Ecology, 92(3), 2011, pp. 633–644� 2011 by the Ecological Society of America

Light-stress avoidance mechanisms in a Sphagnum-dominatedwet coastal Arctic tundra ecosystem in Alaska

D. ZONA,1,2,7 WALTER C. OECHEL,2 JAMES H. RICHARDS,3 STEVEN HASTINGS,4 IRENE KOPETZ,5 HIROKI IKAWA,2

AND STEVEN OBERBAUER6

1Department of Biology, Research Group of Plant and Vegetation Ecology, University of Antwerp, Belgium2Global Change Research Group, San Diego State University, San Diego, California 92182 USA

3Department of Land, Air and Water Resources, University of California, Davis, California 95616 USA4Polar Field Services, Barrow, Alaska 99723 USA

5Amphos XXI, Barcelona, Spain6Florida International University, Miami, Florida 33199 USA

Abstract. The Arctic experiences a high-radiation environment in the summer with 24-hour daylight for more than two months. Damage to plants and ecosystem metabolism can bemuted by overcast conditions common in much of the Arctic. However, with climate change,extreme dry years and clearer skies could lead to the risk of increased photoxidation andphotoinhibition in Arctic primary producers. Mosses, which often exceed the NPP of vascularplants in Arctic areas, are often understudied. As a result, the effect of specific environmentalfactors, including light, on these growth forms is poorly understood. Here, we investigated netecosystem exchange (NEE) at the ecosystem scale, net Sphagnum CO2 exchange (NSE), andphotoinhibition to better understand the impact of light on carbon exchange from a moss-dominated coastal tundra ecosystem during the summer season 2006.

Sphagnum photosynthesis showed photoinhibition early in the season coupled with lowecosystem NEE. However, later in the season, Sphagnummaintained a significant CO2 uptake,probably for the development of subsurface moss layers protected from strong radiation. Wesuggest that the compact canopy structure of Sphagnum reduces light penetration to thesubsurface layers of the moss mat and thereby protects the active photosynthetic tissues fromdamage. This stress avoidance mechanism allowed Sphagnum to constitute a significantpercentage (up to 60%) of the ecosystem net daytime CO2 uptake at the end of the growingseason despite the high levels of radiation experienced.

Key words: Alaska; Arctic tundra; CO2 fluxes; light stress avoidance; photoinhibition; Sphagnum.

INTRODUCTION

Mosses represent a large component of the vegetation

of the Arctic tundra, both in number of species and

percent coverage (Steere 1978, Shaver and Chapin 1991,

Walker et al. 2005). The aboveground living biomass of

mosses range from 120 to more than 200 g/m2 (Oechel

and Sveinbjornsonn 1978, Chapin et al. 1995) and can

exceed 300 g/m2 in certain locations (Murray et al.

1989b). Despite the abundance of bryophytes in Arctic

tundra, few studies (Oechel and Collins 1976, Shaver

and Chapin 1991, Douma et al. 2007, Campioli et al.

2009a, b) have evaluated their role on the seasonal

carbon balance and the influence of environmental

conditions on both their CO2 uptake and the entire

ecosystem CO2 uptake.

Drying and high temperature could decrease moss

photosynthesis (Oechel and Sveinbjornsonn 1978, Mur-

ray et al. 1989a). However several experiments showed

that mosses are generally not water stressed in wet

tundra ecosystems in the high Arctic, as their bases are

embedded in a peat that tends to retain water (Oechel

and Collins 1976, Hickleton and Oechel 1977, Harley et

al. 1989, Murray et al. 1989a). Moreover, elevated

temperature generally is less damaging than high

irradiance as the temperature optima for photosynthesis

of mosses often exceed ambient temperatures in the

Arctic and the temperature optima adjust through rapid

acclimatization (Oechel 1976, Oechel and Collins 1976,

Harley et al. 1989).

While bryophytes can adjust to temperature they

cannot acclimate to high light due to the low nitrogen

(N) levels in their tissues (Clymo and Hayward 1982,

Murray et al. 1993). N deficiency inhibits the protein

synthesis necessary to recover from photochemical

damage (Ohad et al. 1984, Murray et al. 1993, Huang

et al. 2004). In fact, a decrease in photosynthesis due to

high irradiance has been observed in mosses even under

optimum temperature and fully hydrated conditions

(Oechel and Collins 1976, Harley et al. 1989, Murray et

al. 1993, Hajek et al. 2009).

Several studies of tundra net ecosystem exchange

(NEE) in the Arctic have shown midday depression in

Manuscript received 27 April 2010; revised 3 August 2010;accepted 13 August 2010. Corresponding Editor: J. B. Yavitt.

7 E-mail: donatella.zona@ua.ac.be

633

CO2 uptake (Vourlitis et al. 1993, Vourlitis and Oechel

1997, Kwon et al. 2006). However, to our knowledge nostudy linked this observed daily pattern in landscape

NEE to light stress on specific vegetation componentssuch as mosses. To understand the processes responsible

for the midday depression of NEE, measurements of CO2

exchange of individual ecosystem components must belinked to the whole-ecosystem NEE. This connection is

needed to understand the influence of climate changethrough changes in the timing of snowmelt, cloud cover,

and changes in irradiance, on the carbon dynamics in theArctic tundra. Recent increases in surface temperature in

the Arctic are leading to increased growing season length(Rigor et al. 2000, Euskirchen et al. 2006). An earlier

snow melt has been shown to lead to higher carbonuptake in coastal tundra (Kwon et al. 2006). However,

earlier snow melt exposes mosses, the major photosyn-thetically active vegetation at the beginning of the

summer, to high radiation early in the season, possiblynegatively impacting their photosynthetic apparatus.

Climate change can also lead to changes in summer-time circulation and cloudiness (Kay et al. 2008).

Cloudy conditions and diffuse radiation have beenshown to increase productivity in forest and grassland

ecosystems by stimulating the light-limited shade leavesin the understory, increasing light-use efficiency andenhancing CO2 uptake (Gu et al. 2002, Still et al. 2009).

