www.elsevier.com/locate/agrformet

Agricultural and Forest Meteorology 131 (2005) 97–124

Carbon balance of coniferous forests growing in

contrasting climates: Model-based analysis

Belinda E. Medlyn a,*, Paul Berbigier b, Robert Clement c, Achim Grelle d,Denis Loustau e, Sune Linder f, Lisa Wingate c, Paul G. Jarvis c,

Bjarni D. Sigurdsson g, Ross E. McMurtrie a

a School of Biological, Earth and Environmental Sciences, University of NSW, Sydney 2052, Australiab INRA-EPHYSE, BP 81, 33883 Villenave d’Ornon Cedex, France

c School of GeoSciences, University of Edinburgh, King’s Buildings, Mayfield Road, Edinburgh EH9 3JG, UKd Swedish University of Agricultural Sciences, Department of Ecology and Environmental Research,

PO Box 7072, SE-750 07 Uppsala, Swedene INRA-EPHYSE, 69 route d’Arcachon, 33612 CESTAS Cedex, France

f Swedish University of Agricultural Sciences, Southern Swedish Forest Research Centre, PO Box 49, SE-230 53 Alnarp, Swedeng Icelandic Forest Research, Mogilsa, IS-116 Reykjavik, Iceland

Received 30 June 2004; received in revised form 13 May 2005; accepted 23 May 2005

Abstract

Forest carbon exchange contributes significantly to the global carbon balance and is therefore being monitored around theworld,

most notably using eddy covariance technology. In order to extrapolate from these measurements, we need to understand why

carbon balance (or net ecosystem production, NEP) differs among forests. Here, we use a detailed model of forest carbon exchange

applied to three coniferous European forests with differing NEP to pinpoint reasons for the differences among these sites. The model

was parameterised using extensive ecophysiological data obtained at each site. These data gave evidence of major differences

among sites in climate, leaf physiology, respiring biomass, leaf area index, and soil and biomass respiration rates. The model was

compared with eddy covariance data and found to satisfactorily simulate carbon exchange by each forest. Simulations were then run

which interchanged canopy structure, physiology and meteorology among sites, allowing us to quantify the contribution of each

factor to the inter-site differences in gross primary productivity (GPP), ecosystem respiration (RE) and NEP. The most important

factor was the difference in respiration rates, particularly soil respiration rates, among sites. Climate was also very important, with

differences in incident photosynthetically active radiation (PAR) affecting GPP and differences in temperature affecting both GPP

and RE. Effects of leaf area index, respiring biomass and leaf physiology on NEP were secondary, but still substantial. The work

provides detailed quantitative evidence of the major factors causing differences in NEP among coniferous forests.

# 2005 Elsevier B.V. All rights reserved.

Keywords: Eddy correlation; Modelling; Net ecosystem exchange; Climate; Soil respiration; Forest ecosystem

* Corresponding author. Tel.: +61 2 9385 2213; fax: +61 2 9385 1558.

E-mail address: b.medlyn@unsw.edu.au (B.E. Medlyn).

0168-1923/$ – see front matter # 2005 Elsevier B.V. All rights reserved.

doi:10.1016/j.agrformet.2005.05.004

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–12498

Nomenclature

A photosynthetic rate (mmol m�2 s�1)

APAR absorbed photosynthetically active

radiation (MJ m�2 yr�1)

Ca atmospheric CO2 concentration

(mmol mol�1)

Ci intercellular CO2 concentration

(mmol mol�1)

D leaf to air vapour pressure deficit (kPa)

DBH diameter at breast height (cm)

gs stomatal conductance (mol m�2 s�1)

GPP gross primary productivity

(g C m�2 yr�1)

h relative humidity

Jmax potential rate of electron transport

(mmol m�2 s�1)

LAI leaf area index (m2 m�2)

LUE light use efficiency (g C MJ�1)

ME model efficiency

NEP net ecosystem production

(g C m�2 yr�1)

PAR photosynthetically active radiation

(MJ m�2 yr�1)

Q10 proportional increase in respiration rate

with a 10 8C increase in temperature

Rg above-ground growth respiration

(g C m�2 yr�1)

Rm above-ground maintenance respiration

(g C m�2 yr�1)

Rsoil soil respiration (g C m�2 yr�1)

RE ecosystem respiration (g C m�2 yr�1)

Vcmax maximum rate of Rubisco activity

(mmol m�2 s�1)

Greek letters

a quantum yield of electron transport

(mol mol�1)

u curvature of the light response of

electron transport

us soil water content in the rooting zone

(kg m�2)

1. Introduction

It is recognised that the world’s forests contribute

significantly to the global carbon (C) balance, and that

changes in forest C uptake may act as an important

feedback to the current increase in atmospheric carbon

dioxide (Malhi et al., 1999). A large research effort is

therefore currently being directed at monitoring forest

C balance around the world (e.g. Sellers et al., 1997;

Baldocchi et al., 2001; Andreae et al., 2002). An

important component of this research effort is the use

of eddy covariance methodology to measure C balance

of forest patches; this methodology has been applied at

over 100 forest sites (Baldocchi et al., 2001).

Extrapolation of these data to forests globally is

now a priority for researchers. A key problem is

identifying the major controls on C balance, in order to

allow results for individual forest patches to be

generalised (IPCC, 2003).

Forest C balance is known to be affected by a wide

range of different factors. A seminal paper comparing

eddy covariance measurements of forest C exchange in

European forests showed that net ecosystem production

of C (NEP) was linearly related to latitude (Valentini

et al., 2000). However, this relationship was shown not

to hold in North America (Law et al., 2002). It is argued

that in Europe, latitude integrates a number of factors

influencing C uptake such as radiation and precipita-

tion, whereas these factors do not vary with latitude in

the same way across North America (Barr et al., 2002).

Law et al. (2002) suggested that an index combining

temperature and water balance would be better able to

explain variation in forest C exchange across both

continents. The index explained 64% of variation in

gross ecosystem productivity (GPP) in all forests and

grasslands studied, but it was less successful in

explaining variability in NEP. The difference between

GPP and NEP is the ecosystem respiration (RE).

Several authors have calculated that RE varies more

than GPP and hence have suggested that RE is the main

determinant of NEP (Valentini et al., 2000; Pilegaard

et al., 2001; Arain et al., 2002). It is known that

respiration is strongly affected by temperature on a

short-term basis but when compared across sites,

respiration is more strongly related to productivity than

to temperature (Janssens et al., 2001). There is evidence

that both autotrophic and heterotrophic respiration

acclimate to temperature (Luo et al., 2001; Gifford,

2003) and it is thought that substrate limitation may

determine respiration in the long term (Dewar et al.,

1999; Gifford, 2003). There is also evidence that

disturbance, such as ploughing or drainage, may have

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 99

effects on soil respiration that continue for many years

(Barford et al., 2001; Janssens et al., 2001) and there is

therefore an increasing emphasis on site history in

interpreting forest NEP (Schulze et al., 2000; Thornton

et al., 2002).

Hence, NEP is affected by a multiplicity of factors,

including climatic factors such as incident radiation,

temperature, and precipitation, site-specific factors

including soil type, nutrient availability and site history,

and interspecific differences in tree physiology. In order

to be able to generalise from flux studies, we need to be

able to quantify the influence of each of these factors

individually. However, separating out the effects of

each factor is difficult, because of confounding: most

eddy flux sites differ in a number of ways. In an attempt

to reduce the number of confounded factors, some

authors have tried to compare fluxes between sites with

at least one factor in common, such as comparing

ecosystems growing in similar climates (Clark et al.,

1999; Anthoni et al., 2002; Griffis et al., 2003) or

comparing similar ecosystems in contrasting climates

(Barr et al., 2002). An alternative approach, taken here,

is to use a model to directly estimate the contribution of

different factors to the difference in NEP among sites.

We focussed on three coniferous forests growing in

contrasting climatic conditions: a boreal Norway

spruce forest (Flakaliden, Sweden), a maritime Sitka

spruce forest (Griffin, Scotland), and a temperate pine

forest (Bray, France). Net ecosystem production

Table 1

Site characteristics

Bray, France Flakalide

Latitude 448420N 648070NLongitude 08430W 198270EElevation (m a.s.l.) 60 310

Dominant species Pinus pinaster Ait. Picea ab

Year planted 1970 1963

Period modelled March 1997–February 1998 Novemb

Annual mean air

temperature (8C)

12.9 2.0

Soil type Sandy podsol Sandy p

Understorey Grass (Molinia coerulea

L. Moench)

Dwarf sh

Vacciniu

Net ecosystem production

(g C m�2 yr�1)

575 �50

Wood increment

(t C ha�1 yr�1)

4.6 1.1

References Berbigier et al. (2001) Bergh et

varied considerably between these sites; it was

estimated from eddy covariance data to be

�50 g C m�2 yr�1 for Flakaliden, 575 g C m�2 yr�1

for Bray, and 590 g C m�2 yr�1 for Griffin (see

Section 2.1). Our aim was to identify the main

reasons for this difference in NEP among the sites.

We parameterised the forest canopy model MAES-

TRA (Wang and Jarvis, 1990a; Medlyn, 2004) for each

site, using extensive ecophysiological data obtained at

each site. The process of parameterisation enabled us to

identify the major differences among the sites:

differences in climate, canopy structure, leaf physiol-

ogy, and respiration rates were quantified. The model

was used to simulate NEP at each site and results

verified against eddy covariance data. We then ran a

series of simulations designed to quantify the effect of

each inter-site difference on NEP. Single parameters, or

groups of related parameters, were changed one at a

time, and the change in NEP calculated. This process

allowed us to rank individual factors in terms of their

importance in determining NEP.

2. Methods

2.1. Sites

The three eddy flux sites used included a Sitka

spruce forest in Scotland (Griffin), a Norway spruce

n, Sweden Griffin, U.K.

568370N38480W340

ies (L.) Karst Picea sitchensis (Bong.) Carr.

1981

er 1997–October 1998 January 1998–December 1998

6.3

odsolic glacial till Podsolised brown earth;

stagnohumic gley

rubs (chiefly

m spp.); mosses

Calluna vulgaris (L.) Hull; Erica spp.;

sparse under closed canopy

590

6.5

al. (1999) Clement (2004) and

Clement et al. (2003)

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124100

forest in Sweden (Flakaliden), and a maritime pine

forest in southern France (Bray). A summary of site

characteristics is given in Table 1. Griffin is the most

productive of the three sites. It is a young stand just

reaching canopy closure and suffers few climatic

extremes. Flakaliden is subject to harsh winters and

rates of growth are slow; the site is the oldest of the

three and yet the trees are smallest. Bray is subject to

summer droughts, but growth rates are relatively fast

and the trees are approaching maturity.

All three sites formed part of the EUROFLUX

network. In this network, a standard methodology was

used to measure eddy covariance, as described by

Aubinet et al. (2000). A continuous period of 1 year’s

worth of flux data was chosen for each site. For the

Bray and Flakaliden sites, this did not correspond to a

calendar year (Table 1). For ease of comparison

between sites, however, data are rearranged in the

figures to show a calendar year (e.g. January and

February data from Bray are for 1998 but data for

March onwards are for 1997).

Table 1 gives estimates of annual NEP for each site.

For the Flakaliden site, this estimate was taken from

the FLUXNET website (http://www.eosdis.ornl.gov/

FLUXNET) and was gap-filled using a lookup table

method with a correction for low friction velocity

(Falge et al., 2001). For the Bray site, the estimate is

taken from Berbigier et al. (2001), who used

polynomial regressions on meteorological data to fill

gaps in NEP. A similar method was used to obtain

annual NEP for the Griffin site (Clement, 2004).

Annual NEP was similar for Bray and Griffin, but

much smaller for Flakaliden.

2.2. Modelling approach

The heart of the paper is a modelling analysis that

aims to identify the major reasons for differences in

NEP among sites. The analysis consists of a series of

simulations in which drivers or parameters were

changed one at a time from one site to the next. At

each step, the impact of the change on modelled GPP,

RE and NEP was recorded. Each step is essentially a

‘‘what if’’ question: if the only difference between the

two sites was the factor being changed, what would be

the difference in NEP? For example, when incident

PAR is changed, we are asking what the difference in

NEP would be for two sites which were identical apart

from incident PAR. Clearly these simulations do not

represent reality: they are a type of sensitivity analysis

designed to work out which factors are most important

in determining differences in NEP among sites.

