UBC Faculty Research and Publications

Boreal ecosystems sequestered more carbon in warmer years Higuchi, Kaz; Chen, Baozhang; Chen, Jing M.; Black, T. Andrew; Chan, Douglas; Liu, Jane; Tans, Pieter; Worthy, Douglas E. J. 2006-05-31

You don't seem to have a PDF reader installed, try download the pdf

Item Metadata


Black_AGU_2006GL025919.pdf [ 274.35kB ]
JSON: 1.0041924.json
JSON-LD: 1.0041924+ld.json
RDF/XML (Pretty): 1.0041924.xml
RDF/JSON: 1.0041924+rdf.json
Turtle: 1.0041924+rdf-turtle.txt
N-Triples: 1.0041924+rdf-ntriples.txt
Original Record: 1.0041924 +original-record.json
Full Text

Full Text

Boreal ecosystems sequestered more carbon in warmer yearsJing M. Chen,1Baozhang Chen,1Kaz Higuchi,2Jane Liu,3Douglas Chan,2Douglas Worthy,2Pieter Tans,4and Andy Black5Received 5 February 2005; revised 20 March 2005; accepted 13 April 2005; published 19 May 2006.[1] A 13-year (1990–1996, 1999–2004), hourly air CO2record measured on a 40 m tower in northern Canada isanalyzed against interpolated marine boundary layer CO2data representing the free troposphere above the tower. Inwarmer years, the planetary boundary layer was moredepleted with CO2, suggesting that the land area (103–104km2) upwind of the tower sequestered more carbon.After using a novel approach to derive the photosyntheticflux from the air CO2diurnal variation pattern, it isconfirmed that boreal ecosystem photosynthesis increasedmore than ecosystem respiration in warmer years.Citation: Chen, J. M., B. Chen, K. Higuchi, J. Liu, D. Chan,D. Worthy, P. Tans, and A. Black (2006), Boreal ecosystemssequestered more carbon in warmer years, Geophys. Res. Lett., 33,L10803, doi:10.1029/2006GL025919.1. Introduction[2] Atmospheric measurements, as interpreted using at-mospheric transport models [Tans et al., 1990; Denning etal., 1995; Gurney et al., 2002; Rodenbeck et al., 2003] andglobal carbon budgets based on land use history [Houghtonet al., 1999] suggest the existence of a strong carbon sink onland, but the mechanisms are still uncertain [Pacala et al.,2001; Caspersen et al., 2001; Field and Fung, 1999]. Athigh latitudes, the impacts of temperature change on eco-systems are of great concern [Braswell et al., 1997; Oechelet al., 2000]. Greater biospheric activities at higher temper-atures were inferred from remote sensing [Myneni et al.,1997] and atmospheric CO2measurements [Keeling et al.,1996]. From micrometeorological measurements at thestand level, some studies [e.g., Goulden et al., 1998] foundthat warming increased carbon release more than uptake in aboreal forest, while others [e.g., Black et al., 2000] showedthe opposite. The effect of temperature on the forest carboncycle is highly variable depending on species, age and standhistory [Chen et al., 2003], and the boreal landscapeconsists of fragmented forest patches of various ages onvariable soils and mixed with grassland and tundra due tofrequent fire and insect disturbances as well as humanactivities. How these ecosystems collectively respond toclimate change is, therefore, important in understanding themechanisms controlling regional and global carbon cycles,as boreal forests globally store 13% of carbon in above-ground biomass and 43% in soil organic matter[Schlesinger, 1991; Jarvis et al., 2000]. CO2fluxes mea-sured on micrometeorological towers in many flux networksworldwide [Baldocchi et al., 2001] have provided usefulinformation on how various ecosystems behave underdifferent climates. However, such towers can only samplea very small fraction of the land surface as each can onlyrepresent a footprint area of about 1 km2. We seek ways toretrieve carbon cycle information from atmospheric CO2concentration measurements, which have much larger foot-prints (103–104km2)[Lin et al., 2003] than flux towers.2. Data and Site[3] A 13-year (1990–1996, 1999–2004), hourly aver-aged air CO2concentration record measured on a 40-mtower at Fraserdale, northern Ontario, Canada(49C17652029.900N, 81C17634012.300W), is used for this purpose(no data were collected from January 1997 to June 1998).The measurements were made according to the WMO(Global Atmospheric Watch) guidelines, with an accuracyof 0.1 ppm [Higuchi et al., 2003]. Temperature, humidityand wind speed at 20 m and 40 m and