By contrast NEE in peatlands has been shown to beunresponsive to changes in cloudiness and diffuse

radiation (Letts et al. 2005). This result conflicts withpreviously mentioned increased growth of Sphagnum

mosses under attenuated light conditions (Harley et al.1989, Murray et al. 1993).

The focus of this study was to investigate the influenceof Sphagnum, a dominant moss (see Fig. 1) in the Alaskan

Coastal Plain tundra, on NEE throughout the summerseason. We hypothesized that Sphagnum contributes

substantially to the entire ecosystem NEE, and thatphotoinhibition of Sphagnum photosynthesis negatively

affects ecosystem CO2 balance to a significant extent.

MATERIALS AND METHODS

Site description

This study was conducted from the date of snow meltin 2006, 13 June, until 30 August 2006, approximately

the end of the growing season. The study area was in avegetated drained lake basin at the Barrow Environ-

mental Observatory (BEO), about 10 km east of thetown of Barrow, Alaska (South tower 71816051.1700 N,

156835047.2800 W, ;4.5 m above sea level). The vege-tation primarily consists of wet sedge meadow tundra

(Webber 1978). In our site, the main vascular plants areCarex aquatilis, Eriophorum vaginatum, and Dupontia

fisheri (Olivas et al. 2010). Mosses represent the majorcomponent of the living biomass, formed mostly by

Sphagnum arcticum and tescorum (intense brown to darkbrown), S. obtusum and S. orientale (yellow to pale

yellow, found mostly in the wettest microsites), (Shaw et

al. 2008; R. Andrus personal communication). For more

details about the site, soil types, etc., refer to Zona et al.

(2009).

Eddy covariance and environmental measurements

Net ecosystem exchange (NEE) was measured with

eddy covariance (EC) (Swinbank 1951, Desjardins and

Lemon 1974). The basin was divided into three transects

(see Zona et al. 2009). Several environmental variables

(soil moisture, soil temperature, air temperature, relative

humidity, photosynthetic active radiation PAR, net

radiation, and others) were measured near the EC

tower. For details about these measurements, EC data

processing, and quality control, see Zona et al. (2009). In

this paper, we perform the analysis PAR/NEE on the

EC data filtered for u* � 0.1, not gap filled, and not

filtered for wind direction. The EC data used to estimate

daily CO2 uptake were only gap filled by linear

interpolation for gaps � 2 h. See Falge et al. (2001)

for details about quality control and gap-filling meth-

odologies. The wind direction was usually from north to

northeast, and seldom from south or southwest.

The EC data from 08:00 to 18:00 were summed into a

‘‘daytime’’ NEE, day-NEE, as this was the time range

when net Sphagnum exchange (NSE; see Net Sphagnum

exchange) measurements were available.

Biomass and leaf area index (LAI)

Biomass measurements were performed in mid-

August 2006 to determine the most abundant vascular

plants and mosses. The biomass samplings were

performed on three transects in the vegetated drained

lake basin, and three plots (253 25 cm), in each of these

transects were randomly selected for the biomass

sampling for a total of nine plots within the basin. Here

we report the results from the biomass sampling from

the three plots in the southern transect, closest to the

sites of measurements, where the vegetation was fairly

homogenous. The vegetation was removed to a depth of

4 cm in the moss layer in these plots with a knife. This

depth included the greenest and most photosynthetic

layers (Fig. 1). After removal in the field the samples

were brought to the laboratory, where litter, dead

standing biomass, green moss tissue, and above ground

living vascular plants were separated and the predom-

inant genera and species identified.

After separation, these components were over dried at

608C for three days and weighted with an analytical scale.

The seasonal progression of vascular plant develop-

ment was estimated with leaf area index measurements

(LAI). LAI was measured weekly with an optical plant

canopy analyzer (model LAI-2000; LI-COR, Lincoln,

Nebraska, USA) from 7 July to 30 August across a 200-

m transect upwind from the EC tower. In this transect

three measurements at three different points were

performed (total nine measurements for each measuring

date). LAI of the three most abundant species (Carex

aquatilis, Dupontia fisheri, Eriophorum sp.) was also

D. ZONA ET AL.634 Ecology, Vol. 92, No. 3

measured directly in the biomass samples in mid-

August. All leaves of these species were scanned next

to an object of known area (for scale) and the total leaf

area of these three species was divided by the plot area

giving LAI directly.

Sphagnum water content

For Sphagnum water content, three to four samples of

the first 2 cm of the moss layer (Fig. 1) were removed

during each measurement date and sealed in aluminum

cans for wet weight determination. Wet samples were

weighed using a high precision electronic scale (Ohaus

Explorer E01640, resolution of 60.01 g; Ohaus Corpo-

ration, Pine Brook, New Jersey, USA), then dried at

1058C for 24 h, and dry mass determined. Gravimetric

water content was calculated as percentage of dry mass.

Net Sphagnum exchange (NSE)

Net CO2 exchange of Sphagnum (hence forth called

net Sphagnum exchange, NSE) and fluorescence mea-

surements were performed weekly on Sphagnum, in three

different plots across a 200-m transect upwind from the

EC tower. In these three plots, eight samples from the

upper 2 cm of the moss layer, including the capitulum,

were removed from the field. These samples were then

placed inside standard trays (4 cm diameter, see Fig. 1)

to measure NSE and respiration with a Li-6400 portable

photosynthetic system (LI-COR, Lincoln, Nebraska,

USA). The samples were positioned inside a clear

cylindrical chamber (Fig. 1, lower left inset; 7.5 cm

diameter, 3.5 cm long; Li-6400-05, LI-COR) lined with

Teflon to minimize water absorption. Between measure-

ments, the samples were sprayed with distilled water to

maintain optimal hydration. The photosynthesis calcu-

lation was based on the differences in CO2 concentration

between the incoming and outgoing air stream flowing

through the cuvette (multiplied by the air flow and

divided by the dry mass of the mosses). Humidity was

reduced using a magnesium perclorate filter which

prevented condensation in the CO2 analyzer.