The model used was MAESTRA (Medlyn, 2004),

which is a development of the MAESTRO model

(Wang and Jarvis, 1990a). It is available on-line at

www.maestra.unsw.edu.au. This model was chosen

because it is quite detailed and incorporates all of the

major factors of interest. The forest overstorey is

represented as a three-dimensional array of tree

crowns, and for each crown, x and y co-ordinates,

height, trunk space, canopy radius, and one-sided leaf

area are specified. Radiation interception, photosynth-

esis and stomatal conductance are calculated for a

‘‘target’’ crown, which is divided into 72 subvolumes.

Radiation interception for each subvolume is calcu-

lated for both beam and diffuse radiation. Photosyn-

thetic uptake is driven by intercepted radiation, using

the leaf photosynthesis model of Farquhar et al.

(1980), while stomatal conductance may be repre-

sented by either of three models (Ball et al., 1987;

Jarvis, 1976; Leuning, 1995).

For the Bray and Flakaliden sites, carbon uptake by

the understorey is significant (Delzon, 2000; B.D.

Sigurdsson, personal communication) and hence the

model was modified to allow estimation of this

component. Radiation penetration to the forest floor

was calculated using MAESTRA routines and used to

drive a simple sun-shade photosynthesis model

(Medlyn et al., 2000). Routines were also added to

the model to calculate the respiratory component of

the CO2 flux. Above-ground maintenance respiration

is calculated from foliar biomass and above-ground

woody biomass using exponential functions of air

temperature. Total growth respiration is calculated as a

proportion of biomass increment and allocated within

the growth period according to air temperature. Soil

respiration is calculated from functions of soil

temperature developed for each site. Net ecosystem

production is calculated as the difference between

gross photosynthetic uptake and total respiration.

We first verified that the model was able to

adequately reproduce measured CO2 fluxes. First,

model estimates of ecosystem respiration were

compared with night-time CO2 fluxes. Average

measured and modelled fluxes were calculated for

each night for which flux data were available. The

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 101

night-time fluxes at Griffin were corrected for CO2

storage and hence should be comparable with

modelled respiration. At Bray and Flakaliden, the

fluxes were not corrected for storage, so it is expected

that the modelled respiration will exceed measured

CO2 flux except on turbulent nights. A relationship

between night-time flux on turbulent nights and the

temperature of the soil–litter interface, developed by

Berbigier et al. (2001), was used to further evaluate

modelled respiration at the Bray site. Measured

daytime NEP was then compared with modelled

daytime NEP, calculated by subtracting modelled

daytime respiration from modelled daily GPP. Finally,

monthly ensembles (i.e., averages of all data for a

given time of day over the month) of NEP were

calculated for both measured and modelled data. The

goodness of fit of modelled daytime totals and

monthly ensembles to measured data is evaluated

using the model efficiency (ME), which estimates the

proportion of the variance of the data explained by the

1:1 line (Mayer and Butler, 1993).

The sensitivity analysis was then carried out as

follows. The contribution of the understorey to NEP at

each site was recorded separately. Model simulations

were run for the canopy only. Beginning with a

simulation for one of the sites, the model inputs were

changed to those of another site, one group at a time,

cumulatively, until all inputs corresponded to the

second site. At each step, the changes in GPP, RE and

NEP were recorded. The input meteorological data

were changed first, to evaluate the effect of climate

alone. The soil respiration relationship was changed

second; this relationship is shown (see below) to be

very important, and estimated NEP becomes unrea-

listic if the correct relationship is not used for any

given site. Canopy structural and physiological

parameters were then changed. The order in which

these parameters were changed was found not to affect

the conclusions drawn.

2.3. Parameterisation

In this section, we describe in turn each of the

inputs (climatic variables and parameter sets) used to

drive the model. This section is important because

differences in the model inputs should reflect the

actual differences among sites. We provide consider-

able detail about how the inputs were determined for

each site, with the aim of allowing the reader to judge

the validity of each difference in the inputs.

2.3.1. Climate

For each site, half-hourly meteorological data

including incident photosynthetically active radiation

(PAR), air and soil temperatures, relative humidity,

wind speed and precipitation were obtained from the

EUROFLUX database (http://daacl.esd.ornl.gov/

FLUXNET/euro_db.html). Soil water content does

not affect production at Flakaliden (Bergh et al., 1999)

and was assumed not to affect production at Griffin in

the modelled year. At the Bray site, values of soil

water content measured using time-domain reflecto-

metry were included in the meteorological data used to

drive the model (Ogee et al., 2001).

There were strong contrasts in climate among sites.

Annual courses of incident PAR and air temperature

for each site are illustrated in Fig. 1. Incident PAR is

clearly highest at Bray, the southernmost site. Total

annual incident PAR is similar at Griffin and

Flakaliden, but varies more strongly over the course

of the year at Flakaliden. Monthly air temperature is

highest throughout the year at Bray. Winters are

coldest at Flakaliden but mean June and July

temperatures are lowest at Griffin.

2.3.2. Tree dimensions

Average tree diameter at breast height (DBH),

tree height, height to crown base, and crown radius

are given in Table 2. These values were determined

as follows. At Bray, the positions of some 2500 trees

surrounding the flux tower have been recorded and

DBH is measured each year. At the end of 1996 and

1997, heights were measured on 200 trees and a

relationship with DBH developed. At the end of

1995, crown radius and height to crown base were

determined for 14 trees and a relationship with tree

height obtained (Porte, 1999). At Flakaliden, all

measurements were made on a nearby fertilisation/

irrigation experiment (Linder, 1995). Model inputs

were based on one of the Control plots of this

treatment (275 trees). DBH and height of each tree

were measured at the end of 1996 (Bergh et al.,

1999), and canopy radius was measured for each tree

in summer 1997 (Stenberg et al., 1999). At Griffin,

DBH in 1997 was measured on 1000 trees distributed

in fifty 10 m � 10 m sample plots located around the

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124102

Fig. 1. Monthly total photosynthetically active radiation (PAR) and average air temperature for each site, for the periods March 1997–February

1998 for Bray; November 1997–October 1998 for Flakaliden; January 1998–December 1998 for Griffin. Data for Bray and Flakaliden have been

rearranged to simplify comparison.

flux tower. Height versus DBH and canopy radius

versus DBH relationships were developed from

measurements of height and DBH on 200 trees in

1998 and 2001, and of canopy radius of 70 trees in

2001 (L. Wingate, R. Clement, personal commu-

nication).

The major differences in tree dimensions among

sites may be summarised as follows. Trees were tallest

and largest at the Bray site. However, because the

stocking was relatively low at Bray, the ground cover

fraction, or the ratio of total crown area: ground area,

was highest at Griffin and least at Bray.

2.3.3. Leaf area index

Leaf area index at the Bray site was measured

monthly using an optical method (Berbigier and

Bonnefond, 1995). Values in 1997 ranged from 2.6

to 3.1 m2 m�2 (Berbigier et al., 2001). Leaf area index

surrounding the tower at Flakaliden was marginally

less, at 2.2 m2 m�2, determined with a Li-Cor LAI2000

plant canopy analyser (Li-Cor Inc., Lincoln, NE, USA)

calibrated against biomass samples (Wallin et al.,

2001). A detailed study of leaf area index, comparing

optical and destructive methods, has been carried out at

Griffin (Wingate, 1998). Estimates of LAI varied

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 103

Table 2

Model parameters for each site

Parameter Bray Flakaliden Griffin

Canopy structure

Stocking (trees ha�1) 595 2480 2200

Mean tree height/height to

crown base (m)

17.6/10.5 5.0/0.5 6.6/0.3

Mean DBH (cm) 26.9 6.4 9.9

Canopy radius (m) 1.7 1.0 1.5

Leaf area index (m2 m�2) 2.6–3.1 2.2 6.5

Foliage clumping factor 0.76 0.64 0.7

Leaf physiology

Jmax at 25 8C (mmol m�2 s�1) 180 (C), 120 (C + 1) 63 100.6 (top), 71.1 (middle),

41.2 (bottom)

Vcmax at 25 8C (mmol m�2 s�1) 120 (C), 80 (C + 1) 37 48.3 (top), 34.4 (middle),

18.3 (bottom)

a (mol mol�1) 0.2 (C), 0.15 (C + 1) 0.3 0.3

u 0.95 0.71 0.71

Temperature dependence of Jmax

Ha (kJ mol�1) 38.67 100.28 50.03

Hd (kJ mol�1) 200 147.92 201

Topt (8C) 41.16 21.76 32.48

Temperature dependence of Vcmax

Ha (kJ mol�1) 58.52 69.83 59.45

Hd (kJ mol�1) 200

Topt (8C) 31.58

T range over which Jmax,

Vcmax decline to 0 (8C)

10 to 0 See text 6 to �2

Respiration

Foliar respiration rate at 25 8C(mmol m�2 s�1)

0.74 1.13 1.3 (top), 1.0 (middle),

0.7 (bottom)

Q10 for foliage maintenance respiration 2.0 2.0 (summer)–2.8 (winter) 2.0

Above-ground woody biomass (kg m�2) 9.4 2.6 5

Wood respiration rate at 15 8C(mmol kg�1 s�1)

0.117 0.036 0.117

Q10 for wood maintenance respiration 2.13 2.0–2.5 2.0

Understorey

Cover (%) 0–100 Moss 29.1; V. myrtillus

18.7; V. vitis-idea 14; total 73.8

–

Maximum LAI (m2 m�2) 1.88 0.8 –

a (mol mol�1) 0.09 0.234, 0.28, 0.234 –

Jmax (mmol m�2 s�1) at 25 8C 35.6 27.8, 36, 34.4 –

Vcmax (mmol m�2 s�1) at 25 8C 26.0 16.2, 17.5, 21.2 –

Foliar respiration rate at 25 8C(mmol m�2 s�1)

0.19 1.1, 0.5, 1.0 –

See text for sources.

according to method; we used the mean stand LAI

obtained from destructive measurements. At

6.5 m2 m�2, LAI was clearly highest at the Griffin site.

2.3.4. Canopy structure

In MAESTRA, it is possible to vary the leaf

incidence angle distribution, canopy leaf area density

distribution, and foliage clumping factor. The leaf

incidence angle distribution was not measured at any

of the sites and hence was assumed to be spherical—a

common assumption for conifers (Jarvis and Lever-

enz, 1983). The within-crown leaf area density

distribution was measured for P. sitchensis by Wang

and Jarvis (1990b) and for three age classes of foliage

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124104

of P. pinaster by Porte et al. (2000). No information on

leaf area density distribution was available for P.

abies, so for this species, a uniform leaf area density

distribution was assumed. Leaf area density for P.

sitchensis was highest in the mid-canopy and towards

the exterior of the crown, while leaf area density for

P. pinaster was highest at the top of the canopy and

towards the exterior of the crown.

The foliage clumping factor accounts for non-

randomness of foliage distribution due to the clumping

of needles into shoots (Oker-Blom and Kellomaki,

1983). For P. sitchensis, the clumping factor was given

as 0.7 by Wang and Jarvis (1990b). For P. pinaster and

P. abies, mean values of 0.76 and 0.64 were derived

from measurements of shoot projected area ratio made

by Bosc (1999) and Stenberg et al. (1995), respec-

tively. However, there is considerable variation about

the mean in both of these studies, suggesting that the

difference in these parameters among sites is unlikely

to be significant.

2.3.5. Needle reflectance and transmittance of

PAR

These parameters were assumed not to differ

among sites. Values have been determined accurately

only for P. pinaster. Using an integrating sphere,

Berbigier and Bonnefond (1995) found average needle

reflectance in the 400–700 nm waveband was 9% and

transmittance was 1.4%. Approximate values of

needle reflectance and transmittance obtained for P.

sitchensis by Norman and Jarvis (1974) using a

radiation sensor were very similar to those obtained

for P. pinaster, while measurements for P. abies were

unavailable.

2.3.6. Photosynthesis

Photosynthetic rates were estimated according to

the model of Farquhar et al. (1980), for which the key

parameters are the maximum rate of Rubisco activity,

Vcmax, the potential rate of electron transport, Jmax, and

the initial slope, a, and curvature, u, of the light

response of electron transport. The parameters Jmax

and Vcmax may be obtained from A–Ci curves (the

response of photosynthesis to intercellular CO2). In

order to assure comparability of these parameters, A–

Ci curves were obtained for each species and

parameters were fitted to these curves using a

consistent methodology (Medlyn et al., 2002a).

For P. pinaster, A–Ci curves were measured on

current foliage taken from the top of the canopy at

2-month intervals throughout the year, from a stand

neighbouring the Bray site (Medlyn et al., 2002b).

Information on gradients in these parameters through

the canopy and changes with age was also available

from a study on the same stand (Porte and Loustau,

1998). Measurements were made with a Walz

Compact Minicuvette System (Heinz Walz GmbH,

Effeltrich, Germany) on six individual needles laid

across the cuvette. The needles were illuminated from

both sides with saturating light (1400 mmol m�2 s�1).