precipitation werealso measured, allowing for accurate vertical mixing simu-lations under various atmospheric stability conditions. Theinterannual variation in air temperature was very similar tothat at the weather station Kapuskasing, 87 km southwest ofFraserdale. The Globalview CO2matrix data in 41 latitudi-nal bands based on weekly flask samples in the marineboundary layer (MBL) for the 13 years [Conway et al.,1994] were linearly interpolated to represent CO2concen-tration in the free troposphere (FT) at the site as the topboundary condition of the planetary boundary layer (PBL).According to a Landsat TM image at a 30 m resolutionacquired in 1998, the landscape (3600 km2around thetower) consists of 66% of black spruce (Picea mariana)and Jack pine (Pinus banksiana), 20% open land after forestfires and logging, 11% aspen (Populus tremuloides) andpaper birch (Betula papyrifera), and 3% open water. In theprevailing northwest wind direction, the forests are predom-inantly undisturbed.3. Modeling Methodology[4] The diurnal variation in CO2concentration abovevegetation depends on the magnitudes of nighttime ecosys-temrespirationanddaytimenetphotosynthesis.Atmosphericdiffusion also contributes to the diurnal variation becausethe strength of vertical mixing varies greatly from nighttimeto daytime. For the purpose of retrieving ecosystem infor-mation from atmospheric CO2data, we used a model tosimulate both ecosystem and atmospheric processes. Themodel consists of two components: (1) Boreal EcosystemGEOPHYSICAL RESEARCH LETTERS, VOL. 33, L10803, doi:10.1029/2006GL025919, 20061Department of Geography, University of Toronto, Canada.2Meteorological Service of Canada, Toronto, Canada.3Department of Physics, University of Toronto, Toronto, Canada.4CMDL/NOAA, Boulder, Colorado, USA.5Department of Soil Science, University of British Columbia,Vancouver, Canada.Copyright 2006 by the American Geophysical Union.0094-8276/06/2006GL025919$05.00L10803 1of4Productivity Simulator (BEPS) [Liu et al., 2002], whichsimulates ecosystem processes including water balance,photosynthesis [Farquhar et al., 1980], and autotrophicand heterotrophic respiration, and radiation and energybalances of the canopy and the soil surface; and (2) theVertical Diffusion Scheme (VDS) [Chen et al., 2004], whichsimulates CO2diffusion within the planetary boundary layer(PBL) under both stable and unstable atmospheric condi-tions. The combined BEPS-VDS model simulated well themeasured hourly CO2concentration at 40 m for the 13 years(r2= 0.71, the root mean square error, RMSE = 5.32 ppm,n = 103858). For 10-day averaged hourly values, theagreement between measurements and the model is signif-icantly improved (r2= 0.84, RMSE = 1.06 ppm, n = 11306)as the effects of horizontal advection and infrequent strongvertical diffusion associated with synoptic events becomeless significant in longer time periods. The 10-day averageddiurnal amplitudes of measured and modeled CO2agreevery well (r2= 0.96) over the 13 years.[5] In order to gain information on ecosystem behavior, amethodology is developed to separate the effects of atmo-spheric diffusion and ecosystem metabolism on the CO2concentration measurements. Figure 1 shows an example ofmeasured and simulated hourly CO2concentrations on atypical day (11 July 1996). The simulated values generallyfollow closely the measured values in the diurnal cycle. Toinvestigate the effect of daytime photosynthesis on themeasured CO2, we turned off the gross primary productivity(GPP) in BEPS from sunrise to sunset. As shown inFigure 1, the simulated CO2with GPP = 0 increasesconsiderably from the measured CO2. This increase isexpected as the carbon uptake by photosynthesis is artifi-cially terminated while the total ecosystem respiration (bothheterotrophic and autotrophic) remains unchanged. As at-mospheric diffusion is unchanged in both simulations andhas the same effect on the measured and modeled CO2, thedifference between the simulated and measured values istherefore solely due to photosynthesis. In this way, thesignal of photosynthesis is extracted from the CO2timeseries. Physically, the hourly average difference in CO2(DCi, in ppm) between the measured and simulated (withGPP = 0) cases reflects the