Measurements were performed from about 08:00 to

23:00, every two hours. Some of the measurement dates

showed in this paper presented a shorter time range due

to equipment failure or untenable environmental condi-

tions (e.g., sometimes the instrument powered down due

to freezing or rain). Air temperature was recorded inside

the Li-6400 chamber using a type-E thermocouple. NSE

was gap filled by linear interpolation, and summed from

08:00 to 18:00 for cumulative daily NSE.

Dark respiration measurements were performed on

each sample after light measurements using a black

cover to shade the entire chamber. Measurements were

logged two to five minutes after shading to allow

adjustment of mosses to the dark.

At the end of each set of measurements, moss dry

weight was determined, and NSE and dark respiration

were standardized for dry mass. The average dry mass of

Sphagnum present in the 0.0625-m2 plots where biomass

measurements were performed was extrapolated to 1 m2

and used to convert the Sphagnum net assimilation and

FIG. 1. Sphagnum mosses in the trays used for the NSE measurements at the beginning of July 2006. The bottom left insetshows a detail of the Li-6400-05 chamber used for these measurements; the bottom right inset shows a cross section of a Sphagnummoss sample removed on 27 July 2007. Note the greener layer ;1 cm below the capitulum (0.4 inches in the figure).

March 2011 635LIGHT-STRESS AVOIDANCE MECHANISMS

dark respiration measurements into the same units as

NEE measured by the EC towers (g CO2�m�2�h�1). Thisallowed a direct comparison of the contribution of

Sphagnum to entire NEE. Every scaling methodology to

compare moss to entire ecosystem productivity brings

some uncertainties connected to the effects of chambers

on net CO2 exchange. Methodologies based on destruc-

tive biomass samplings to estimate nonvascular contri-

bution to ecosystem NPP (Shaver and Chapin 1991,

Campioli et al. 2009b) are limited to a few sampling

times a year and do not allow examining a daily trend

and the influence of environmental variables including

light. On the other hand, removal of vascular plants and

the use of the chambers over the remaining vegetation to

measure net moss CO2 exchange (Douma et al. 2007)

include vascular plant root and soil respiration. Conse-

quently, we decided to measure only the green moss CO2

flux directly, and we consider our methodology among

the most accurate available technologies for the

objectives of this study. Because of the fairly low LAI

in our site, the light environment of the mosses in the

chambers was pretty similar to the one in the field.

Calibration of the Li-6400 was performed every two

to four weeks using ultra-high-purity nitrogen as the

CO2 and H2O zero and 729 ppm CO2 in air standard gas

(certified grade 61 ppm) (Matheson Gas Product,

Montgomeryville, Pennsylvania, USA) for the CO2

span. A dew point generator (Li-610, LI-COR) was

used to produce an air stream with a known water vapor

dew point (typically 78C lower than the ambient air

temperature).

Sphagnum fluorescence

Fluorescence measurements were performed to assess

the impact of light on Sphagnum net CO2 uptake. This

technique is necessary to investigate the performance of

photosystem II under field conditions (Genty et al. 1989,

Maxwell and Johnson 2000). These measurements were

performed using an Opti-Sciences OS1-FL modulated

fluorometer (Opti-Sciences, Tyngsboro, Massachusetts,

USA) equipped with a fiber-optic cable. Photosystem II

(PSII) activity was evaluated using the yield of quantum

efficiency (light-adapted test, UPSII or DF/F 0m) under

photosynthetic conditions and the dark-adapted ratio of

variable to maximal fluorescence (Fv/Fm, or maximum

quantum yield). These two tests were performed on 10–

12 different points in the Sphagnum layers (five or sixmeasurements on the upper part of Sphagnum, the

capitulum, and five or six measurements 1 cm below it,corresponding to the greenest and most photosynthet-

ically active layer of Sphagnum canopy, see Fig. 1, lowerright inset). These measurements were performed everytwo hours at the same time as the NSE measurements.

The light-adapted test was performed using openbody clips holding the fiber-optic 0.6 mm and 608 from

the capitulum of the mosses. This test is theoreticallyproportional to the operating quantum efficiency of

PSII photochemistry, and was used to determine theproportion of the light absorbed by photosystem II used

in photochemistry (Genty et al. 1989). After the light-adapted test, the same samples were placed in standard

Opti-Sciences clips and dark adapted for 15 minutes.The leaf clips were wrapped in aluminum foil to prevent

drying of the bryophyte material and ensure completedark conditions, as described by Marschall and Proctor

(2004).Initial fluorescence (F0) was determined using far-red

light (700 nm) that excites PSI and positions the openPSII centers in a relaxed state. In this condition, QA, the

first electron acceptor of PSII, is fully reduced. At thisstage a saturating pulse, provided by a filtered 35-W

halogen lamp (;15 000 lmol�m�2�s�1 PAR) allowed themeasurement of maximum fluorescence (Fm); Fv, the

variable fluorescence, was determined as Fv ¼ Fm � F0.The maximum quantum efficiency of photosystem II(PSII) primary photochemistry was calculated as Fv/Fm

¼ (Fm � F0)/Fm.Fv/Fm traces were recorded at the beginning of the

field season to ensure an adequate duration for thesaturation pulse, which was set to 0.8 s. Fluorescence

measurements were performed on different samplesfrom the ones used for gas exchange measurements to

reduce stress.

Statistical analysis

General linear modeling (Systat version 10; Systat

Software 2002) was used to identify the explanatorypower and significance of PAR on NEE, on NSE, and

test the significance of the interaction term PAR 3 date.

RESULTS

Environmental and vegetation measurements

The summer season 2006 was fairly wet with totalprecipitation of 62 mm from 13 June to 31 August. The

average air temperature from 13 June to 31 August was38C. The decrease in sun angle and the increase in cloud

cover from June to August caused a decrease in PAR(i.e., decrease in maximum and minimum PAR, data not

shown). The generally cloudier conditions later in thesummer (mostly August) also probably increased the

diffuse radiation. Leaf area index (LAI) increasedduring peak season followed by a decrease caused by

vascular plant senescence late in the season (Table 1).