Measurements were made at five different tempera-

tures from 15 to 35 8C. Values were converted to a

half-total leaf surface area basis, as required by

the model, using a total: projected area ratio of

2.57.

For P. sitchensis, A–Ci curves were measured on

foliage in situ at three different levels in the canopy.

These measurements were made at a site in southern

Scotland with LAI similar to that of the Griffin stand

(Meir et al., 2002). Measurements were made with a

CIRAS-1 gas exchange system (PP Systems,

Hitchin, Hertfordshire, UK) on an intact shoot.

Saturating illumination (800–2000 mmol m�2 s�1)

was supplied from one side of the cuvette and the

other side lined with reflective plastic. Air tempera-

ture averaged 22 8C. The total: projected area ratio

was assumed to be 2.31 (Barclay and Goodman,

2000).

For P. abies, photosynthesis measurements were

made for an elevated CO2 experiment using branch

bags at the Flakaliden irrigation/fertilisation experi-

ment (Roberntz and Stockfors, 1998). The measure-

ments used here were made on unbagged control

branches in an untreated control plot. Measurements

were made on current-year shoots on the third or

fourth whorl of branches, using a Li-6200 infrared gas

analyser (Li-Cor Inc., Lincoln, NE, USA). Illumina-

tion of 1000 mmol m�2 s�1 was provided from one

side only, and chamber temperature averaged 19.5 8C.

The total: projected area ratio was assumed to be 2.74

(Riederer et al., 1988).

Samples of the A–Ci responses obtained are

shown in Fig. 2. There is a considerable difference

between the relationships, with that for P. pinaster

being the uppermost and that for P. abies lowermost.

The initial slope and saturating values of these

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 105

Fig. 2. Samples of measured photosynthesis–intercellular CO2 (A–Ci) responses used for each site. At Bray, the response was measured in

January, on current needles, on a cut branch taken from the third whorl. At Flakaliden, the response was measured in September, on current

needles, on a branch in the third whorl. Data for Griffin were measured at a plantation with similar leaf area index in southern Scotland. The

response was measured on 1-year-old shoots at the top of the canopy. All curves were taken at 20 8C and all are expressed on a half-total leaf

surface area basis. Data supplied by Peter Roberntz (Flakaliden), Patrick Meir (Griffin) and Belinda Medlyn (Bray).

relationships are used to derive Vcmax and Jmax,

respectively. The parameter values derived for each

species are given in Table 2. As indicated by the

relative positions of the relationships shown in

Fig. 2, these parameters are highest for P. pinaster

and lowest for P. abies.

Values for the parameters of the light response of

Jmax, a and u, were taken from Porte and Loustau

(1998) for P. pinaster. They assumed u to be constant

and equal to 0.95 and then fitted a to light-response

curves of photosynthesis made at ambient CO2

concentration. We corrected these values for absorp-

tance, which was assumed to be 89.5% in the visible

waveband (Berbigier and Bonnefond, 1995). For P.

sitchensis, we used values of a and u estimated by

Ibrom et al. (unpublished manuscript) from a study by

Leverenz and Jarvis (1979), who measured the light

response of shoots which had been trimmed such that

needles were arranged in a single layer, perpendicular

to the light source. As no comparable measurements

were available for P. abies, it was assumed that a and u

were identical to those for P. sitchensis. The difference

in these parameters implies that photosynthesis

increases more rapidly with increasing PAR in the

spruces as opposed to the pine.

2.3.7. Dependence of photosynthesis on

temperature

The temperature response of the parameters Jmax

and Vcmax was measured for P. pinaster at 2-month

intervals (Medlyn et al., 2002b). The temperature

response given here was obtained by fitting the

equations given by Medlyn et al. (2002a) to data from

all periods combined. For P. sitchensis, the tempera-

ture response of Jmax and Vcmax was estimated from

the temperature response of photosynthesis given by

Neilson et al. (1972). For P. abies, the temperature

responses were assumed to be the same as those of

Pinus sylvestris growing in similar boreal conditions

in Finland (Wang et al., 1996; Medlyn et al., 2002a).

The temperature responses are quite different across

the sites, with the optimal temperature for Jmax being

20 8C lower at the northernmost site (Flakaliden)

compared to the southernmost (Bray).

There is very little data available on the effects of

very low temperatures on photosynthesis. Inspection

of the flux data, however, showed that CO2 uptake

decreased rapidly below threshold temperatures at the

two milder sites. Based on this observation, the model

was modified such that Jmax and Vcmax were assumed

to decline to zero at low temperatures at these sites. At

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124106

Flakaliden, photosynthesis is inhibited over the winter

by frozen soil. The recovery of photosynthesis in

spring was modelled according to Bergh et al. (1998).

2.3.8. Stomatal conductance

Stomatal conductance data were obtained from

trees at each site. At the Bray site, stomatal

conductance was measured by Bosc (1999) using an

automatic ‘‘branch bag’’ style cuvette. The cuvette

was installed on a current-year branch near the top of

the canopy and measurements of branch stomatal

conductance were made every 20 min over a period of

40 days in September and October 1997. These data

span a drought period and hence allow the effect of

low soil water content on stomatal conductance to be

estimated. Similarly, at Griffin, branch bags were

operated as cuvettes for 3 days in July 2001 and

measurements of branch gas exchange made every

30 min (Wingate, 2003). For the Flakaliden site,

stomatal conductance was taken from the same set of

shoot gas exchange data used to derive photosynthetic

parameters. It was found that stomatal conductance

did not change during the measurement of A–Ci curves

and hence mean stomatal conductance from these

measurements was calculated and related to the

Fig. 3. Stomatal conductance data obtained from each site. The x-axis is the

multiplied by relative humidity (h) divided by atmospheric CO2 concentration

for the other sites, the full datasets are shown. Data are expressed per unit hal

Lisa Wingate (Griffin) and Alexandre Bosc (Bray). Lines are fitted to data

conditions prevailing before the beginning of the

measurement (Roberntz and Stockfors, 1998).

These data are compared in Fig. 3, with the

stomatal index proposed by Ball et al. (1987) on the x-

axis. As with photosynthetic rate, stomatal conduc-

tance appears to vary among the three species, being

highest for P. pinaster and lowest for P. abies. The

relationships shown in Fig. 3 are all significantly

different (P < 0.01).

The stomatal model of Leuning (1995) was fitted to

the data from P. pinaster and P. sitchensis. For P.

sitchensis, the following model was obtained:

gs ¼ 0:037 þ 6:33A

Cað1 þ D=0:5Þ (1)

where gs is stomatal conductance (mol m�2 s�1), A is

photosynthesis (mmol m�2 s�1), Ca is atmospheric

CO2 (mmol mol�1) and D is leaf to air vapour pressure

deficit (kPa). For P. pinaster, the fitted model includes

a dependence on soil water content, us:

gs ¼�

0:015 þ 8:63A

Cað1 þ D=1:39Þ

�f ðusÞ (2)

The dependence on soil water content was taken

from a study of sapflux of P. pinaster (Granier and

stomatal index proposed by Ball et al. (1987), viz. photosynthesis (A)

(Ca). For Bray, only data at high relative soil water content are shown;

f-total leaf surface area. Data supplied by Peter Roberntz (Flakaliden),

. Solid line, Bray; dashed line, Flakaliden; dotted line, Griffin.

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 107

Loustau, 1994) and is given by

f ðusÞ ¼ 1 � 0:0156 exp

�4:269

us;max � us

us;max � us;min

�(3)

where us,max and us,min are the maximal and minimal

root zone soil water content, respectively. For the P.

abies data, it was not possible to fit the model of

Leuning (1995) because of the small range of vapour

pressure deficits in the data set and hence the model of

Ball et al. (1987) was fitted instead, obtaining:

gs ¼ 0:034 þ 3:5Ah

Ca

(4)

where h is relative humidity.

2.3.9. Foliage respiration

A detailed study of foliage maintenance respiration

was made on P. abies (Stockfors, 1997; Roberntz and

Stockfors, 1998; Stockfors and Linder, 1998b). The

measurements were made on detached needles from

current-year shoots taken from the third whorl of

branches at the Flakaliden irrigation-fertilisation

experiment. Needles were dark-adapted for at least

one hour before measurements were made using an

oxygen electrode. Measurements at 10 and 20 8C were

made on seven occasions through the year. Needle

respiration varied significantly through the year, with

wintertime rates double those during summer (Stock-

fors and Linder, 1998b). The temperature response of

needle respiration also varied over the course of the

year: Q10 ranged from 2 in summer to 2.8 in winter

(Stockfors and Linder, 1998b).

Needle maintenance respiration of P. sitchensis

trees was measured during the course of an elevated

CO2 experiment using branch bags at a site in southern

Scotland (Barton, 1997; Barton and Jarvis, 1999).

Here, measurements were made on shoots in situ using

an infrared gas analyser. Respiration rates were lower

for dark-adapted shoots during daytime than at night,

suggesting that foliage respiration rate is reduced in

daytime (Barton, 1997). Here, the night-time rate was

used as the base rate for foliage at the top of the

canopy. Respiration rate per unit foliage area was

assumed to decline with canopy depth, as found by

Meir et al. (2002).

For P. pinaster, needle maintenance respiration

rates were estimated from A–Ci curves measured in

situ at three levels in the canopy and on two age classes

of foliage (Porte and Loustau, 1998). In these

measurements, no attempt was made to dark-adapt

the needles. The respiration rate did not vary

significantly through the canopy. Information on

Q10 was not available for P. sitchensis or P. pinaster.

For these two species, Q10 was assumed to be 2.

For all three sites, it was assumed that foliage

respiration rate is reduced by 60% during the day

(Brooks and Farquhar, 1985). Foliage respiration

parameter values corrected to 25 8C are shown in

Table 2. Foliage respiration rate is highest for P.

sitchensis, followed by P. abies then P. pinaster.

However, it is difficult to be confident that this

difference between species is real, due to the

considerable differences in experimental protocol

outlined above.

There was no evidence for any difference in growth

respiration rates among sites, so growth respiration of

foliage and wood were calculated as constant fractions

(respectively 0.25 g C g�1 C and 0.15 g C g�1 C;

Chung and Barnes, 1977) of total production for all

sites.

2.3.10. Above-ground woody respiration

Detailed studies of stem respiration have been

made at two of the three sites. At the Bray site, Bosc

et al. (2003) measured respiration of excised branch

and stem sections which ranged in diameter from 0.9

to 4 cm for branches and 6 to 24 cm for stems.

Measurements were made in the laboratory under

controlled conditions in order to accurately determine

respiration at a constant temperature (15 8C). Respira-

tion rates were found to decrease with the age of the

section measured, irrespective of whether rates were

expressed per unit surface area, volume, or nitrogen.

Measurements were made in October, well after the

end of the growing season, and respiration rates were

assumed to represent maintenance respiration only.

Field measurements on branch segments were also

made using an automated cuvette over a period of 3–7

days (Bosc et al., 2003), allowing the temperature

response to be determined. A Q10 of 2.13 was

obtained.

At the Flakaliden irrigation-fertilisation experi-

ment, stem respiration was measured in situ on the

north-facing side of stems at a height of 1–1.5 m on

seven occasions between June 1993 and April 1994

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124108

(Stockfors and Linder, 1998a). On each occasion,

continuous measurements were made over a period of

24–48 h to allow determination of the temperature

response. The Q10 varied from 2.0 to 2.5 over the

growing season (Table 2). In a second study, stem

respiration was measured in situ using an automated

system throughout 1999 (Ceschia, 2001). Respiration

was measured at three different heights. During

summer 1999, automated measurements were also

made on first-order branches at three heights. In

October 1999, two trees were harvested and measure-

ments of respiration made on stem and branch

segments under controlled conditions in the labora-

tory. In this study, maintenance respiration rates were

also shown to depend on segment diameter, whether

expressed on a surface or volume area basis. The rates

obtained for P. abies are compared with those

measured by Bosc et al. (2003) on P. pinaster in

Fig. 4. It will be seen that, for a wood segment of

similar diameter, maintenance respiration rates at a

constant temperature are considerably higher for

P. pinaster.

The information from these studies was used to

parameterise a simple model of woody respiration.

Maintenance respiration at a reference temperature

was assumed constant throughout the year and was

calculated on a mass basis. The parameters derived for

the Bray and Flakaliden sites are given in Table 2. As

Fig. 4. Measured woody maintenance respiration as a function of stem or

(2001) (Flakaliden). All data corrected to 15 8C.

would be expected from Fig. 4, the effective wood

maintenance respiration rate was considerably higher

for Bray than for Flakaliden. No information was

available on stem respiration rates for the Griffin site

or for P. sitchensis. For this site, parameters were taken

from Bray because the trees there were growing at a

similar rate to those at Griffin.

