop load on photosynthesis, dry
isotope ratio mass spectrometry. J. Chromatogr. A 842:351–371. matter production and partitioning of Crispin/M.27 apple trees. Tree
McDowell, N.G., D.R. Bowling, B.J. Bond, J. Irvine, B.E. Law, P. Anthoni Physiol. 11:19–33.
and J.R. Ehleringer. 2004. Response of the carbon isotopic content Palmroth, S., R. Oren, H.R. McCarthy, K.H. Johnsen, A.C. Finzi, J.R.
of ecosystem, leaf, and soil respiration to meteorological and physi- Butnor, M.G. Ryan and W.H. Schlesinger. 2006. Aboveground sink
ological driving factors in a Pinus ponderosa ecosystem. Glob. strength in forests controls the allocation of carbon below ground
Biogeochem. Cycles 18:1013. and its [CO2]-induced enhancement. Proc. Natl Acad. Sci. USA
McDowell, N.G., D.J. Beerling, D.D. Breshears, R.A. Fisher, K.F. Raffa 103:19362–19367.
and M. Stitt. 2011. The interdependence of mechanisms underlying Paterson, E., A.J. Midwood and P. Millard. 2009. Through the eye of the
climate-driven vegetation mortality. Trends Ecol. Evol. 26:523–532. needle: a review of isotope approaches to quantify microbial pro-
McLaughlin, S.B., R.K. McConathy and B. Beste. 1979. Seasonal cesses mediating soil carbon balance. New Phytol. 184:19–33.
changes in within-canopy allocation of 14C-photosynthate by white Plain, C., D. Gérant, P. Maillard, M. Dannoura, Y. Dong, B. Zeller, P. Priault,
oak. For. Sci. 25:361–370. F. Parent and D. Epron. 2009. Tracing of recently assimilated carbon
Mencuccini, M. and T. Hölttä. 2010. The significance of phloem trans- in respiration at high temporal resolution in the field with a tuneable
port for the speed with which canopy photosynthesis and below- diode laser absorption spectrometer after in situ 13CO2 pulse labelling