TABLE 1. Leaf area index (LAI) for each of the indicated datesof summer 2006 at the measurement site.

Date LAI

7 Jul 0.31 6 0.1512 Jul 0.39 6 0.1917 Jul 0.38 6 0.1226 Jul 0.47 6 0.031 Aug 0.64 6 0.3914 Aug 0.52 6 0.3126 Aug 0.42 6 0.1330 Aug 0.47 6 0.29

Note: Values are means 6 SE (n ¼ 9).

D. ZONA ET AL.636 Ecology, Vol. 92, No. 3

Based on destructive sampling measurements in mid-

August, the most abundant vascular plants were Carex

aquatilis (LAI of 0.43), Eriophorum sp. (LAI of 0.13),

and Dupontia fisheri (LAI of 0.02). The sum of these

estimates led to a total LAI of 0.57 which was similar to

the LAI estimate of 0.52 on 14 August derived using the

indirect method (the LAI-2000). The large majority of

mosses belonged to the two genera Sphagnum and

Polytricum. While Polytricum accounted for only 2.4%

of the total dry weight, Sphagnum ssp. was about 74%.

The average dry mass of green/photosynthetic Sphag-

num at our site was 368 g/m2, higher than previous

measurements reported for the Arctic tundra (Oechel

and Sveinbjornsonn 1978, Murray et al. 1989b, Chapin

et al. 1995). The waterlogged conditions of the soils at

our site likely favored Sphagnum establishment and

growth. The dry mass of Carex aquatilis was around 6%

of the total dry mass, Eriophorum sp. was 1.6% of the

total dry weight, and Dupontia fisheri was 0.3% of the

total dry mass.

Optimum water content for photosynthesis in the

Sphagnaceae generally ranges between 700% and 2000%

of dry mass (Ueno and Kanda 2006). The water content

of Sphagnum measured in this study was consistently

within this range, indicating that the moss mat was

always fully hydrated, but not supersaturated (which

would also reduce photosynthetic rates). Therefore,

water stress or supersaturation was excluded as limiting

factors on photosynthesis during the course of these

observations. This was further supported by the lack of

a significant correlation between moss water content and

rates of NSE.

Net ecosystem exchange (NEE), net Sphagnum

exchange (NSE), and fluorescence

Light response curves of NEE and NSE for different

time periods during the summer season 2006, showed a

differential response to radiation input (Fig. 2). Nega-

tive signs for NEE and NSE indicate CO2 uptake

(photosynthesis higher than respiratory loss) and

positive signs indicate CO2 release (respiratory loss

higher than photosynthetic uptake). From 21 to 28 June,

the ecosystem was a CO2 sink at moderate PAR levels

below about 800 lmol�m�2�s�1 and above this value it

switched to a CO2 source (Fig. 2). During this period,

the polynomial fit NEE and PAR showed the highest r2

FIG. 2. Light response curves of net ecosystem exchange (NEE; open circles) and net Sphagnum exchange (NSE; solid circles)for the indicated periods. The regression between NEE and PAR for the period 21–28 June is a polynomial fit with r2¼ 0.23 and P, 0.001; for the period 29 June to 13 July, the regression is a rectangular hyperbolic fit with r2¼ 0.48 and P , 0.001; for the period14–31 July, it is a rectangular hyperbolic fit with r2 ¼ 0.68 and P , 0.001; and for the period 10–31 August, it is a rectangularhyperbolic fit with r2¼ 0.51 and P , 0.001. The regression between NSE and PAR for 25 June is a linear fit with r2¼ 0.69 and P¼0.05; for 2, 3, and 10 July, it is a linear fit with r2¼ 0.14 and P¼ 0.1; for 18 and 28 July, it is a polynomial fit with r2¼ 0.35 and P¼0.07; and for 12, 20, and 30 August, it is a rectangular hyperbolic fit with r2¼ 0.53 and P , 0.001.

March 2011 637LIGHT-STRESS AVOIDANCE MECHANISMS

(0.23) while the rectangular hyperbolic model (where

NEE increases with PAR until a saturation value)

presented r2 ¼ 0.06. At this time also NSE indicated

decreased photosynthetic activity with increase in PAR

(Fig. 2). During this part of the season the sensitivity of

Sphagnum to strong radiation lead to a significant

midday depression of carbon fixation (Fig. 3). This was

particularly evident on 25 June, when the clear sky and

strong PAR during the first half of the day caused the

ecosystem to be a CO2 source. Around 12:00 the sharp

FIG. 3. (a, b) Incoming photosynthetically active radiation (PAR) measured above the plant canopy (;1.5 m above ground),(c, d) net ecosystem exchange (NEE) measured by the eddy covariance (EC) towers and net Sphagnum exchange (NSE) measuredwith the chamber method, (e, f ) Fv/Fm in the capitulum and 1 cm below it, and (g, h) yield of quantum efficiency (UPSII) in thecapitulum and 1 cm below it, on 25 June and 3 July 2006. Error bars indicate 6SE.

D. ZONA ET AL.638 Ecology, Vol. 92, No. 3

decrease in PAR, due to sudden increase in cloudiness,

switched NEE and NSE to CO2 uptake. Strong

radiation decreased the efficiency of photosystems

during the day mostly in the capitulum (as shown by

the lower values of Fv/Fm and yield when PAR was

highest and for a short time thereafter, Fig. 3).

Later in the season (from 29 June to 13 July) the

correlation between NEE and PAR changed and the

ecosystem remained a CO2 sink with increasing radia-

tion, and it was well described by a rectangular

hyperbolic model (Fig. 2). At this time NSE presented

a fairly noisy dependence on light but the relationship to

PAR and diurnal measurements showed decreased

photosynthetic activity at high light (Fig. 2). Vascular

plants started to develop, becoming progressively more

important for the tundra NEE (see increase in LAI,

Table 1).