2.3.11. Characterisation of understorey

The understorey was significant at the Bray and

Flakaliden sites, but sparse at the Griffin site, and

hence was modelled at the first two sites only.

The understorey at Bray was chiefly composed of a

grass, Molinia caerulea (L.) Moench, and formed the

subject of a detailed study by Delzon (2000). Grass

clumps were harvested on a fortnightly basis

throughout the growing season to determine above-

ground biomass and leaf area. Leaf extension was also

measured non-destructively on a weekly basis. Gas

exchange measurements, including responses to light,

CO2 and temperature, were made on grass clumps

removed from the site to a nearby greenhouse. These

measurements were used to parameterise a model of

photosynthesis and tested using further gas exchange

measurements made on grass clumps in situ. Key

parameters taken from this work and used to

parameterise the understorey model used here, are

given in Table 2.

branch diameter. Data from Bosc et al. (2003) (Bray) and Ceschia

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 109

The understorey at Flakaliden was studied in

autumn 1998 by manually harvesting fifty 20 cm �20 cm plots that were randomly distributed over

control plots of the irrigation-fertilisation experiment.

The samples from each plot were divided into different

understorey species, measured for leaf area by LI-

3000 area meter (Li-Cor Inc., Lincoln, NE, USA),

dried in an oven at 80 8C for 48 h and weighed. The

three main understorey classes were Vaccinium

myrtillus L., V. vitis-idea L. and various moss species,

with 23, 32 and 33% of the understorey biomass,

respectively. In September 1998, light-response

curves and A–Ci curves were measured for the three

main understorey species using a LI-6200 Portable

Photosynthesis System (Li-Cor Inc., Lincoln, NE,

USA) and a light source. These data were used to

parameterise the biochemical model of Farquhar et al.

(1980). The parameters derived are given in Table 2.

2.3.12. Soil CO2 efflux

Soil CO2 efflux at the Bray site was measured over

the period January 2000–June 2002 using both (1) a

continuous automated open-chamber system with five

chambers of 25 cm diameter and (2) a portable manual

system (SRC-1, PP Systems, Hitchin, Hertfordshire,

England) (Pargade, 2000). The automated system was

used to estimate the Q10 of soil respiration on a monthly

basis, while the portable system was used to calculate a

spatial average (30 replicates) of the respiration rate at

15 8C (D. Loustau, personal communication).

At Flakaliden, soil CO2 efflux around the tower was

measured using a differential chamber system

Fig. 5. Regressions of soil respiration vs. soil temperature at 0–10 cm dept

mean annual soil temperature is marked by an asterisk for each site.

described by Moren and Lindroth (2000). Measure-

ments were made on three occasions at 36 sample sites

with varying soil moisture content. Also, at the nearby

fertilisation-irrigation experiment, a detailed study of

soil respiration was made as part of a soil warming

study (Stromgren, 2001). An automated chamber

system was used, similar to that used at the Bray site.

Continuous measurements were made from June to

October and relationships with temperature derived

for each month.

At Griffin, respiration was measured with a portable

CO2 analyser equipped with a soil chamber (SRC-1 and

EGM-3, PP Systems, Hitchin, Hertfordshire, England).

Measurements were made on 24 occasions between

August 2000 and June 2001. Sample sites were chosen

to reflect variations in site microtopography (L.

Wingate, personal communication). A relationship

with soil temperature was developed.

The relationships derived for August at each site

are compared in Fig. 5. For most values of soil

temperature, soil CO2 efflux is greatest at Flakaliden

and least at Bray.

3. Results

3.1. Model-data comparison

Although there are distinct limitations to the value

of a comparison between model output and eddy

covariance measurements (see Medlyn et al., 2005 for

a full discussion), we did verify that the model was

h for August for each site. The soil respiration rate corresponding to

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124110

able to adequately reproduce measured fluxes. We first

looked at night-time fluxes, as these consist of

respiration alone, whereas daytime fluxes include

photosynthetic uptake as well as respiratory losses.

We compared modelled night-time respiration with

measured night-time CO2 flux (Fig. 6). At Bray, as

expected, the modelled respiration values exceed the

night-time CO2 fluxes most of the time, with the

modelled values generally following the upper limit of

measurements. The night-time fluxes at Bray were not

corrected for CO2 storage within the canopy, which

tends to occur on calm nights. However, the modelled

values coincide well with respiration values estimated

Fig. 6. Annual courses of modelled and measured ecosystem respiration

values, which are estimated as average night-time CO2 efflux. Solid line

indicates respiration estimated from a relationship between night-time CO

interface. Data for Bray and Flakaliden have been rearranged to show a

from the relationship between CO2 flux on turbulent

nights (u* > 0.4 m s�1) and the soil–litter interface

temperature. At both Flakaliden and Griffin, modelled

respiration follows fairly closely the upper limit of

measured night-time CO2 flux, with a large number of

values overestimated (although not so severely as at

Bray). Many authors have noted previously that, even

where night-time fluxes are corrected for CO2 storage,

fluxes tend to be low when turbulence is low,

suggesting a systematic underestimation by the flux

measurement system (Lavigne et al., 1997; Aubinet

et al., 2000; Falge et al., 2001; Barr et al., 2002). We

therefore regard the congruence between the modelled

for (a) Bray, (b) Flakaliden and (c) Griffin. Circles show measured

is modelled average night-time respiration. In (a), the dotted line

2 efflux on turbulent nights, and the temperature of the litter–soil

calendar year to simplify comparison (see Fig. 1).

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 111

values and the upper limit of the night-time CO2 fluxes

as evidence that ecosystem respiration is being

satisfactorily modelled.

We then looked at daytime fluxes. Daytime NEP

was modelled by subtracting modelled daytime

respiration from modelled GPP, and was compared

with measured daytime CO2 flux (Fig. 7). In general,

Fig. 7. Annual courses of measured (solid circles) and modelled (lines) day

(c) Griffin. The figure also shows (on right) modelled vs. measured daily N

lines. ME is the model efficiency. Data for Bray and Flakaliden have been re

variation in daytime GPP is captured well by the

model. Model performance is particularly good for the

Griffin and Flakaliden sites. The model performs less

well at the Bray site, where fluxes tend to be

overestimated in spring and underestimated in mid-

summer. These results are also clear from the

ensemble comparisons, shown in Fig. 8. The ensemble

time net ecosystem production (NEP) for (a) Bray, (b) Flakaliden and

EP. The solid lines are regression lines and the dashed lines are 1:1

arranged to show a calendar year to simplify comparison (see Fig. 1).

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124112

values are averages of NEP for each half-hour over each

month. The comparison of modelled with measured

ensemble NEP is excellent for the Griffin and

Flakaliden sites. At Bray, daytime CO2 uptake increases

steadily between February and June, but this pattern is

Fig. 8. Measured (solid circles) and modelled (lines) monthly half-hour ne

and (c) Griffin. Ensembles are calculated by averaging NEP for each half

modelled vs. measured ensemble NEP. The solid lines are regression lines a

Bray and Flakaliden have been rearranged to show a calendar year to sim

not captured by the model. We believe this problem is

caused by the modelling of respiration (see Section 4).

Fig. 8 shows again that modelled night-time CO2

release from respiration is considerably larger than

measured night-time CO2 flux, particularly at Bray.

t ecosystem production (NEP) ensembles for (a) Bray, (b) Flakaliden

-hour with reliable CO2 flux data. The figure also shows (on right)

nd the dashed lines are 1:1 lines. ME is the model efficiency. Data for

plify comparison (see Fig. 1).

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 113

Table 3

Model analysis of effects of individual parameters on components of NEP. The analysis begins with the Bray simulation. One parameter or set of

parameters is changed at a time to the Flakaliden simulation, then to the Griffin simulation, and back to the Bray simulation. The column DNEP

gives the change in NEP due to the change in the corresponding parameter(s). Values in bold give the total change of NEP due to the change in a

block of related parameters. Components of NEP listed are GPP (gross primary production), Rm (above-ground maintenance respiration), Rg

(above-ground growth respiration), Rsoil (soil respiration) and RE (total ecosystem respiration = Rm + Rg + Rsoil). Also included are APAR

(absorbed photosynthetically active radiation,) and LUE (light use efficiency = GPP/APAR)

APAR

(MJ m�2

yr�1)

LUE

(g C MJ�1)

GPP

(g C m�2

yr�1)

Rm

(g C m�2

yr�1)

Rg

(g C m�2

yr�1)

Rsoil

(g C m�2

yr�1)

RE

(g C m�2

yr�1)

NEP

(g C m�2

yr�1)

DNEP

(g C m�2

yr�1)

(a) Bray to Flakaliden

Bray 1656 1.16 1926 872 116 664 1652 274

SBray understorey 1278 1.37 1752 839 97 664 1600 152 �122

+Flakaliden climate 75Soil water 1278 1.43 1826 839 97 664 1600 226 74

PAR 784 1.52 1193 839 97 664 1600 �407 �633

T, D + other climate 756 1.34 1012 339 97 190 626 386 788

Overwintering 756 1.13 854 339 97 190 626 228 �158

+Flakaliden respiration rates 756 1.13 854 297 97 569 963 �109 �336

+Flakaliden canopy structure 118Leaf area index 708 1.11 783 250 72 569 891 �108 1

Biomass + other structure 769 1.09 840 206 47 569 822 18 126

Clumping 720 1.12 810 206 47 569 822 �12 �30

Leaf area density 746 1.11 830 206 47 569 822 9 20

+Flakaliden physiology �31Vcmax, Jmax 746 0.81 605 206 47 569 822 �217 �226

Temperature dependence

of photosynthesis

746 0.94 698 206 47 569 822 �124 93

a, u 746 1.14 847 206 47 569 822 26 150

Stomatal conductance 746 1.07 799 206 47 569 822 �22 �48

+Flakaliden understorey 1026 0.95 979 295 47 569 911 68 90

(b) Flakaliden to Griffin

Flakaliden 1026 0.95 979 295 47 569 911 68

SFlakaliden understorey 746 1.07 800 206 47 569 822 �22 �90

+Griffin climate 22�Overwintering 746 1.35 1006 206 47 569 822 185 207

PAR 697 1.39 972 209 47 569 826 146 �39

T, D + other climate 690 1.68 1158 254 47 857 1158 0 �146

+Griffin respiration rates 690 1.68 1156 233 47 449 729 427 427

+Griffin canopy structure 244Leaf area index 891 2.12 1890 525 56 449 1029 860 434

Biomass + other structure 1018 1.99 2024 627 138 449 1213 811 �49

Clumping 1029 1.94 1996 627 138 449 1213 783 �28

Leaf area density 974 1.93 1884 627 138 449 1213 670 �112

+Griffin physiology �52Vcmax, Jmax 974 2.07 2019 627 138 449 1213 806 135

Temperature dependence

of photosynthesis

974 1.85 1798 627 138 449 1213 585 �221

Stomatal conductance 976 1.88 1831 627 138 449 1213 618 34

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124114

Table 3 (Continued )

APAR

(MJ m�2

yr�1)

LUE

(g C MJ�1)

GPP

(g C m�2

yr�1)

Rm

(g C m�2

yr�1)

Rg

(g C m�2

yr�1)

Rsoil

(g C m�2

yr�1)

RE

(g C m�2

yr�1)

NEP

(g C m�2

yr�1)

DNEP

(g C m�2

yr�1)

(c) Griffin to Bray

Griffin 976 1.88 1831 627 138 449 1213 618

+Bray climate �1705PAR 1865 1.23 2301 636 138 449 1223 1078 460

T, D + other climate 1846 1.35 2495 1217 138 2195 3549 �1054 �2131

Soil water 1846 1.33 2462 1217 138 2195 3549 �1087 �33

+Bray respiration rates 1846 1.33 2447 964 138 664 1766 681 1767

+Bray canopy structure �629Leaf area index 1502 1.13 1698 508 154 664 1326 371 �309

Biomass + other structure 1149 1.21 1392 839 97 664 1600 �208 �580

Clumping 1174 1.17 1375 839 97 664 1600 �225 �17

Leaf area density 1286 1.28 1652 839 97 664 1600 52 277

+Bray physiology 100Vcmax, Jmax 1286 1.59 2049 839 97 664 1600 448 396

Temperature dependence

of photosynthesis

1286 1.59 2048 839 97 664 1600 447 �1

a, u 1286 1.26 1624 839 97 664 1600 23 �424

Stomatal conductance 1278 1.37 1752 839 97 664 1600 152 129

+Bray understorey 1656 1.16 1926 872 116 664 1652 274 122

Annual totals of NEP can also be compared.