accumulating reduction of CO2by GPP. Assuming that this reduction is uniform in themixed layer, the simulated mixed layer height (zi)andthe average dry air density (rair) can then be used to estimatethe time-integrated (since sunrise) GPP per unit surface areaas DCirairzi(mol mC02). As the air moves across thelandscape, this effect of GPP on air CO2gradually accu-mulates. For hour i after sunrise, the total accumulatedeffect is DCirairzi, and GPP in this hour is (DCirairziC0DCiC01rairziC01,inmolmC02). The daily total GPP thenequalsPSSi¼SRþ1DCiziC0DCiC01ziC01ðÞrair, where SR is the hourof sunrise and SS is sunset. The accumulation of thisphotosynthesis effect starts at sunrise and moves with theair from sunrise to sunset, and the tower CO2measurementstherefore integrate the influence of the land surface of dailyair travel length upwind of the tower. This simple method-ology makes no assumptions related to horizontal homoge-neity. Since no flux measurements were made at theFraserdale site, this methodology was validated at a towerflux site in a black spruce forest in Saskatchewan, where theupwind area is covered by forests of similar density. Halfhourly carbon fluxes in 1999 were converted into GPPusing an existing method developed at the Saskatchewansite [Griffis et al., 2003], and the concentration-deriveddaily GPP was highly correlated with that derived fromeddy covariance flux measurements (r2= 0.82, RMSE =0.11 g C mC02dC01, n = 186).4. Results and Discussion[6] A simple analysis of the CO2record against FT datareveals important temperature- dependent ecosystem signals(Figure 2a): the annual mean difference in CO2(DCFT-PBL)between FT and the daily minimum measured at 40 mincreased with the annual mean air temperature. The dailyminimum CO2value represented closely the mean value inthe well mixed PBL [Chen et al., 2004, 2005], and the dailyDCFT-PBLresulted from the net difference between grossprimary productivity (GPP) in daytime and ecosystemrespiration (ER) in both nighttime and daytime, as well asthe mixing between FT and PBL [Bakwin et al., 1998]. Theincrease in the annual mean DCFT-PBLwith temperaturesuggests that GPP increased considerably faster with tem-perature than did ER. Daily balloon temperature soundingsat Moosonee (200 km N from Faserdale) and Maniwaki(540 km SE) weather stations in the same years were usedto determine the very weak correlations between the annualPBL height and the annual mean temperature (r2= 0.12 and0.19, respectively). The PBL height increased 2% and 5%from the coldest to warmest year at these two locations,respectively, and bias estimates in Figure 2a are based onthe 5% increase. The difference in the frequency of south-erly or northerly airflows was about 4% between twocoldest (1992 and 1993) and two warmest (1999 and2001) years. Since southerly flows had a lower CO2concentration than the northerly flows by C241 ppm in thegrowing season (largest in the year), the flow direction hadFigure 1. An example of modeled and measured hourlyvalues of atmospheric CO2on 11 July 1996 at 40 m atFraserdale. The agreement indicates that both ecosystemmetabolism (photosynthesis and respiration) and atmo-spheric diffusion are well modeled. A new series is obtainedfrom sunrise to sunset (indicated by triangles) after turningoff the gross primary productivity (GPP) in the model. Inthe absence of GPP, the concentration remained higher thanthe corresponding measured values. The vertical line is thedifference between measured and simulated (with GPP = 0)CO2, that is, DCiused for estimating the cumulativedifference resulting from GPP since sunrise.L10803 CHEN ET AL.: BOREAL ECOSYSTEMS L108032of4small impacts on DCFT-PBLon a yearly basis depending onthe frequency. The total bias error from these two largestsources would only decrease, to the largest extent possible,the slope of DCFT-PBLagainst temperature (Figure 2a) byC2415%. The annual mean air pressure and temperature wereuncorrelated at Fraserdale for the 13 years and Kapuskasingfor 20 years (r2= 0.14 and 0.0003, respectively), suggestingthat the frequency of low and high pressure systemsaffecting the vertical mixing regime had only very smallinterannual variations. The coldest year of 1992 after thePinatubo volcano eruption is an outlier possibly because ofthe positive effect of the increased diffuse radiation onphotosynthesis. Without the 1992 data point, the r2valueincreases to 0.87.