From 14 July to early August, the ecosystem became

progressively a stronger CO2 sink, and NEE reached

saturation at around 800–1000 lmol�m�2�s�1. During

this time NSE also reached its maximum value. From 18

to 28 July, NSE saturated at approximately 600

lmol�m�2�s�1 (Fig. 2). Vascular plant development

reached the seasonal maximum at the beginning of

August (see Table 1) when Sphagnum constituted a

smaller, but still considerable fraction of NEE (see Fig.

4). The photosynthetic capacity of the capitulum became

lower, as demonstrated by their lower Fv/Fm and yield

during the day (Fig. 4). However, Fv/Fm and yield of the

capitulum partially recovered at night (Fig. 4).

At the end of the summer season (10–31 August) NEE

showed a more scattered dependence on light. Vascular

plant senescence led to a decrease in LAI (Table 1). NSE

showed a similar trend as in the previous period, with a

slightly different saturation level (around 800

lmol�m�2�s�1; Fig. 2). With the senescence of the

vascular plants, Sphagnum became dominant again

accounting for almost the entire NEE earlier in the

morning (;93%) and about half later in the day (on 12

August; see Fig. 4). The photosynthetic capacity of the

capitulum remained low during the day and partly

recovered at night (see Fv/Fm and yield; Fig. 4).

The change in light dependence of NEE and of NSE is

demonstrated by the fact that the interaction term PAR

3 date is very significant in explaining both NEE and

NSE (in a GLM P , 0.001, r2¼ 0.38 and P¼ 0.001, r2¼0.38, respectively).

At the beginning of the summer the strong radiation

(Fig. 5a) resulted in low net CO2 sink activity by NSE

limiting NEE (Fig. 5c). The capitulum of Sphagnum

showed a progressive decrease in Fv/Fm, while the

subsurface layer (1 cm) Fv/Fm values were nearly

constant over the summer, showing only a slight

increase at peak season followed by a small decrease at

the end of the summer (Fig. 5b). Later in the season

Sphagnum was able to maintain significant CO2 uptake,

which constituted up to 60% of the NEE during mid to

late August (Fig. 5c).

DISCUSSION

From 21 to 28 June, the ecosystem initiated CO2

uptake. At this time, the photosynthetic apparatus of

Sphagnum was already light stressed. The light stress isindicated by the transition from CO2 sink to source with

increased radiation (Fig. 2), by the relatively low Fv/Fm,and by the decrease in yield of quantum efficiency over

the course of the day (Fig. 3, 25 June). The lightinhibition of NSE affected the entire ecosystem NEE,

which became a CO2 source above a PAR of 800lmol�m�2�s�1 (Fig. 2). During this time, the rectangular

hyperbolic model was not able to significantly modelNEE, while a polynomial fit with a decrease in NEE at

high PAR more adequately described the observedtrend.

In July, vascular plants started developing, as shownby the increase in LAI (Table 1). Probably also

Sphagnum started growing new tissues and startedreallocating its photosynthetic capacity in subsurface

layers protected from direct radiation. This is shown bythe slight increase in Fv/Fm 1 cm below the capitulum

(Fig. 5b). Overall, different layers responded in differingways to an increase in radiation (i.e., top layers wereinhibited, subsurface layers became more active) and the

net effect resulted in a complex and variable response toincreasing light (Fig. 2). The growth of vascular plants

partially shaded the mosses, even if the low LAI onlysuggests a minor light attenuation. The growth of

vascular plants probably greatly contributed to theincreased CO2 uptake and the change in the light

dependence of the entire ecosystem, which no longershowed inhibition at high light levels (Fig. 2). In fact,

light-curve experiments performed on Carex in this siteshowed that this species was not light inhibited at this

time (S. Oberbauer, unpublished data).After mid-July, PAR decreased (data not shown)

probably due to a combination of increase in cloudinessand decrease in sun angle. This decrease in radiation

input at the same time as an increase in vascular plantcanopy coverage decreased the light stress on the

photosynthetic apparatus of Sphagnum. At this time,probably Sphagnum continued to grow and reallocatemost of its photosynthetic capacity in subsurface layers,

as shown by the higher Fv/Fm and yield 1 cm below thecapitulum (Fig. 4). Moreover, an increase in radiation

above 800 lmol�m�2�s�1 slightly decreased NSE, but didnot inhibit it completely as in previous periods (Fig. 2).

Between the end of July and the beginning of August,when maximum development of vascular plants oc-

curred, Sphagnum constituted a reduced, but stillsubstantial fraction (about 31% on 28 July), of the

entire ecosystem NEE (Fig. 5c). Later in the season,when vascular plants started to senesce, the relative

contribution of Sphagnum to the entire ecosystem NEEagain increased, and accounted for on average 60% of

the total ecosystem NEE on 12 August and 30 August(Fig. 5c). In summary, while NEE and NSE showed

indications of midday depression early in the season,

March 2011 639LIGHT-STRESS AVOIDANCE MECHANISMS

later in the season this effect was no longer present (Fig.

2). The light inhibition early in the season significantly

influenced the entire ecosystem CO2 uptake, decreasing

NEE.

The seasonal trend in Fv/Fm values suggested the

occurrence of photoinhibition mostly in the top layers of

Sphagnum. Fv/Fm in the capitulum progressively de-

creased during the summer season to values less than 0.5

FIG. 4. (a, b) Incoming photosynthetically active radiation (PAR) measured above the plant canopy (;1.5 m above ground),(c, d) net ecosystem exchange (NEE) measured by the eddy covariance (EC) towers and net Sphagnum exchange (NSE) measuredwith the chamber method, (e, f ) Fv/Fm in the capitulum and 1 cm below it, and (g, h) yield of quantum efficiency (UPSII) in thecapitulum and 1 cm below it, on 28 July and 12 August. Error bars indicate 6SE.