Modelled annual NEP for each site is given in Table 3

(first row of each section). These values can be

compared with the NEP estimated from flux data given

in Table 1. Modelled NEP overestimates the value

given for Flakaliden and underestimates that for Bray,

but the relative NEP of the three sites remains the

same.

In summary, model agreement with measurements

is satisfactory. This comparison cannot be used to say

that the model is correct, but does show that it can

reproduce the major features of the dataset, giving

some confidence in the realism of the model and its

suitability for our purpose.

3.2. Analysis of NEP

The results of the analysis of the difference in NEP

among sites are given in Table 3. In this analysis,

parameter sets were changed one at a time and the

effects on GPP, RE and NEP are recorded. The aim is to

show the relative importance of each parameter. There

is substantial uncertainty attached each value in Table 3,

owing to the uncertainty in parameter values and

model structure. These uncertainties are extremely

difficult to quantify. However, as a rough rule of thumb

based on considerable experience with the model,

we suggest that a change greater than 100 g C m�2 yr�1

(i.e., approx. 20% of NEP at the Bray site) can be

regarded as significant.

Beginning with the Bray site (Table 3a), the

contribution of the understorey was first removed,

reducing modelled NEP by 122 g C m�2 yr�1. The

climate was changed to that of the Flakaliden site in

four steps: first the soil water content, then the PAR,

thirdly all other climatic factors, primarily air and soil

temperature and relative humidity, and finally the

winter conditions preventing photosynthesis at Fla-

kaliden were taken into account. Low soil water

content at Bray was estimated to reduce NEP by

74 g C m�2 yr�1. Reducing the PAR to that of

Flakaliden had a dramatic effect on GPP, reducing

it by 633 g C m�2 yr�1, but reducing the temperature

had an even more dramatic effect on respiration,

reducing it by 968 g C m�2 yr�1. Winter conditions at

Flakaliden also served to reduce GPP significantly.

However, the effect on NEP of changing climate alone

from Bray to Flakaliden was quite small, an increase

of just 75 g C m�2 yr�1. Changing respiration rates

from Bray to Flakaliden caused a large decrease in

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 115

NEP, of 336 g C m�2 yr�1, largely due to the increase

in soil respiration rate. Changing canopy structure

from Bray to Flakaliden served to increase NEP,

largely due to the decrease in maintenance and growth

respiration occasioned when respiring biomass was

changed from the large, tall trees at Bray to the small

ones at Flakaliden. Overall, there was little effect of

changing leaf physiology from Bray to Flakaliden,

despite the large differences in Jmax and Vcmax

parameters (see Fig. 3). The decrease in these

parameters caused a large decrease in NEP of

226 g C m�2 yr�1, but this was largely compensated

for by the change in temperature dependence of these

parameters, and the increase in the quantum yield. The

understorey at Flakaliden was estimated to add

90 g C m�2 yr�1 to the NEP.

Comparing the Flakaliden and Griffin sites

(Table 3b), changing the boreal climate of Flakaliden

to the maritime climate at Griffin had almost no effect

on NEP overall. The effect of PAR on GPP was small,

and although GPP was increased considerably by the

removal of the winter dormancy of photosynthesis,

increasing the temperature from Flakaliden to that of

Griffin caused a compensatory increase in respiration.

As with the Bray–Flakaliden comparison, a large

change in NEP was seen when respiration rates were

changed, because of the large difference in soil

respiration rate. Changing the canopy structure from

Flakaliden to Griffin also affected NEP considerably,

with an increase of 244 g C m�2 yr�1. The increase in

LAI had a large impact on GPP, as did the increased

canopy area. Changing the leaf physiology parameters

had a relatively small overall effect on NEP.

Comparing Griffin and Bray (Table 3c), the change

in climate decreased NEP dramatically. The increase

in PAR and temperature increased GPP by

664 g C m�2 yr�1, but maintenance and soil respira-

tion also increased, by 2326 g C m�2 yr�1, far out-

weighing the increase in GPP. Changing respiration

rates also had a major effect: maintenance respiration

was decreased by 252 g C m�2 yr�1 and soil respira-

tion by 1531 g C m�2 yr�1. The change in canopy

structure also decreased NEP significantly. Auto-

trophic respiration decreased considerably, because of

the smaller LAI at Bray, but the low LAI and canopy

area decreased GPP further. As with the earlier

comparisons, leaf physiology had a relatively small

overall effect on NEP. There was a large increase in

NEP due to the higher Jmax and Vcmax, but this increase

was counterbalanced by the effect of a lower quantum

yield.

Both GPP and RE thus varied considerably among

the sites, and the sensitivity analysis shown in Table 3

helps us to understand the reasons for this variation.

GPP was highest at Bray, slightly lower at Griffin, and

considerably lower at Flakaliden. Bray has the highest

GPP because it has the highest incident PAR and

temperature and the most productive leaf physiology.

Griffin has the next largest GPP because of its large

LAI; neither climate nor leaf physiology cause much

difference in GPP between Flakaliden and Griffin.

Above-ground autotrophic respiration is similar at the

Bray and Griffin sites, despite the much larger stem

biomass at Bray, because of the large leaf biomass at

Griffin. At Flakaliden, total biomass is low, so above-

ground respiration is low. Soil respiration was highest

at Bray, slightly lower at Flakaliden, and lower again

at Griffin, and these differences are due to a

combination of soil temperature, which is highest at

Bray, and the relationship between soil respiration and

temperature, which is highest at Flakaliden. We thus

see that climate alone does not determine NEP; LAI,

soil respiration, and respiring biomass are also key

determinants.

4. Discussion

The number of eddy flux sites measuring NEP of

forests and other ecosystems around the world is

growing rapidly, and many estimates of annual NEP

are now available from these data (Baldocchi et al.,

2001). However, we are limited in our ability to

generalise from these experimental sites because the

analysis of the differences among sites is still in its

early stages (IPCC, 2003). This analysis is complex

because of the multiplicity of factors that affect NEP.

Straightforward comparison of flux datasets shows

only the gross differences among ecosystems, whereas

we need to understand the importance of individual

factors in determining NEP. As most inter-site

comparisons involve many confounded factors, the

logical way to separate out their influence is to use a

model. A modelling approach has been taken

previously by Law et al. (2001), who used the model

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124116

3PG to separate out the effects of LAI and climate on

the NEP of young and old stands of ponderosa pine.

Here, we used the model MAESTRA to separate

out the effects of climate, stand structure and

physiology on NEP of three coniferous forest stands

growing in contrasting climates. The results of the

analysis show clearly why annual NEP does not vary

straightforwardly among sites. Climate alone has

several different effects: an increase in PAR increases

GPP, but the concomitant increase in temperature also

has a large effect on respiration and may or may not

increase GPP, depending on whether low winter

temperatures restrict photosynthesis, or whether high

summer temperatures result in drought. Soil respira-

tion rates also differed considerably among sites, and

these had a huge effect on modelled NEP. Leaf area

index and respiring biomass were also found to have

substantial effects on NEP. Individual aspects of leaf

physiology could influence NEP significantly, but,

interestingly, changing the suite of leaf physiology

parameters from one site to another suggested that the

overall effect of leaf physiology was small.

For this analysis to be credible, we must have

confidence in the model used, and in the assumed

differences in parameters. We now consider each of

these in turn.

4.1. Model validity

We believe that the model is appropriate for this

analysis because it embodies current best understanding

of forest gas exchange and because it can reproduce the

main features of the datasets. The MAESTRA model

has a long pedigree: it was developed from the

MAESTRO model (Wang and Jarvis, 1990a), which

was first published in 1990 and has been successfully

applied in a wide range of contexts since then (Medlyn,

2004). The model is very similar in structure to many

other models applied to eddy covariance data (e.g.

Baldocchi and Wilson, 2001) and it seems likely that

most current models would yield similar conclusions to

this study. However, it is important to be aware that

there is considerable uncertainty in model structure,

parameter values, and the data (for more detail, see

Medlyn et al., 2005). To account for this uncertainty, we

suggest that only relatively large shifts in model outputs

(changes greater than 100 g C m�2) should be treated as

significant.

Although comparison with the flux data cannot

prove the model to be correct, it shows that the model

can, in general, adequately reproduce the datasets.

Fluxes at the Griffin site were simulated well

throughout the year. At Flakaliden, there was a slight

overestimation (by 9%) of daily CO2 uptake through

the year. This error is within the range of uncertainty

of eddy flux data (Anthoni et al., 1999). There is

however one major discrepancy between model output

and data at the Bray site, where CO2 uptake was

overestimated in spring and underestimated in

summer. This problem may be related to the way

that we model respiration. Note that in Fig. 8, where

monthly ensembles of NEP are compared, midday

NEP from flux measurements for the Bray site

increases steadily during the first half of the year,

whereas modelled midday NEP begins to decrease

after March. In the model, this decrease occurs

because, as temperature increases from spring to

summer, the respiration rate increases faster than GPP.

This effect is not seen in the flux data. Fig. 9 illustrates

the importance in the model of the temperature

dependence of respiration. Monthly ensemble NEP is

plotted against incident PAR for the months of

February and May and it can be seen that the response

of NEP to PAR is similar in both months. Adding

modelled respiration, which depends on temperature,

gives GPP. The response of GPP to PAR is seen to be

quite different between the months. The temperature

dependence of GPP is inadequate to explain this large

increase between February and May. The implication

is that the temperature dependence of respiration in the

model is too large.

There has recently been considerable debate about

the adequacy of our current ability to model

respiration (Dewar et al., 1999; Cannell and Thornley,

2000; Gifford, 2003). Several authors have suggested

that, instead of a temperature-driven growth-main-

tenance paradigm, plant respiration should instead be

linked to photosynthetic production (Dewar et al.,

1999; Gifford, 2003; Whitehead et al., 2004). Our

results provide further evidence that the standard

respiration model is inadequate. We attempted to test

the respiration component of the model against flux

data using night-time CO2 efflux (Fig. 6). The

modelled respiration tended to overestimate the flux

data, as has been found by a number of other studies

(Lavigne et al., 1997; Law et al., 1999). Current

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 117

Fig. 9. (a) Half-hourly averages of net ecosystem production (NEP) from flux measurements vs. half-hourly averages of incident photo-

synthetically active radiation (PAR) for the Bray site in February (solid circles) and May (open squares). (b) Half-hourly averages of gross

primary productivity (GPP), calculated by summing NEP and modelled respiration, vs. half-hourly averages of incident PAR for the Bray site in

February (solid circles) and May (open squares).

thinking is that eddy covariance underestimates night-

time fluxes, particularly in stable conditions (Aubinet

et al., 2003) and hence the discrepancy shown in Fig. 6

does not necessarily indicate a problem with the

model. Reliable estimates of night-time CO2 flux

would be invaluable in aiding us to develop better

respiration models.

4.2. Major differences among sites

4.2.1. Soil respiration

Differences in the rate of soil respiration at a given

soil temperature were one of the most important

influences on NEP. The responses used in the model

were very different (Fig. 5), with respiration rate at a

given soil temperature being generally highest for

Flakaliden and lowest for Bray. Janssens et al. (2001)

attributed the high soil CO2 efflux at Griffin and

Flakaliden to the effects of disturbance at plantation

establishment (ploughing and fertilisation at Griffin;

clear-felling, prescribed burning, and soil scarification

at Flakaliden), while the low soil CO2 efflux at Bray

was attributed to low soil moisture content. These

explanations are clearly speculative.

A possible alternative hypothesis is acclimation of

respiration to soil temperature. Fig. 5 illustrates that

soil respiration at the average soil temperature for each

site is similar, a result which is consistent with the

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124118

outcome of a soil warming experiment at Flakaliden

(Stromgren, 2001). After 4 years of warming, the soil

respiration rate at the increased temperature in the

warmed plot was similar to the rate at ambient

temperature in the unwarmed plot. However, acclima-

tion to soil warming may have different causes than

acclimation to soil temperature in different climates.

In the soil warming experiment, a major hypothesis for

the acclimation was a reduction in substrate avail-

ability for heterotrophic respiration (Eliasson et al.,

2005), whereas a comparison of soil carbon pools

across the three sites considered here suggests that

substrate availability may have been lowest in the

coldest site, Flakaliden. Soil carbon in the top 50 cm

(including litter layer) at Flakaliden was estimated at

4.5 kg C m�2 (Andersson et al., 2002), whereas it was

11 kg C m�2 in the top 30 cm at Griffin (Wingate,

2003) and 10.5 kg C m�2 in the top 30 cm at Bray (D.