ground respiration are linked. New Phytol. 185:189–203. of 20-year-old beech trees. Tree Physiol. 29:1433–1447.
Mikan, C.J., D.R. Zak, M.E. Kubiske and K.S. Pregitzer. 2000. Combined Poorter, H., K.J. Niklas, P.B. Reich, J. Oleksyn, P. Poot and L. Mommer.
effects of atmospheric CO2 and N availability on the belowground 2012. Biomass allocation to leaves, stems and roots: meta-analyses
carbon and nitrogen dynamics of aspen mesocosms. Oecologia of interspecific variation and environmental control. New Phytol.
124:432–445. 193:30–50.
Millard, P. and G. Grelet. 2010. Nitrogen storage and remobilization by Powers, E.M. and J.D. Marshall. 2011. Pulse labeling of dissolved
trees: ecophysiological relevance in a changing world. Tree Physiol. 13C-carbonate into tree xylem: developing a new method to deter-
30:1083–1095. mine the fate of recently fixed photosynthate. Rapid Commun. Mass
Millard, P., M. Sommerkorn and G.-A. Grelet. 2007. Environmental Spectrom. 25:33–40.
change and carbon limitation in trees: a biochemical, ecophysiologi- Pumpanen, J., J. Heinonsalo, T. Rasilo, K.-R. Hurme and H. Ilvesniemi.
cal and ecosystem appraisal. New Phytol. 175:11–28. 2009. Carbon balance and allocation of assimilated CO2 in Scots
Miltner, A., H. Richnow, F. Kopinke and M. Kastner. 2004. Assimilation pine, Norway spruce, and silver birch seedlings determined with gas
of CO2 by soil microorganisms and transformation into soil organic exchange measurements and 14C pulse labelling. Trees Struct.
matter. Org. Geochem. 35:1015–1024. Funct. 23:611–621.
Mordacq, L., M. Mousseau and E. Deleens. 1986. A 13C method of Rangel-Castro, J.I., K. Killham, N. Ostle, G.W. Nicol, I.C. Anderson,
estimation of carbon allocation to roots in a young chestnut coppice. C.M. Scrimgeour, P. Ineson, A. Meharg and J.I. Prosser. 2005a.
Plant Cell Environ. 9:735–739. Stable isotope probing analysis of the influence of liming on root
Mortazavi, B., M.H. Conte, J.P. Chanton, M.C. Smith, J.C. Weber, J. exudate utilization by soil microorganisms. Environ. Microbiol.
Crumsey and J. Ghashghaie. 2009. Does the 13C of foliage-respired 7:828–838.
CO2 and biochemical pools reflect the 13C of recently assimilated Rangel-Castro, J.I., J.I. Prosser, N. Ostle, C.M. Scrimgeour, K. Killham
carbon? Plant Cell Environ. 32:1310–1323. and A.A. Meharg. 2005b. Flux and turnover of fixed carbon in soil
Navarro, M.N.V., C. Jourdan, T. Sileye, et al. 2008. Fruit development, microbial biomass of limed and unlimed plots of an upland grassland
not GPP, drives seasonal variation in NPP in a tropical palm planta- ecosystem. Environ. Microbiol. 7:544–552.
tion. Tree Physiol. 28:1661–1674. Reynolds, J.F. and J. Chen. 1996. Modelling whole-plant allocation in
Neufeld, J., M. Dumont, J. Vohra and J. Murrell. 2007. Methodological relation to carbon and nitrogen supply: coordination versus optimi-
considerations for the use of stable isotope probing in microbial zation: opinion. Plant Soil. 185:65–74.
ecology. Microb. Ecol. 53:435–442. Ritter, W., C.P. Andersen, R. Matyssek and T.E.E. Grams. 2011.
Ngao, J., D. Epron, N. Delpierre, N. Bréda, A. Granier and B. Longdoz. Carbon flux to woody tissues in a beech/spruce forest during
2012. Spatial variability of soil CO2 efflux linked to soil parameters summer and in response to chronic O3 exposure. Biogeosciences
and ecosystem characteristics in a temperate beech forest. Agric. 8:3127–3138.
For. Meteorol. 154–155:136–146. Roeb, G. and S.J. Britz. 1991. Short-term fluctuations in the transport
Nogués, S., C. Damesin, G. Tcherkez, F. Maunoury, G. Cornic and J. of assimilates to the ear of wheat measured with steady-state
Ghashghaie. 2006. 13C/12C isotope labelling to study leaf carbon 11C-CO -labelling of the flag leaf. J. Exp. Bot. 42:469–475.
2
respiration and allocation in twigs of field-grown beech trees. Rapid Rouhier, H., G. Billès, L. Billès and P. Bottner. 1996. Carbon fluxes in
Commun. Mass Spectrom. 20:219–226. the rhizosphere of sweet chestnut seedlings (Castanea sativa)
Offermann, C., J.P. Ferrio, J. Holst, R. Grote, R. Siegwolf and A. Gessler. grown under two atmospheric CO2 concentrations: 14C partitioning
2011. The long way down—are carbon and oxygen isotopes signals after pulse labelling. Plant Soil. 180:101–111.
in the tree ring uncoupled from canopy physiological processes? Ruehr, N.K., C.A. Offermann, A. Gessler, J.B. Winkler, J.P. Ferrio, N.
Tree Physiol. 31:1088–1102. Buchmann and R.L. Barnard. 2009. Drought effects on allocation of
Ostle, N., P. Ineson, D. Benham and D. Sleep. 2000. Carbon assimila- recent carbon: from beech leaves to soil CO2 efflux. New Phytol.
tion and turnover in grassland vegetation using an in situ 13CO2 184:950–961.
pulse labelling system. Rapid Commun. Mass Spectrom. Ryan, M.G., R.M. Hubbard, S. Pongracic, R.J. Raison and R.E. McMurtrie.
14:1345–1350. 1996. Foliage, fine-root, woody-tissue and stand respiration in Pinus
Ostle, N., A.S. Whiteley, M.J. Bailey, D. Sleep, P. Ineson and M. radiata in relation to nitrogen status. Tree Physiol. 16:333–343.
Manefield. 2003. Active microbial RNA turnover in a grassland soil Sala, A., D.R. Woodruff and F.C. Meinzer. 2012. Carbon dynamics in
estimated using a 13CO2 spike. Soil Biol. Biochem. 35:877–885. trees: feast or famine? Tree Physiol., 32:764–775.
Palacio, S., E. Paterson, A. Sim, A.J. Hester and P. Millard. 2011. Salmon, Y., R. Barnard and N. Buchmann. 2011. Ontogeny and leaf gas
Browsing affects intra-ring carbon allocation in species with con- exchange mediate the carbon isotopic signature of herbaceous
trasting wood anatomy. Tree Physiol. 31:150–159. plants. Plant Cell Environ. 34:465–479.