D. ZONA ET AL.640 Ecology, Vol. 92, No. 3

(see Fig. 5b). Unstressed vascular plants typically have

Fv/Fm of 0.8 (Bjorkman and Demmig 1987). Values

higher than 0.7 have been measured in unstressed algae

(Hader et al. 2002), unstressed Antarctic mosses

(Lovelock et al. 1995), and in Sphagnum in low light

conditions (Murray et al. 1993, Rice et al. 2005, Hajek et

al. 2009).

However, even if inhibited, Sphagnum was able to

maintain a significant CO2 uptake throughout the entire

summer. The capitulum of Sphagnum appeared progres-

sively browned as the season progressed for to the

development of brown pigments, which is probably

connected to their significant light stress and decreased

photosynthetic capacity, while the subsurface layers

appeared green (Fig. 1) and were photosynthetically

active (Fig. 5b). This is in agreement with the observed

yellow-brown top of Sphagnum hummocks and contem-

poraneous bright green color at the base of the cushions,

attributed to bleaching of tissues exposed to high light

(Harley et al. 1989).

The development of subsurface layers protected from

light stress, together with the decrease in PAR, allowed

Sphagnum to form a significant fraction of the NEE later

in the season (about 60%, Fig. 5c). Not being able to

photoacclimate as a result of low nitrogen content and

an inability to maintain rapid protein synthesis neces-

sary for recovery from photochemical damage (Clymo

and Hayward 1982, Ohad et al. 1984, Henley et al. 1991,

Murray et al. 1993), Sphagnum has apparently evolved a

morphological rather than physiological adaptation to

FIG. 5. (a) Photosynthetically active radiation (PAR) measured above the plant canopy (;1.5 m above ground) averaged from08:00 to 18:00, (b) Fv/Fm in the capitulum and 1 cm below it, averaged per each measurement date, and (c) net ecosystem exchange(NEE) and net Sphagnum exchange (NSE), summed from 08:00 to 18:00, and percentage contribution of NSE to the total NEE.Error bars indicate 6SE.

March 2011 641LIGHT-STRESS AVOIDANCE MECHANISMS

high light levels, and grows in compact mats whose

brown top layers form a protecting screen, preventing

light stress of the subsurface photosynthetic layers.

The characteristic reddish-violet color of Sphagnum

has been connected to secondary cell-wall pigments,

sphagnorubins (Gerdol et al. 1998). These pigments

have antioxidant activity which increase tolerance to

radiation and photo-oxidative damage (Husain et al.

1987, Krol et al. 1995, Steyn et al. 2002). However, here

we showed that their photoprotective role is not enough

to prevent damage of the photosystems of the upper

moss layer, the capitulum. This result is in agreement

with the observed higher photosynthetic capacity in the

subapical segments in several Sphagnum species from

open habitats (Hajek et al. 2009) and with the

observation that Sphagnum could tolerate both high

and low radiation inputs (Bonnett et al. 2010).

Considering the water saturated condition of the moss

mat for the majority of the summer season and similar

saturated conditions found in the wet-sedge and tussock

tundra by several researchers (Oechel and Collins 1976,

Harley et al. 1989, Murray et al. 1989a), we suggest that

the compact canopy structure of Sphagnum could

represent an evolutionary strategy to minimize light

penetration to the subsurface layers of the moss mat.

Past estimates showed that in tundra ecosystems the

contribution of mosses to total NEE has been estimated

to reach a maximum of 25% in a dwarf-shrub tundra

(Campioli et al. 2009b), 28% in tussock tundra (Shaver

and Chapin 1991), and around 60% in a moss

dominated tundra (Douma et al. 2007). Our results are

close to the last estimate. In fact, we showed a significant

contribution of Sphagnum to total NEE in the Arctic

tundra and large variation of this contribution through-

out the day (up to 93% early in the morning on 12

August), and throughout the growing season (about

60% of the daily NEE on 12 and 31 August, Fig. 5c).

Due to the dominant role of Sphagnum in this

ecosystem, we suggest that climate change and species

composition change could lead to a significant change in

carbon dynamics and change the light response of NEE

in the Arctic tundra.

Climate change is increasing climate variability,

changing cloudiness, decreasing snow cover, and in-

creasing in the length of the growing season (Rigor et al.

2000, Euskirchen et al. 2006). Higher radiation input

stimulated vascular plant photosynthetic uptake, while

inhibiting mosses, so that the tundra maintained CO2

uptake in different light conditions. This could be the

reason why peatlands do not show increased productiv-

ity with increased cloudiness (Letts et al. 2005). The

concept of plant–environment system as a dynamic unit

that reacts as a whole, which has been defined as

‘‘holocoenotic’’ (Billings 1952), could be extended to

different vegetation components of the Arctic ecosys-

tems with complementary strategies to adjust to or avoid

the environmental stress. Different vegetation compo-

nents of the Arctic tundra, and even different parts of

the same vegetation presents complementary responses

to environmental stress, allowing the entire ecosystem to

maintain CO2 uptake in different light conditions. The

net of these different processes makes the light

dependence of NEE in tundra ecosystems very complex,

and difficult to model without studying the single

components and the entire ecosystem at the same time.

ACKNOWLEDGMENTS

Funding for this study was provided by NSF (OPP-0421588).Logistic support was provided by NSF, and Glenn Sheehan(BASC). We thank the SDSU field crew and all the BASC andUIC workers and bear guides, especially Hermann. We alsothank Rommel Zulueta and Joseph Verfaillie for field supportand suggestions on experimental design, Paulo Olivas forbiomass measurements and help on the experiment, andRichard Andrus for Sphagnum species identification. We thankalso Matteo Campioli for a thorough revision of the manuscriptand Sara Vicca for the help in the statistical analysis. Theresearch is conducted on the privately owned Ukpeavik InupiatCorporation (UIC), and we thank UIC for providing access totheir laboratory where some of the measurements wereperformed.