Loustau, personal communication).

The methodology used to measure soil respiration

at the three sites differed slightly, which may also have

contributed to the difference in measured soil

respiration rates. It is known that soil respiration

rates are quite sensitive to the measurement system.

Comparisons of measurement systems by Norman

et al. (1997) and Janssens et al. (2000) suggest that the

system used at the Griffin site (CIRAS-1 closed

dynamic system) may give higher respiration rates

than the open dynamic systems used at the other sites.

However, the differences in respiration found by those

comparisons are considerably smaller (10–46%,

Janssens et al., 2000) than the differences among

sites found here (3–10 times larger at Flakaliden and

Griffin than at Bray). There have been some additional

measurements at the Bray site yielding higher rates of

soil respiration (D. Loustau, personal communica-

tion), but using these rates makes it impossible to close

the carbon balance at this site. In summary, the

evidence is good for quite different soil respiration

rates at the three sites, but we do not completely

understand why these rates differ.

4.2.2. Above-ground respiration

Both respiring biomass and maintenance respira-

tion rates have moderate impacts on NEP (Table 3).

There are clear differences in woody biomass among

the three sites, which are substantiated by inventories:

woody biomass is more than three times greater at

Bray than at Flakaliden (Table 2). Biomass is

evidently a function of both plantation age and growth

rate. It is interesting that woody maintenance

respiration rates are also considerably larger for Bray

than for Flakaliden (Fig. 4), even when segment

diameter is taken into account. We hypothesise that the

maintenance respiration rate is larger at Bray because

of increased cellular activity, i.e. increased growth

rate. This difference also lends support to the

argument that respiration rate should be linked to

photosynthesis, discussed above. This hypothesis

should be explored further, carefully accounting for

within-tree variation in rates.

4.2.3. Total respiration

Previous analyses comparing these sites have

stressed the importance of respiration in determining

NEP (Valentini et al., 2000; Janssens et al., 2001). This

study confirms the importance of respiration but there

are some points of disagreement with previous studies.

Valentini et al. (2000) suggested that GPP was

constant with latitude and that differences in NEP

were entirely driven by differences in ecosystem

respiration. In contrast, the current analysis suggests

large differences in GPP among sites, largely driven

by the considerable differences in climate, with

secondary effects of needle photosynthetic rates and

leaf area index. The study by Janssens et al. (2001)

also shows considerable variation in GPP among sites,

and they suggest that RE is related to GPP. They focus

on the role of soil respiration in explaining the high

rates of ecosystem respiration in the northern sites,

citing the effect of ploughing at establishment to

explain high RE at the Griffin site. Here, however, the

estimated ratio of RE:GPP was similar for Bray and

Flakaliden (0.86 and 0.93, respectively), it was

considerably lower for Griffin (0.66). Soil respiration

was actually quite low at Griffin; plant respiration was

relatively high owing to the large LAI, but the LAI

also resulted in a large GPP and hence a lower RE:GPP

ratio.

4.2.4. Leaf area index

Mean LAI was similar at the Bray and Flakaliden

sites but much higher at Griffin, with significant

consequences for NEP (Table 3). Light interception,

photosynthesis, and respiration were all increased, but

the overall effect was a large increase in NEP. LAI has

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 119

also been shown to be a key determinant of NEP in a

study comparing young and old ponderosa pine forests

(Anthoni et al., 2002).

The high LAI at Griffin is due to a combination of

factors including fertility, climate, and age. Bray and

Flakaliden are both relatively infertile sites and their

LAI can be increased dramatically with the regular

addition of fertiliser (Bergh et al., 1999; D. Loustau,

personal communication). The climate at Griffin is

mild, with neither the harsh winters of Flakaliden nor

the summer droughts of Bray, both of which restrict

leaf area development. Finally, Griffin forest is the

youngest of the three plantations and had only recently

reached canopy closure.

4.2.5. Canopy structure

Inspection of Table 3 shows that the leaf area

density distribution, both within crowns (leaf area

density) and the crown size (included in biomass +

other structure) has a significant effect on light

interception, GPP and thus NEP. The importance of

leaf area density for GPP and NEP is explored further

by Wang and Jarvis (1990b). The clumping of foliage

into shoots, in contrast, has almost no effect on NEP.

The parameter describing this clumping was identified

by Baldocchi and Wilson (2001) as a key parameter

for deciduous trees. However, here it varied only

slightly among sites and indeed the difference was

unlikely to be significant, explaining the low

sensitivity of the analysis to this parameter.

4.2.6. Leaf physiology

There appeared to be significant differences among

species in foliar physiology. There were large

differences in A–Ci responses (Fig. 2), in the

relationship between stomatal conductance and

photosynthetic rate (Fig. 3), and in the quantum yield

of electron transport.

The difference in A–Ci responses and hence the

photosynthetic parameters Jmax and Vcmax may have

been in part mediated by needle nitrogen content. The

nitrogen content of the needles measured for Fig. 2

was 1.8, 2.0 and 3.7 g m�2 on a half-total leaf surface

area basis for P. sitchensis, P. abies, and P. pinaster,

respectively. We note that it is also possible that

differences in experimental procedure such as lighting

and needle arrangement contributed to the difference

in responses. However, the differences in the stomatal

conductance relationships followed the same pattern

among species as the A–Ci responses, with the slope of

stomatal conductance versus photosynthesis being

largest for P. pinaster and smallest for P. abies (Fig. 3).

As photosynthesis is included on the x-axis, metho-

dological differences in lighting and needle arrange-

ment are less likely to be of importance here. There are

few other comparative studies of how the relationship

shown in Fig. 3 varies among species; it would be

interesting to know if the relative rankings shown in

Figs. 2 and 3 generalise to other species.

The evidence for the differences in the quantum

yield parameter is, unfortunately, quite flimsy,

particularly as quantum yield is a highly sensitive

parameter in the model (Medlyn et al., 2005). Earlier

versions of the model assumed that this parameter was

fixed and equal to its theoretical value of 0.425 (Evans

and Farquhar, 1991), but comparisons of model output

with eddy flux data showed that this value was clearly

too high (Ibrom et al., unpublished manuscript). It is

not altogether clear why this parameter is less than its

theoretical value, nor how it should be measured at

leaf scale. The two parameter values used here were

obtained by (i) fitting the parameter to light-response

curves (Porte and Loustau, 1998) and (ii) estimating

the parameter from previous measurements of needle

gas exchange response to incident radiation (Leverenz

and Jarvis, 1979; Ibrom et al., unpublished manu-

script). Neither set of measurements was made at a

saturating CO2 concentration. In the first case the

curvature parameter u was also fitted to data (and was

highly correlated with a), while in the second case a

fixed value of u was assumed. Because of these

differences in methodology, it is uncertain whether the

difference in quantum yield, although considerable

(0.2 versus 0.3), is genuine. There are also few other

estimates in the literature that could be used to

compare these values against. Better quantification of

this parameter is needed.

Interestingly, despite the large differences in

parameterisations for needle physiology, the overall

effect of physiology on NEP was quite small.

Changing individual parameters gave large changes

in NEP, but the changes due to different parameters

appeared to compensate, so the overall effect of

changing the total suite of physiological parameters

from one species to another was small. This result

poses the question of whether the small effect of

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124120

physiology is coincidental in these cases, or whether

the physiological parameters are always correlated in

such a way as to have little overall effect on NEP. The

modelling study by McMurtrie et al. (1994), in which

conifer GPP was found to be roughly proportional to

utilisable absorbed PAR in different climates, suggests

that physiology does not play a large role in

determining GPP. A similar conclusion is suggested

by the work of Buchmann and Schulze (1999), who

found that the mean maximum rate of light-saturated

assimilation in eddy flux studies on conifers increased

consistently with increasing leaf area index. However,

further research is required to confirm this conclusion,

and also, if found to be correct, to answer the question

of why significant leaf-level differences in physiology

are compensatory at the canopy level.

4.2.7. Climate

The finding that climate has several opposing

effects on NEP echoes the results of many other

studies. For example, Law et al. (2002), in a

comparative study of 34 eddy flux sites, showed that

GEP was related to site water balance and tempera-

ture, but that RE was poorly correlated with

temperature across sites, with the result that the only

discernible influence of climate on NEP was a weak

correlation with site water balance. Barr et al. (2002)

also illustrate how temperature does not have a

consistent effect on NEP. They showed that years with

a warm spring in a temperate deciduous forest had

reduced NEP because of increased respiration and

drought stress, while NEP was increased by warm

springs in a boreal deciduous forest owing to earlier

leaf-out. The likelihood that differences in LAI are

related to climate adds yet another dimension to the

complex response of NEP to climate.

4.3. Further research needs

This study indicates that considerable further

research is needed before we will be able to

extrapolate from current eddy flux studies to reliably

predict NEP for other forest ecosystems.

We would argue that one important step would be

to analyse data from other flux sites using the detailed,

model-based methodology used here. This methodol-

ogy allows us to separate out the individual effects of

climate, structure and physiology on NEP for any

given site. Detailed characterisations of structure and

physiology are available for many sites but are often

not linked quantitatively to NEP.

Soil respiration represents another important

knowledge gap. The analysis presented here shows

that soil respiration rates are one of the most important

controls on NEP. However, they can at present only be

modelled using site-specific empirical relationships

and we have little idea of how to generalise from these

relationships to other sites. Process-based models of

soil respiration rates are urgently required.

We also show that modelling of plant respiration is

problematic. There is disagreement about whether the

growth-maintenance paradigm of plant respiration

(Amthor, 2000) is adequate to model annual respira-

tion rates, or whether respiration rates should be linked

to photosynthetic uptake on this timescale (Gifford,

2003). Night-time eddy flux data should be of use in

enabling us to quantify ecosystem respiration more

accurately, but unfortunately these data are affected by

high uncertainty, particularly when turbulence is low

(Massman and Lee, 2002; Baldocchi, 2003). Reducing

the uncertainty in night-time fluxes would aid

significantly in developing better models of ecosystem

respiration.

Finally, the exercise of comparing physiological

parameters across sites has highlighted our poor

understanding of how some of these parameters (such

as stomatal conductance, quantum yield, and respira-

tion rates) vary among different forest systems. In

these comparisons, confounding due to methodologi-

cal differences was a serious problem. Further

comparative studies of these parameters, that aim to

link variation in parameters with factors such as

climate, species, and nutrition, would be very helpful

in our attempts to generalise. However, the use of

comparable methodologies in such studies is essential.

5. Conclusions

This study focussed on three forest sites whose

carbon balance, or NEP, was shown to differ using

eddy flux measurements. A modelling analysis was

used to identify the main factors causing the difference

in NEP. Differences in climate affected both carbon

uptake and release but could not alone account for the

differences in NEP. Differences in soil respiration

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 121

rates were found to be extremely important, but their

variation among sites was difficult to explain.

Differences in plant respiration rates, respiring

biomass, and leaf area index also contributed

significantly to the difference in NEP among sites.

If we are to generalise from flux studies to other forest

patches, each of these factors needs to be quantified.

Acknowledgements

BM was supported by an Australia Research

Council Post Doctoral Fellowship. RM was supported

by an Australia Research Council Discovery Grant.

We are very grateful to Johan Bergh, Eric Ceschia,

Michael Freeman, Peter Roberntz, Monika Stromgren,

Jan Stockfors, Britta Widen, and Jerry Skoglund and

his undergraduate students (Sweden), Alexandre

Bosc, Sylvain Delzon, and Annabel Porte (France),

and Craig Barton and Patrick Meir (U.K.) for

supplying data used in parameterisation. Thanks to

Ying-Ping Wang, as ever, for supplying the original

MAESTRO model.

References

Amthor, J.S., 2000. The McCree–de Wit–Penning de Vries–Thorn-

ley respiration paradigms: 30 years later. Ann. Bot. 86, 1–20.

Andersson, P., Berggren, D., Nilsson, I., 2002. Indices for nitrogen

status and nitrate leaching from soils under Norway spruce

(Picea abies (L.) Karst) stands in Sweden. For. Ecol. Manage.

157, 39–53.

Andreae, M.O., Artaxo, P., Brandao, C., Carswell, F.E., Ciccioli, P.,

da Costa, A.L., Culf, A.D., Esteves, J.L., Gash, J.H.C., Grace, J.,

Kabat, P., Lelieveld, J., Malhi, Y., Manzi, A.O., Meixner, F.X.,

Nobre, A.D., Nobre, C., Ruivo, M.d.L.P., Silva-Dias, M.A.,

Stefani, P., Valentini, R., Jouanne, J., von Waterloo, M.J.,

2002. Biogeochemical cycling of carbon, water, energy, trace

gases, and aerosols in Amazonia: the LBA-EUSTACH experi-

ments. J. Geophys. Res. 107 (D20). doi:10.1029/2001JD000524.