798 Epron et al.
Schneider, A. and K. Schmitz. 1989. Seasonal course of translocation Pinus: comparative shapes and some fine structure profiles. Can. J.
and distribution of 14C-labelled photoassimilate in young trees of Bot. 57:845–863.
Larix decidua Mill. Trees Struct. Funct. 3:185–191. Thorpe, M.R., A. Lacointe and P.E.H. Minchin. 2011. Modelling phloem
Schnyder, H., R. Schäufele and R. Wenzel. 2004. Mobile, outdoor con- transport within a pruned dwarf bean: a 2-source-3-sink system.
tinuous-flow isotope-ratio mass spectrometer system for automated Funct. Plant Biol. 38:127–138.
high-frequency 13C- and 18O-CO2 analysis for Keeling plot applica- Tohjima, Y., K. Katsumata, I. Morino, H. Mukai, T. Machida, I. Akama, T.
tions. Rapid Commun. Mass Spectrom. 18:3068–3074. Amari and U. Tsunogai. 2009. Theoretical and experimental evalua-
Schutz, A., W. Bond and M. Cramer. 2009. Juggling carbon: allocation tion of the isotope effect of NDIR analyzer on atmospheric CO2 mea-
patterns of a dominant tree in a fire-prone savanna. Oecologia surement. J. Geophys. Res. 114:D13302.
160:235–246. Treonis, A.M., N.J. Ostle, A.W. Stott, R. Primrose, S.J. Grayston and P.
Simard, S.W., D.M. Durall and M.D. Jones. 1997a. Carbon allocation and Ineson. 2004. Identification of groups of metabolically-active rhizo-
carbon transfer between Betula papyrifera and Pseudotsuga menziesii sphere microorganisms by stable isotope probing of PLFAs. Soil
seedlings using a 13C pulse-labeling method. Plant Soil 191:41–55. Biol. Biochem. 36:533–537.
Simard, S.W., M.D. Jones, D.M. Durall, D.A. Perry, D.D. Myrold and R. Vance, E.D., P.C. Brookes and D.S. Jenkinson. 1987. An extraction
Molina. 1997b. Reciprocal transfer of carbon isotopes between method for measuring soil microbial biomass C. Soil Biol. Biochem.