[7] Seasonal variations in DCFT-PBL(Figure 2b) revealthe reason for its large temperature sensitivity. In winters,marked by daily mean temperature (T) below C05C176C,DCFT-PBLwas negative and decreased slowly with increas-ing T, indicating a small increase of ER with temperature.At T > C1760, DCFT-PBLincreased rapidly, suggesting that thenet uptake of CO2by the surface, that is GPP-ER, increasedrapidly with T. As the T increase in the growing season(May-August) was only slightly less than the annual Tincrease (65–85%), an increase in the annual T resultedin an increase in the net carbon uptake. The actual amountof the net carbon uptake (in mol C mC02tC01, where t is a timeperiod of interest) equals the change in DCFT-PBL(in ppmtC01or 44.64 C2 10C06mol C mC03tC01at the sea level and T =273C176K) times the mixed layer height (m). Since the mixedlayer height in summers was about 50% higher than that inwinters, we expect that the difference in the temperaturesensitivity of (GPP – ER) between summers and winterswas also about 50% larger than what is indicated as theslope in Figure 2b. This also confirms the importance of thetiming of spring warming in ecosystem carbon cycling.[8] Using the methodology described in Section 2, dailyGPP values are derived and summed to annual values. Astrong linear relationship is found between the annualconcentration-derived GPP and annual mean air temperature(r2= 0.71, or 0.69 for active growing season mean temper-ature) (Figure 3). Other meteorological factors were weaklycorrelated with GPP (r2= 0.04 and 0.13 for precipitationand radiation, respectively). The ratio of annual evapotrans-piration modeled by BEPS to precipitation ranged from 0.40to 0.73 in these 13 years, suggesting that water was not alimiting factor for growth in this area. Also shown inFigure 3 is the annual ER modeled with consideration ofboth temperature and moisture effects [Lloyd and Taylor,1994; Potter, 1997] using a multiple layer soil model. Theactual modeled ER has an equivalent Q10value of 2.4because of the increase in the active layer in summers. TheER modeling is constrained (to <4%) by the CO2concen-tration measurements, as the nighttime CO2increase to themaximum was highly sensitive to ER, especially in calmnights with a large T inversion, when a 4% increase in ERcaused a 1.0 ppm increase in modeled CO2concentration at40 m. An optimization method was used to find ER modelparameters that produce the minimum RMSE betweenmodeled and measured CO2at 40 m. Consistent with thefinding that the net uptake of CO2by ecosystems increasedwith T (Figure 2), the concentration-derived GPP had alarger T sensitivity than that of ER (Figure 3).Figure 2. Interannual and seasonal temperature dependen-cies of atmospheric CO2over a boreal region. (a) Theannually-averaged difference in CO2(DCFT-PBL) betweenthe daily minimum in the planetary boundary layer (PBL)and the free troposphere (FT) increased with air tempera-ture. The vertical bars indicate bias errors due totemperature dependencies of the mixed layer height (leftof each data point) and the wind direction (right of each datapoint). This increase in DCFT-PBLsuggests that the PBL ismore depleted with CO2in warmer years. The slope ofDCFT-PBLagainst temperature is highly significant (p <0.0008 in the t test). (b) 10-day mean DCFT-PBLvalues vs.temperature (T), indicating that in the growing season (T >0C176C) an increase in air temperature generally induced anincrease in the PBL CO2depletion.Figure 3. Sensitivities of gross primary productivity(GPP) and ecosystem respiration (ER) to temperature inboreal ecosystems. The vertical bars indicate their errors.The concentration-derived GPP increased more withtemperature than did ER, providing a reason for the largerPBL CO2depletion in warmer years (Figure 2). Thestandard error in the slope against temperature is 0.1184 and0.1091 mol mC02yC01C176CC01for GPP and ER, respectively, andthese two slopes are significantly different in the t test (p <0.017).L10803 CHEN ET AL.: BOREAL ECOSYSTEMS L108033of4[9] We used the same model to explore