Anthoni, P.M., Law, B.E., Unsworth, M.H., 1999. Carbon and water

vapor exchange of an open-canopied Ponderosa pine ecosystem.

Agric. For. Meteorol. 95, 151–168.

Anthoni, P.M., Unsworth, M.H., Law, B.E., Irvine, J., Baldocchi,

D.D., Van Tuyl, S., Moore, D., 2002. Seasonal differences in

carbon and water vapor exchange in young and old-growth

ponderosa pine ecosystems. Agric. For. Meteorol. 111, 203–222.

Arain, M.A., Black, T.A., Barr, A.G., Jarvis, P.G., Massheder, J.M.,

Verseghy, D.L., Nesic, Z., 2002. Effects of seasonal and inter-

annual climate variability on net ecosystem productivity of

boreal deciduous and conifer forests. Can. J. For. Res. 32,

878–891.

Aubinet, M., Grelle, A., Ibrom, A., Rannik, U., Moncrieff, J., Foken,

T., Kowalski, A.S., Martin, P.H., Berbigier, P., Bernhofer, C.,

Clement, R., Elbers, J., Granier, A., Grunwald, T., Morgenstern,

K., Pilegaard, K., Rebmann, C., Snijders, W., Valentini, R.,

Vesala, T., 2000. Estimates of the annual net carbon and water

exchange of forests: the EUROFLUX methodology. Adv. Ecol.

Res. 30, 113–175.

Aubinet, M., Clement, R., Elbers, J.A., Folken, T., Grelle, A., Ibrom,

A., Moncrieff, J., Pilegaard, K., Rannik, U., Rebmann, C., 2003.

Methodology for data acquisition, storage and treatment. In:

Valentini, R. (Ed.), Fluxes of Carbon, Water and Energy of

European Forests. Springer-Verlag, Berlin, Heidelberg, pp. 9–

35.

Baldocchi, D.D., 2003. Assessing the eddy covariance technique for

evaluating carbon dioxide exchange rates of ecosystems: past,

present and future. Glob. Change Biol. 9, 479–492.

Baldocchi, D.D., Wilson, K.B., 2001. Modeling CO2 and water

vapor exchange of a temperate broadleaved forest across hourly

to decadal time scales. Ecol. Modell. 142, 155–184.

Baldocchi, D., Falge, E., Gu, L.H., Olson, R., Hollinger, D., Run-

ning, S., Anthoni, P., Bernhofer, C., Davis, K., Evans, R.,

Fuentes, J., Goldstein, A., Katul, G., Law, B., Lee, X.H., Malhi,

Y., Meyers, T., Munger, W., Oechel, W., Paw, U.K.T., Pilegaard,

K., Schmid, H.P., Valentini, R., Verma, S., Vesala, T., Wilson, K.,

Wofsy, S., 2001. FLUXNET: a new tool to study the temporal

and spatial variability of ecosystem-scale carbon dioxide, water

vapor, and energy flux densities. Bull. Am. Meteorol. Soc. 82,

2415–2434.

Ball, J.T., Woodrow, I.E., Berry, J.A., 1987. A model predicting

stomatal conductance and its contribution to the control of

photosynthesis under different environmental conditions. In:

Biggins, J. (Ed.), Progress in Photosynthesis Research. Marti-

nus-Nijhoff Publishers. Dordrecht, The Netherlands, pp. 221–

224.

Barclay, H.J., Goodman, D., 2000. Conversion of total to projected

leaf area index in conifers. Can. J. Bot. 78, 447–454.

Barford, C., Wofsy, S.C., Goulden, M.L., Munger, J.W., Pyle, E.H.,

Urbanski, S., Hutyra, L., Saleska, S.R., Fitzjarrald, D., Moore,

K., 2001. Factors controlling long- and short-term sequestration

of atmospheric CO2 in a mid-latitude forest. Science 294, 1688–

1691.

Barr, A.G., Griffis, T.J., Black, T.A., Lee, X.H., Staebler, R.M.,

Fuentes, J.D., Chen, Z., Morgenstern, K., 2002. Comparing the

carbon budgets of boreal and temperate deciduous forest stands.

Can. J. For. Res. 32, 813–822.

Barton, C.V.M., 1997. Effects of elevated atmospheric carbon

dioxide concentration on growth and physiology of Sitka spruce

(Picea sitchensis (Bong.) Carr.). Ph.D. Edinburgh University, p.

203.

Barton, C.V.M., Jarvis, P.G., 1999. Growth response of branches of

Picea sitchensis to four years exposure to elevated atmospheric

carbon dioxide concentration. New Phytol. 144, 233–243.

Berbigier, P., Bonnefond, J.M., 1995. Measurement and modelling

of radiation transmission within a stand of maritime pine (Pinus

pinaster Ait.). Ann. Sci. For. 52, 23–42.

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124122

Berbigier, P., Bonnefond, J.M., Mellmann, P., 2001. CO2 and water

vapour fluxes for 2 years above Euroflux forest site. Agric. For.

Meteorol. 108, 183–197.

Bergh, J., McMurtrie, R.E., Linder, S., 1998. Climatic factors

controlling the productivity of Norway spruce: a model-based

analysis. For. Ecol. Manage. 110, 127–139.

Bergh, J., Linder, S., Lundmark, T., Elfving, B., 1999. The effect of

water and nutrient availability on the productivity of Norway

spruce in northern and southern Sweden. For. Ecol. Manage.

119, 51–62.

Bosc, A., 1999. Etude experimentale du fonctionnement hydrique et

carbone des organes aeriens du Pin maritime (Pinus pinaster

Ait.). Ph.D. Universite Victor Segalen Bordeaux 2.

Bosc, A., De Grandcourt, A., Loustau, D., 2003. Variability of stem

and branch maintenance respiration in a Pinus pinaster tree. Tree

Physiol. 23, 227–236.

Brooks, A., Farquhar, G.D., 1985. Effect of temperature on the CO2/

O2 specificity of ribulose-1,5-bisphosphate carboxylase/oxyge-

nase and the rate of respiration in the light. Planta 165, 397–406.

Buchmann, N., Schulze, E.D., 1999. Net CO2 and H2O fluxes of

terrestrial ecosystems. Glob. Biogeochem. Cycles 13, 751–760.

Cannell, M.G.R., Thornley, J.M.H., 2000. Modelling the compo-

nents of plant respiration: some guiding principles. Ann. Bot. 85,

45–54.

Ceschia, E., 2001. Environmental effects on spatial and seasonal

variations of stem respiration in European beech and Norway

spruce. Doctoral Thesis. Acta Univ. Agric. Suec., Silvestria 219,

56 pp.

Chung, H.H., Barnes, R.L., 1977. Photosynthate allocation in Pinus

taeda. I. Substrate requirements for synthesis of shoot biomass.

Can. J. For. Res. 7, 106–111.

Clark, K.L., Gholz, H.L., Moncrieff, J.B., Cropley, F., Loescher,

H.W., 1999. Environmental controls over net exchanges of

carbon dioxide from contrasting Florida ecosystems. Ecol. Appl.

9, 936–948.

Clement, R., Moncrieff, J.B., Jarvis, P.G., 2003. Net carbon pro-

ductivity of Sitka spruce forest in Scotland. Scottish Forestry 57,

5–10.

Clement, R., 2004. Mass and energy exchange of a plantation forest

in Scotland using micrometeorological methods. Ph.D. Univer-

sity of Edinburgh.

Delzon, S., 2000. Photosynthese de la Molinie (Molinia coerulea

(L.) Moench) en sous-bois de Pin maritime, aux niveaux de la

feuille et du couvert. D.E.A. Universite d’Orsay-Paris XI, 35 pp.

Dewar, R.C., Medlyn, B.E., McMurtrie, R.E., 1999. Acclimation of

the respiration photosynthesis ratio to temperature: insights from

a model. Glob. Change Biol. 5, 615–622.

Eliasson, P.E., McMurtrie, R.E., Pepper, D.A., Stromgren, M.,

Linder, S., Agren, G.I., 2005. The response of heterotrophic

CO2 flux to soil warming. Glob. Change Biol. 11, 167–181.

Evans, J.R., Farquhar, G.D., 1991. Modeling canopy photosynthesis

from biochemistry of the C3 chloroplast. In: Boote, K.J.,Loomis,

R.S. (Eds.), Modeling Crop Photosynthesis—From Biochemis-

try to Canopy. Crop Science Society of America, Madison, WI,

pp. 1–17.

Falge, E., Baldocchi, D., Olson, R., Anthoni, P., Aubinet, M.,

Bernhofer, C., Burba, G., Ceulemans, R., Clement, R., Dolman,

H., Granier, A., Gross, P., Grunwald, T., Hollinger, D., Jensen,

N.O., Katul, G., Keronen, P., Kowalski, A.S., Lai, C.T., Law,

B.E., Meyers, T., Moncrieff, J., Moors, E., Munger, J.W.,

Pilegaard, K., Rannik, U., Rebmann, C., Suyker, A., Tenhunen,

J., Tu, K., Verma, S., Vesala, T., Wilson, K., Wofsy, S., 2001. Gap

filling strategies for defensible annual sums of net ecosystem

exchange. Agric. For. Meteorol. 107, 43–69.

Farquhar, G.D., Von Caemmerer, S., Berry, J.A., 1980. A biochem-

ical model of photosynthetic carbon dioxide assimilation in

leaves of 3-carbon pathway species. Planta 149, 78–90.

Gifford, R.M., 2003. Plant respiration in productivity models:

conceptualisation, representation and issues for global terrestrial

carbon-cycle research. Funct. Plant Biol. 30, 171–186.

Granier, A., Loustau, D., 1994. Measuring and modelling the

transpiration of a maritime pine canopy from sap-flow data.

Agric. For. Meteorol. 71, 61–81.

Griffis, T.J., Black, T.A., Morgenstern, K., Barr, A.G., Nesic, Z.,

Drewitt, G.B., Gaumont-Guay, D., McCaughey, J.H., 2003.

Ecophysiological controls on the carbon balances of three

southern boreal forests. Agric. For. Meteorol. 117, 53–71.

IPCC, 2003. IPCC meeting on current scientific understanding of the

processes affecting terrestrial carbon stocks and human influ-

ences upon them. In: Schimel, D., Manning, M. (Eds.), IPCC

Expert Meeting Report. IPCC Secretariat, Geneva, Switzerland,

p. 34.

Janssens, I.A., Kowalski, A., Longdoz, B., Ceulemans, R., 2000.

Assessing forest soil CO2 efflux: an in situ comparison of four

techniques. Tree Physiol. 20, 23–32.

Janssens, I.A., Lankreijer, H., Matteucci, G., Kowalski, A.S., Buch-

mann, N., Epron, D., Pilegaard, K., Kutsch, W., Longdoz, B.,

Grunwald, T., Montagnani, L., Dore, S., Rebmann, C., Moors,

E.J., Grelle, A., Rannik, U., Morgenstern, K., Berbigier, P.,

Ibrom, A., Moncrieff, J., Aubinet, M., Bernhofer, C., Jensen,

N.O., Vesala, T., Granier, A., Schulze, E.-D., Lindroth, A.,

Dolman, A.J., Jarvis, P.G., Ceulemans, R., Valentini, R.,

2001. Productivity overshadows temperature in determining soil

and ecosystem respiration across European forests. Glob.

Change Biol. 7, 269–278.

Jarvis, P.G., 1976. The interpretation of the variations in leaf water

potential and stomatal conductance found in canopies in the

field. Phil. Trans. Roy. Soc. London, Ser. B 273, 593–610.

Jarvis, P.G., Leverenz, J.W., 1983. Productivity of temperate, decid-

uous and evergreen forests. In: Lange, O.L., Nobel, P.S., Os-

mond, C.B., Ziegler, H. (Eds.), Encyclopedia of Plant

Physiology, 12D Physiological Plant Ecology IV. Ecosystem

Processes: Mineral Cycling, Productivity and Man’s Influence.

Springer-Verlag, Berlin, pp. 234–280.

Lavigne, M.B., Ryan, M.G., Anderson, D.E., Baldocchi, D.D., Crill,

P.M., Fitzjarrald, D.R., Goulden, M.L., Gower, S.T., Massheder,

J.M., McCaughey, J.H., Rayment, M., Striegl, R.G., 1997.