ectomycorrhizal Betula papyrifera and Pseudotsuga menziesii. New 19:703–707.
Phytol. 137:529–542. Vizoso, S., D. Gérant, J.M. Guehl, R. Joffre, M. Chalot, P. Gross and P.
Søe, A.R.B. and N. Buchmann. 2005. Spatial and temporal variations in Maillard. 2008. Do elevation of CO2 concentration and nitrogen fer-
soil respiration in relation to stand structure and soil parameters in tilization alter storage and remobilization of carbon and nitrogen in
an unmanaged beech forest. Tree Physiol. 25:1427–1436. pedunculate oak saplings? Tree Physiol. 28:1729–1739.
Subke, J.-A., H.W. Vallack, M. Tord, S.G. Keel, D.B. Metcalfe, P. Högberg von Felten, S., S. Hättenschwiler, M. Saurer and R. Siegwolf. 2007.
and P. Ineson. 2009. Short-term dynamics of abiotic and biotic soil Carbon allocation in shoots of alpine treeline conifers in a CO2
13CO effluxes after in situ 13CO pulse labelling of a boreal pine for- enriched environment. Trees Struct. Funct. 21:283–294.
2 2
est. New Phytol. 183:349–357. Warembourg, F.R. and E.A. Paul. 1973. The use of 14CO2 canopy tech-
Talhelm, A., S. Qadir, M. Powers, K. Bradley, A. Friend and K. Pregitzer. niques for measuring carbon transfer through the plant-soil system.
2007. 13C labeling of plant assimilates using a simple canopy-scale Plant Soil 38:331–345.
open air system. Plant Soil. 296:227–234. Warren, J.M., C.M. Iversen, C.T. Garten Jr., et al. 2012. Timing and mag-
Taneva, L., J.S. Pippen, W.H. Schlesinger and M.A. Gonzalez-Meler. nitude of C partitioning through a young loblolly pine (Pinus taeda L.)
2006. The turnover of carbon pools contributing to soil CO2 and soil stand using 13C labeling and shade treatments. Tree Physiol.
respiration in a temperate forest exposed to elevated CO2 concen- 32:799–813.
tration. Glob. Change Biol. 12:983–994. Werner, C. and A. Gessler. 2011. Diel variations in the carbon isotope
Teskey, R.O. and M.A. McGuire. 2007. Measurement of stem respira- composition of respired CO2 and associated carbon sources: a review
tion of sycamore (Platanus occidentalis L.) trees involves internal of dynamics and mechanisms. Biogeosciences 8:2437–2459.
and external fluxes of CO2 and possible transport of CO2 from roots. Wingate, L., J. Ogée, R. Burlett, A. Bosc, M. Devaux, J. Grace, D. Loustau
Plant Cell Environ. 30:570–579. and A. Gessler. 2010. Photosynthetic carbon isotope discrimination
Teste, F.P., S.W. Simard, D.M. Durall, R.D. Guy and S.M. Berch. 2010. Net and its relationship to the carbon isotope signals of stem, soil and
carbon transfer between Pseudotsuga menziesii var. glauca seedlings ecosystem respiration. New Phytol. 188:576–589.
in the field is influenced by soil disturbance. J. Ecol. 98:429–439. Yoneyama, T., L.L. Handley, C.M. Scrimgeour, D.B. Fisher and J.A.
Thompson, R.G., D.S. Fensom, R.R. Anderson, R. Drouin and W. Leiper. Raven. 1997. Variations of the natural abundances of nitrogen and
1979. Translocation of 11C from leaves of Helianthus, Heracleum, carbon isotopes in Triticum aestivum, with special reference to
Nymphoides, Ipomoea, Tropaeolum, Zea, Fraxinus, Ulmus, Picea, and phloem and xylem exudates. New Phytol. 137:205–213.

Tps 057

Uploaded by

Document Informationclick to expand document information

Copyright:

Available Formats

Tps 057

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Tps 057

Uploaded by

Copyright:

Available Formats

Tree Physiology 32, 776–798

Invited review: Part of an invited issue on carbon allocation

Pulse-labelling trees to study carbon allocation dynamics: a review

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Keywords: carbon isotope, forest, partitioning, residence time, transfer time.

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Online at http://www.treephys.oxfordjournals.org

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Volume 32, 2012

Flux Methods Type of tree Species References Main finding

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Online at http://www.treephys.oxfordjournals.org

Flux Methods Type of tree Species References Main finding

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Volume 32, 2012

Flux Methods Type of tree Species References Main finding

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Online at http://www.treephys.oxfordjournals.org

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Volume 32, 2012

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Online at http://www.treephys.oxfordjournals.org

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tracing the labelled C

Expression of isotope composition

Tree Physiology Volume 32, 2012

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Online at http://www.treephys.oxfordjournals.org

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Volume 32, 2012

t­reatments, or even among individuals within a given treat-

Variations in the rate of C transfer among species

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Online at http://www.treephys.oxfordjournals.org

Species Methods References Velocity (m h−1)

Broadleaved Ash, elm 11C (seedlings) Thompson et al. (1979) 0.3–6.0

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Volume 32, 2012

Species Methods References Half-life (days)

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Half-life of labelled C in respiratory substrate pools Compartmental modelling of respiratory C pools

Tree Physiology Online at http://www.treephys.oxfordjournals.org

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Volume 32, 2012

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Online at http://www.treephys.oxfordjournals.org

experiments were thus useful to disentangle the contribution

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Volume 32, 2012

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Online at http://www.treephys.oxfordjournals.org

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Volume 32, 2012

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Online at http://www.treephys.oxfordjournals.org

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

Tree Physiology Volume 32, 2012

Downloaded from https://academic.oup.com/treephys/article/32/6/776/1664870 by guest on 15 May 2023

treatments, or even among individuals within a given treat-