the possiblereasons for the difference in the T sensitivity between GPPand ER. The large T sensitivity of GPP shown in Figure 3could not be captured by the model (r2= 0.54, RMSE = 20.5gCmC02yC01) when the nutrient availability was keptconstant, but was well simulated (r2= 0.79, RMSE = 8.3gCmC02yC01) when coupled carbon (C) and nitrogen (N)dynamics in soil and vegetation were included [Chen et al.,2003] based on C:N ratios of vegetation and soil [Dickinsonet al., 2002]. At higher T, the decomposition of soil organicmatter is faster, producing more mineralized N available forimmediate uptake by plant roots [Braswell et al., 1997;Jarvis et al., 2000]. As boreal ecosystems are nutrientlimited and plant growth is sensitive to the amount ofavailable nitrogen, more mineralized N at higher T leadsto higher productivity. These model experiments, thoughexplorative, suggest that nutrient conditions in the soilplayed an important role in the response of boreal ecosys-tems to Tchanges [Jarvis et al., 2000], in agreement with Nmineralization data from a 10-year soil heating experimentin a temperate forest [Melillo et al., 2002]. Our result is ingeneral agreement with the finding from a 5-year, 5C176C soilwarming experiment inducing an accumulated increase ofabout 80% in growth in a boreal forest [Jarvis et al., 2000].This suggests that in global carbon cycle modeling, it isimportant to consider coupled carbon and nutrient dynamics.[10] The retrieved GPP and ER values constrained by theconcentration measurements suggest that boreal ecosystemsin the vicinity of the Fraserdale tower were collectively acarbon sink of 10.8 ± 14.2 g C mC02yC01in these 13 years,which is in agreement with previous work based on remotesensing [Chen et al., 2003]. However, the uncertainties inthe absolute values of GPP and ER are still of the sameorder of magnitude as the difference between them. As therecord gets longer, these uncertainties would becomesmaller. Tower flux measurements allow immediate assess-ments of carbon balance within a small footprint, whileconcentration measurements can provide reliable informa-tion on the ecosystem response to climate change for muchlarger areas. The fact that the temperature sensitivity of GPPis larger than that of ER suggests that global warming couldlead to increased carbon sequestration in boreal ecosystems.[11] Acknowledgments. This work is supported by the CanadianFoundation for Climate and Atmospheric Sciences. Lin Huang andAlexander Shashkov of Atmospheric Science and Technology Directorate,Environment Canada provided useful comments.ReferencesBakwin, P. S., et al. (1998), Measurements of carbon dioxide on very talltowers: Results of the NOAA/CMDL program, Tellus, Ser. B, 50, 401–415.Baldocchi, D., et al. (2001), Fluxnet: A new tool to study the temporal andspatial variability of ecosystem-scale carbon dioxide, water vapour, andenergy flux densities, Bull. Am. Meteorol. Soc., 82, 2415–2434.Black, T. A., W. J. Chen, A. G. Barr, M. A. Arain, Z. Chen, Z. Nesic, E. H.Hogg, H. H. Neumann, and P. C. Yang (2000), Increased carbon seques-tration by a boreal deciduous forest in years with a warm spring, Geo-phys. Res. Lett., 27, 1271–1274.Braswell, B. H., D. S. Schimel, E. Linder, and B. Moore (1997), Theresponse of global terrestrial ecosystems to interannual temperature varia-bility, Science, 278, 870–872.Caspersen, J. P., et al. (2001), Contributions of land-use history to carbonaccumulation in U.S. forests, Science, 290, 1148–1151.Chen, J. M., et al. (2003), Spatial distribution of carbon sources and sinks inCanada’s forests, Tellus, Ser. B, 55, 622–641.Chen, B., J. M. Chen, J. Liu, D. Chan, K. Higuchi, and A. Shashkov(2004), A vertical diffusion scheme to estimate the atmospheric rectifiereffect, J. Geophys. Res., 109, D04306, doi:10.1029/2003JD003925.Chen, B., J. M. Chen, and D. E. J. Worthy (2005), Interannual variability inthe atmospheric CO2rectification over a boreal forest region, J. Geophys.Res., 110, D16301, doi:10.1029/2004JD005546.Conway, T. J., P. P. Tans, L. S. Waterman, K. K. Thoning, D. R. Kitzis,K. A. Masarie, and N. Zhang (1994), Evidence for interannual variabilityof the carbon cycle from the National Oceanic and Atmospheric Admin-istration/Climate Monitoring