Comparing nocturnal eddy covariance measurements to esti-

mates of ecosystem respiration made by scaling chamber mea-

surements at six coniferous boreal sites. J. Geophys. Res. 102,

28977–28985.

Law, B.E., Goldstein, A.H., Anthoni, P.M., Unsworth, M.H., Panek,

J.A., Bauer, M.R.J., Fracheboud, M., Huiltman, N., 2001. Car-

bon dioxide and water vapor exchange by young and old

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124 123

ponderosa pine ecosystems during a dry summer. Tree Physiol.

21, 299–308.

Law, B.E., Ryan, M.G., Anthoni, P.M., 1999. Seasonal and annual

respiration of a ponderosa pine ecosystem. Glob. Change Biol. 5,

169–182.

Law, B.E., Falge, E., Gu, L., Baldocchi, D.D., Bakwin, P., Berbigier,

P., Davis, K., Dolman, A.J., Falk, M., Fuentes, J.D., Goldstein,

A., Granier, A., Grelle, A., Hollinger, D., Janssens, I.A., Jarvis,

P.G., Jensen, N.O., Katul, G., Mahli, Y., Matteucci, G., Meyers,

T., Monson, R., Munger, W., Oechel, W., Olson, R., Pilegaard,

K., Paw, U.K.T., Thorgeirsson, H., Valentini, R., Verma, S.,

Vesala, T., Wilson, K., Wofsy, S., 2002. Environmental controls

over carbon dioxide and water vapor exchange of terrestrial

vegetation. Agric. For. Meteorol. 113, 97–120.

Leuning, R., 1995. A critical appraisal of a coupled stomatal-

photosynthesis model for C3 plants. Plant Cell Environ. 18,

339–357.

Leverenz, J.W., Jarvis, P.G., 1979. Photosynthesis in Sitka spruce

(Picea sitchensis (Bong.) Carr.). VIII. The effects of light flux

density and direction on the rate of net photosynthesis and the

stomatal conductance of needles. J. Appl. Ecol. 16, 919–932.

Linder, S., 1995. Foliar analysis for detecting and correcting nutrient

imbalances in Norway spruce. Ecol. Bull. (Copenhagen) 44,

178–190.

Luo, Y.Q., Wan, S.Q., Hui, D.F., Wallace, L.L., 2001. Acclimatiza-

tion of soil respiration to warming in a tall grass prairie. Nature

413, 622–625.

Malhi, Y., Baldocchi, D., Jarvis, P.G., 1999. The carbon balance of

tropical, temperate and boreal forests. Plant Cell Environ. 22,

715–740.

Massman, W.J., Lee, X., 2002. Eddy covariance flux corrections and

uncertainties in long term studies of carbon and energy

exchanges. Agric. For. Meteorol. 104, 185–198.

Mayer, D.G., Butler, D.G., 1993. Statistical validation. Ecol. Model.

68, 21–32.

McMurtrie, R.E., Gholz, H.L., Linder, S., Gower, S.T., 1994.

Climatic factors controlling the productivity of pine stands: a

model-based analysis. Ecol. Bull. (Copenhagen) 43, 173–188.

Medlyn, B.E., 2004. A MAESTRO retrospective. In: McNaughton,

K. (Ed.), Forests at the Land–Atmosphere Interface. CAB

International, pp. 105–121.

Medlyn, B.E., McMurtrie, R.E., Dewar, R.C., Jeffreys, M.P., 2000.

Soil processes dominate the long-term response of forest net

primary productivity to increased temperature and atmospheric

CO2 concentration. Can. J. For. Res. 30, 873–888.

Medlyn, B.E., Dreyer, E., Ellsworth, D., Forstreuter, M., Harley,

P.C., Kirschbaum, M.U.F., LeRoux, X., Loustau, D., Montpied,

P., Strassemeyer, J., Walcroft, A., Wang, K., 2002a. Temperature

response of parameters of a biochemically-based model of

photosynthesis. II. A review of experimental data. Plant Cell

Environ. 25, 1167–1179.

Medlyn, B.E., Loustau, D., Delzon, S., 2002b. Temperature response

of parameters of a biochemically-based model of photosynth-

esis. I. Seasonal changes in mature maritime pine (Pinus pina-

ster Ait.). Plant Cell Environ. 25, 1155–1165.

Medlyn, B.E., Robinson, A.P., Clement, R., McMurtrie, R.E., 2005.

On the validation of models of forest CO2 exchange using eddy

covariance data: some perils and pitfalls. Tree Physiol. 25, 839–

857.

Meir, P., Kruijt, B., Broadmeadow, M., Barbosa, E., Kull, O.,

Carswell, F., Nobre, A., Jarvis, P.G., 2002. Acclimation of

photosynthetic capacity to irradiance in tree canopies in relation

to leaf nitrogen concentration and leaf mass per unit area. Plant

Cell Environ. 25, 343–357.

Moren, A.-S., Lindroth, A., 2000. CO2 exchange at the floor of a

boreal forest. Agric. For. Meteorol. 101, 1–14.

Neilson, R.E., Ludlow, M.M., Jarvis, P.G., 1972. Photosynthesis of

Sitka spruce (Picea sitchensis (Bong.) Carr.). II. Response to

temperature. J. Appl. Ecol. 9, 721–745.

Norman, J.M., Jarvis, P.G., 1974. Photosynthesis in Sitka spruce

(Picea sitchensis (Bong.) Carr.). III. Measurements of canopy

structure and interception of radiation. J. Appl. Ecol. 11, 375–

398.

Norman, J.M., Kucharik, C.J., Gower, S.T., Baldocchi, D.D., Crill,

P.M., Rayment, M., Savage, K., Striegl, R.G., 1997. A compar-

ison of six methods for measuring soil-surface carbon dioxide

fluxes. J. Geophys. Res. 102, 8771–28777.

Ogee, J., Lamaud, E., Brunet, Y., Berbigier, P., Bonnefond, J.M.,

2001. A long-term study of soil heat flux under a forest canopy.

Agric. For. Meteorol. 106, 173–186.

Oker-Blom, P., Kellomaki, S., 1983. Effect of grouping foliage on

the within-stand and within-crown light regime: comparison of

random and grouping canopy models. Agric. For. Meteorol. 28,

143–155.

Pargade, J., 2000. Analyses des variations spatio-temporelles du flux

de CO2 d’un sol forestier mesure par la methode dynamique.

DEA. Universite Henri Poincare Nancy I. p30.

Pilegaard, K., Hummelshøj, P., Jensen, N.O., Chen, Z., 2001. Two

years of continuous CO2 eddy-flux measurements over a Danish

beech forest. Agric. For. Meteorol. 107, 29–41.

Porte, A., 1999. Modelisation des effets du bilan hydrique sur la

production primaire et la croissance d’un couvert de pins

maritimes (Pinus pinaster Ait.) en lande humide. Ph.D. Uni-

versite d’Orsay, Paris XI.

Porte, A., Loustau, D., 1998. Variability of the photosynthetic

characteristics of mature needles within the crown of a 25-

year-old Pinus pinaster. Tree Physiol. 18, 223–232.

Porte, A., Bosc, A., Champion, I., Loustau, D., 2000. Estimating the

foliage area of maritime pine (Pinus pinaster Ait.) branches and

crowns with application to modelling the foliage area distribu-

tion in the crown. Ann. For. Sci. 57, 73–86.

Riederer, M., Kurbasik, K., Steinbrecher, R., Voss, A., 1988. Surface

areas, lengths and volumes of Picea abies (L.) Karst needles:

determination, biological variability and effect of environmental

factors. Trees 2, 165–172.

Roberntz, P., Stockfors, J., 1998. Effects of elevated CO2 concen-

tration and nutrition on net photosynthesis, stomatal conduc-

tance and needle respiration of field-grown Norway spruce trees.

Tree Physiol. 18, 233–241.

Schulze, E.D., Wirth, C., Heimann, M., 2000. Managing forests after

Kyoto. Science 289, 2058–2059.

Sellers, P.J., Hall, F.G., Kelly, R.D., Black, A., Baldocchi, D., Berry,

J., Ryan, M., Ranson, K.J., Crill, P.M., Lettenmaier, D.P.,

Margolis, H., Cihlar, J., Newcomer, J., Fitzjarrald, D., Jarvis,

B.E. Medlyn et al. / Agricultural and Forest Meteorology 131 (2005) 97–124124

P.G., Gower, S.T., Halliwell, D., Williams, D., Goodison, B.,

Wickland, D.E., Guertin, F.E., 1997. BOREAS in 1997: experi-

ment overview, scientific results, and future directions. J. Geo-

phys. Res. 102, 28731–28769.

Stenberg, P., Linder, S., Smolander, H., 1995. Variation in the ratio

of shoot silhouette area to needle area in fertilized and unferti-

lized Norway spruce trees. Tree Physiol. 15, 705–712.

Stenberg, P., Kangas, T., Smolander, H., Linder, S., 1999. Shoot

structure, canopy openness, and light interception in Norway

spruce. Plant Cell Environ. 22, 1133–1142.

Stockfors, J., 1997. Respiratory losses in Norway spruce: the effect

of growth and nutrition. Doctoral Thesis. Acta Univ. Agric.

Suec., Silvestria 20, 41 pp.

Stockfors, J., Linder, S., 1998a. Effect of nitrogen on the seasonal

course of growth and maintenance respiration in stems of Nor-

way spruce trees. Tree Physiol. 18, 155–166.

Stockfors, J., Linder, S., 1998b. The effect of nutrition on the

seasonal course of needle respiration in Norway spruce stands.

Trees 12, 130–138.

Stromgren, M., 2001. Soil-surface CO2 flux and growth in a boreal

Norway spruce stand. Effects of soil warming and nutrition.

Doctoral Thesis. Acta Univ. Agric. Suec., Silvestria 220, 44 pp.

Thornton, P.E., Law, B.E., Gholz, H.L., Clark, K.L., Falge, E.,

Ellsworth, D.S., Golstein, A.H., Monson, R.K., Hollinger, D.,

Falk, M., Chen, J., Sparks, J.P., 2002. Modeling and measuring

the effects of disturbance history and climate on carbon and

water budgets in evergreen needleleaf forests. Agric. For.

Meteorol. 113, 185–222.

Valentini, R., Matteucci, G., Dolman, A.J., Schulze, E.-D.,

Rebmann, C., Moors, E.J., Granier, A., Gross, P., Jensen,

N.O., Pilegaard, K., Lindroth, A., Grelle, A., Bernhofer, C.,

Grunwald, T., Aubinet, M., Ceulemanns, R., Kowalski, A.S.,

Vesala, T., Rannik, U., Berbigier, P., Loustau, D., Gudmundsson,

J., Thorgeirsson, H., Ibrom, A., Morgenstern, K., Clement, R.,

Moncrieff, J., Montagnani, L., Minerbi, S., Jarvis, P.G., 2000.

Respiration as the main determinant of carbon balance in

European forests. Nature 404, 861–865.

Wallin, G., Linder, S., Lindroth, A., Rantfors, M., Flemberg, S.,

Grelle, A., 2001. Carbon dioxide exchange in Norway spruce at

the shoot, tree and ecosystem scale. Tree Physiol. 21, 969–976.

Wang, K., Kellomaki, S., Laitinen, K., 1996. Acclimation of photo-

synthetic parameters in Scots pine after three years exposure to

elevated temperature and CO2. Agric. For. Meteorol. 82, 195–

217.

Wang, Y.-P., Jarvis, P.G., 1990a. Description and validation of an

array model—MAESTRO. Agric. For. Meteorol. 51, 257–280.

Wang, Y.-P., Jarvis, P.G., 1990b. Influence of crown structural

properties on PAR absorption, photosynthesis, and transpiration

in Sitka spruce: application of a model (MAESTRO). Tree

Physiol. 7, 297–316.

Whitehead, D., Griffin, K.L., Turnbull, M.H., Tissue, D.T., Engel,

V.C., Brown, K.J., Schuster, W.S.F., Walcroft, A.S., 2004.

Response of total night-time respiration to differences in total

daily photosynthesis for leaves in a Quercus rubra L. canopy:

implications for modelling canopy CO2 exchange. Glob. Change

Biol. 10, 925–938.

Wingate, L., 1998. Calibration of the hemispherical photographic

and beam transmittance techniques to measure leaf area index

distribution in a Picea sitchensis (Bong.) Carr. plantation. B.Sc.

University of Edinburgh, p. 47.

Wingate, L., 2003. The contribution of photosynthesis and respira-

tion to the net ecosystem exchange and ecosystem 13C discri-

mination of a Sitka spruce plantation. Ph.D. University of

Edinburgh, p. 268.