LITERATURE CITED

Billings, W. D. 1952. The environmental complex in relation toplant growth and distribution. Quarterly Reviews of Biology27:251–265.

Bjorkman, O., and B. Demmig. 1987. Photon yield of O2

evolution and chlorophyll fluorescence characteristics at 77 Kamong vascular plants of diverse origins. Planta 170:489–504.

Bonnett, S. A. F., N. Ostle, and C. Freeman. 2010. Short-termeffect of deep shade and enhanced nitrogen supply onSphagnum capillifolium morphophysiology. Plant Ecology207:347–358.

Campioli, M., A. Michelsen, A. Demey, A. Vermeulen, R.Samson, and R. Lemeur. 2009a. Net primary production andcarbon stocks for subarctic mesic-dry tundras with contrast-ing microtopography, altitude, and dominant species. Eco-systems 12:760–776.

Campioli, M., R. Samson, A. Michelsen, S. Jonasson, R.Baxter, and R. Lemeur. 2009b. Nonvascular contribution toecosystem NPP in a subarctic heath during early and lategrowing season. Plant Ecology 202:41–53.

Chapin, F. S., III, G. R. Shaver, A. E. Giblin, K. J.Nadelhoffer, and J. A. Laundre. 1995. Responses of Arctictundra to experimental and observed changes in climate.Ecology 76:694–711.

Clymo, R. S., and P. M. Hayward. 1982. The ecology ofSphagnum. Pages 229–289 in A. J. E. Smith, editor.Bryophyte ecology. Chapman and Hall, London, UK.

Desjardins, R. L., and E. R. Lemon. 1974. Limitations of aneddy-correlation technique for the determination of thecarbon dioxide and sensible heat fluxes. Boundary LayerMeteorology 5:475–488.

Douma, J. C., M. T. Van Wijk, S. I. Lang, and G. R. Shaver.2007. The contribution of mosses to the carbon and waterexchange of arctic ecosystems: quantification and relation-ships with system properties. Plant, Cell and Environment 30:1205–1215.

Euskirchen, E. S., A. D. McGuire, D. W. Kicklighter, Q.Zhuang, J. S. Clein, R. J. Dargaville, D. G. Dye, J. S.Kimball, K. C. McDonald, J. M. Melillo, V. E. Romanov-sky, and N. V. Smith. 2006. Importance of recent shifts in soilthermal dynamics on growing season length, productivity,and carbon sequestration in terrestrial high-latitude ecosys-tems. Global Change Biology 12:731–750.

D. ZONA ET AL.642 Ecology, Vol. 92, No. 3

Falge, E., et al. 2001. Gap filling strategies for defensible annualsums of net ecosystem exchange. Agricultural and ForestMeteorology 107:43–69.

Genty, B., J. M. Briantais, and N. R. Baker. 1989. Therelationship between the quantum yield of photosyntheticelectron-transport and quenching of chlorophyll fluores-cence. Biochimica et Biophysica Acta 990:87–92.

Gerdol, R., A. Bonora, R. Marchesini, R. Gualandri, and S.Pancaldi. 1998. Growth response of Sphagnum capillifoliumto nighttime temperature and nutrient level: mechanisms andimplications for global change. Arctic and Alpine Research30:388–395.

Gu, L., D. Baldocchi, S. B. Verma, T. A. Black, T. Vesala,E. M. Falge, and P. R. Dowty. 2002. Advantages of diffuseradiation for terrestrial ecosystem productivity. Journal ofGeophysical Research 107:1–23.

Hader, D. P., M. Lebert, R. P. Sinha, E. S. Barbieri, and E. W.Helbling. 2002. Role of protective and repair mechanisms inthe inhibition of photosynthesis in marine macroalgae.Photochemistry and Photobiological Sciences 1:809–814.

Hajek, T., E.-S. Tuittila, M. Ilomets, and R. Laiho. 2009. Lightresponses of mire mosses: a key to survival after water-leveldrawdown? Oikos 118:240–250.

Harley, P. C., J. D. Tenhunen, K. J. Murray, and J. Beyers.1989. Irradiance and temperature effects on photosynthesisof tussock tundra Sphagnum mosses from the foothills of thePhilip Smith Mountains, Alaska. Oecologia 79:251–259.

Henley, W. J., G. Levavasseur, L. A. Franklin, C. B. Osmond,and J. Ramus. 1991. Photoacclimation and photoinhibitionin Ulva rotundata as influenced by nitrogen availability.Planta 184:235–243.

Hickleton, P. R., and W. C. Oechel. 1977. The influence of lightintensity and temperature on the field carbon dioxideexchange of Dicranum fuscescens in the subarctic. Arcticand Alpine Research 9:407–419.

Huang, Z. A., D. A. Jiang, Y. Yang, J. W. Sun, and S. H. Jin.2004. Effects of nitrogen deficiency on gas exchange,chlorophyll fluorescence, and antioxidant enzymes in leavesof rice plants. Photosynthetica 42:357–364.

Husain, S. R., J. Cillard, and P. Cillard. 1987. Hydroxyl radicalscavenging activity of flavonoids. Phytochemistry 26:2489–2491.

Kay, J. E., T. L’Ecuyer, A. Gettelman, G. Stephens, and C.O’Dell. 2008. The contribution of cloud and radiationanomalies to the 2007 Arctic sea ice extent minimum.Geophysical Research Letters 35:L08503.

Krol, M., G. R. Gray, N. P. A. Huner, V. M. Hurry, G. Oquist,and L. Malek. 1995. Low-temperature stress and photope-riod affect an increased tolerance to photoinhibition in Pinusbanksiana seedlings. Canadian Journal of Botany 73:1119–1127.

Kwon, H.-J., W. C. Oechel, R. C. Zulueta, and S. J.Hastings. 2006. Effects of climate variability on carbonsequestration among adjacent wet sedge tundra and moisttussock tundra ecosystems. Journal of Geophysical Re-search 111:G03014.

Letts, M. G., P. M. Lafleur, and N. T. Roulet. 2005. On therelationship between cloudiness and net ecosystem carbondioxide exchange in a peatland ecosystem. Ecoscience 12:53–59.

Lovelock, C. E., C. B. Osmond, and R. D. Seppelt. 1995.Photoinhibition in the Antarctic moss Grimmia antarcticiCard. when exposed to cycles of freezing and thawing. Plant,Cell and Environment 18:1395–1402.

Marschall, M., and M. C. F. Proctor. 2004. Are bryophytesshade plants? Photosynthetic light responses and proportionsof chlorophyll a, chlorophyll b and total carotenoids. Annalsof Botany 94:593–603.

Maxwell, K., and G. N. Johnson. 2000. Chlorophyll fluores-cence-a practical guide. Journal of Experimental Botany 51:659–668.

Murray, K. J., P. C. Harley, J. Beyers, H. Walz, and J. D.Tenhunen. 1989a. Water content effects on photosyntheticresponse of Sphagnum mosses from the foothills of the PhilipSmith Mountains, Alaska. Oecologia 79:244–250.

Murray, K. J., J. D. Tenhunen, and J. Kummerow. 1989b.Limitations on Sphagnum growth and net primary produc-tion in the foothills of the Philip Smith Mountains, Alaska.Oecologia 80:256–262.

Murray, K. J., J. D. Tenhunen, and R. S. Nowak. 1993.Photoinhibition as a control on photosynthesis and produc-tion of Sphagnum mosses. Oecologia 96:200–207.

Oechel, W. C. 1976. Seasonal patterns of temperature responseof CO2 flux and acclimation in Arctic mosses growing in situ.Photosynthetica 10:447–456.

Oechel, W. C., and N. J. Collins. 1976. Comparative CO2

exchange patterns in mosses from two tundra habitats atBarrow, Alaska. Canadian Journal of Botany 54:1355–1369.

Oechel, W. C., and B. Sveinbjornsonn. 1978. Primaryproduction processes in Arctic bryophytes at Barrow,Alaska. Pages 269–298 in L. L. Tieszen, editor. Vegetationand production ecology of an Alaskan Arctic tundra.Springer, New York, New York, USA.

Ohad, I., D. Kyle, and C. Arntzen. 1984. Membrane proteindamage and repair: removal and replacement of inactivated32-kilodalton polypeptides in chloroplast membranes. Jour-nal of Cell Biology 99:481–485.

Olivas, P. C., S. F. Oberbauer, C. E. Tweedie, W. C. Oechel,and A. Kuchy. 2010. Responses of CO2 flux components ofAlaskan Coastal Plain tundra to shifts in water table. Journalof Geophysical Research: Biosciences. In press.

Rice, S. K., C. Gutman, and N. Krouglicof. 2005. Laserscanning reveals bryophyte canopy structure. New Phytolo-gist 166:695–704.

Rigor, I. G., R. L. Colony, and S. Martin. 2000. Variations insurface air temperature observations in the Arctic, 1979–97.Journal of Climate 13:896–914.

Shaver, G. R., and F. S. Chapin III. 1991. Production: biomassrelationships and element cycling in contrasting Arcticvegetation types. Ecological Monographs 61:1–31.

Shaw, A. J., R. E. Andrus, and B. Shaw. 2008. Sphagnumberingiense sp. nov. (Bryophyta) from Arctic Alaska, basedon morphological and molecular data. Systematic Botany 33:469–477.

Steere, W. C. 1978. The mosses of Arctic Alaska. J. Cramer,Vaduz, Germany.

Steyn, W. J., S. J. E. Wand, D. M. Holcroft, and G. Jacobs.2002. Anthocyanins in vegetative tissues: a proposed unifiedfunction in photoprotection. New Phytologist 155:349–361.

Still, C. J., W. J. Riley, S. C. Biraud, D. C. Noone, N. H.Buenning, J. T. Randerson, M. S. Torn, J. Welker, J. W. C.White, R. Vachon, G. D. Farquhar, and J. A. Berry. 2009.Influence of clouds and diffuse radiation on ecosystem-atmosphere CO2 and CO18O exchanges. Journal of Geo-physical Research, Biogeosciences 114:G01018.

Swinbank, W. C. 1951. The measurement of vertical transfer ofheat and water vapour by eddies in the lower atmosphere.Journal of Meteorology 72:135–132.

Systat Software. 2002. Systat version 10. Systat Software, Inc.,Chicago, Illinois, USA.

Ueno, T., and H. Kanda. 2006. Photosynthetic response of thearctic semi-aquatic moss Calliergon giganteum to watercontent. Aquatic Botany 85:241–243.

Vourlitis, G. L., and W. C. Oechel. 1997. Landscape-scale CO2,H2O vapor, and energy flux of moist-wet coastal tundraecosystems over two growing-seasons. Journal of Ecology 85:575–590.

March 2011 643LIGHT-STRESS AVOIDANCE MECHANISMS

Vourlitis, G. L., W. C. Oechel, S. J. Hastings, and M. A.Jenkins. 1993. The effect of soil moisture and thaw depth onCH4 flux from wet coastal tundra ecosystem on the northslope of Alaska. Chemosphere 26:329–338.

Walker, D. A., et al. 2005. The circumpolar Arctic vegetationmap. Journal of Vegetation Science 16:267–282.

Webber, P. J. 1978. Spatial and temporal variation of thevegetation and its productivity, Barrow, Alaska. Pages 37–

112 in L. L. Tieszen, editor. Vegetation and productionecology of an Alaskan Arctic tundra. Springer, New York,New York, USA.

Zona, D., W. C. Oechel, J. Kochendorfer, K. T. Paw U, A. N.Salyuk, P. Olivas, S. F. Oberbauer, and D. A. Lipson. 2009.Methane fluxes during the initiation of a large-scale watertable manipulation experiment in the Alaskan Arctic tundra.Global Biogeochemical Cycles 23:GB2013.

D. ZONA ET AL.644 Ecology, Vol. 92, No. 3