and Diagnostics Laboratory Global AirSampling Network, J. Geophys. Res., 99(D11), 22,831–22,856, (Dataavailable at www.cmdl.noaa.gov/ccgg/)Denning, A. S., I. Y. Fung, and D. A. Randall (1995), Latitudinal gradientof atmospheric CO2due to seasonal exchange with land biota, Nature,376, 240–243.Dickinson, R. E., et al. (2002), Nitrogen controls on climate model evapo-transpiration, J. Clim., 15, 278–295.Farquhar, G. D., S. Caemmerer, and S. Berry (1980), A biochemical modelof photosynthetic CO2assimilation in leaves of C3species, Planta, 149,78–90.Field, C. B., and I. Y. Fung (1999), The not-so-big U.S. carbon sink,Science, 285, 544–545.Goulden, M. L., et al. (1998), Sensitivity of boreal forest carbon balance tosoil thaw, Science, 279, 214–217.Griffis, T. J., et al. (2003), Ecophysiological controls on the carbon balancesof three southern boreal forests, Agric. For. Meteorol., 117, 53–71.Gurney, R. K., et al. (2002), Towards robust regional estimates of CO2sources and sinks using atmospheric transport models, Nature, 415,626–630.Higuchi, K., et al. (2003), Regional source/sink impact on the diurnal,seasonal and inter-annual variations in atmospheric CO2at a boreal forestsite in Canada, Tellus, Ser. B, 55, 115–125.Houghton, R. A., J. L. Hachler, and K. T. Lawrence (1999), The U.S.carbon budget: Contributions from land-use change, Science, 285,574–578.Jarvis, P., E. Linder, and S. Botany (2000), Constraints to growth of borealforests, Nature, 405, 904–905.Keeling, R. F., S. C. Piper, and M. Heimann (1996), Global and hemi-spheric CO2sinks deduced from changes in atmospheric O2concentra-tion, Nature, 381, 218–221.Lin, C., et al. (2003), A near-field tool for simulating the upstream influenceof atmospheric observations: The stochastic time-inverted lagrangiantransport (STILT) model, J. Geophys. Res., 108(D16), 4493,doi:10.1029/2002JD003161.Liu, J., et al. (2002), Net primary productivity mapped for Canada at 1–kmresolution, Global Ecol. Biogeogr., 11, 115–129.Lloyd, J., and J. A. Taylor (1994), On the temperature dependence of soilrespiration, Fuct. Ecol., 8, 315–323.Melillo, J. M., et al. (2002), Soil warming and carbon-cycle feedbacks tothe climate system, Science, 298, 2173–2176.Myneni, R. B., et al. (1997), Increased plant growth in the northern highlatitudes from 1988 to 1991, Nature, 386, 698–702.Oechel, W. C., et al. (2000), Acclimation of ecosystem CO2exchange in theAlaskan Arctic in response to decadal climate warming, Nature, 406,978–981.Pacala, S. W., et al. (2001), Consistent land- and atmospheric-based UScarbon sink estimates, Science, 292, 2316–2320.Potter, C. S. (1997), An ecosystem simulation model for methane produc-tion and emission from wetlands, Global Biogeochem. Cycles, 11, 495–506.Rodenbeck, C., et al. (2003), CO2flux history 1982–2001 inferred fromatmospheric data using a global inversion of atmospheric transport,Atmos. Chem. Phys., 3, 1919–1964.Schlesinger, W. H. (1991), Biogeochemistry: An Analysis of GlobalChange, Elsevier, New York.Tans, P. P., I. Y. Fung, and T. Takahashi (1990), Observation constraints onthe global atmospheric CO2budget, Science, 247, 1431–1438.C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0C0A. Black, Department of Soil Science, University of British Columbia,MCML 129 - 2357 Main Mall, Vancouver, BC, Canada V6T 1Z4.D. Chan, K. Higuchi, and D. Worthy, Meteorological Service of Canada,4905 Dufferin Street Downsview, Toronto, ON, Canada M3H 5T4.B. Chen and J. M. Chen, Department of Geography, University ofToronto, 100 St. George St., Room 5047, Toronto, ON, Canada M5S 3G3.(chenj@geog.utoronto.ca)J. Liu, Department of Physics, University of Toronto, 60 St. George St.,Toronto, ON, Canada M5S 1A7.P. Tans, CMDL/NOAA, 325 Broadway R/GMD1, Boulder, CO 80305–3328, USA.L10803 CHEN ET AL.: BOREAL ECOSYSTEMS L108034of4


Citation Scheme:


Citations by CSL (citeproc-js)

Usage Statistics

Country Views Downloads
China 7 12
United States 4 30
Germany 4 0
Russia 1 0
City Views Downloads
Shenzhen 7 12
Ashburn 4 0
Unknown 4 0
Saint Petersburg 1 0

{[{ mDataHeader[type] }]} {[{ month[type] }]} {[{ tData[type] }]}
Download Stats



Customize your widget with the following options, then copy and paste the code below into the HTML of your page to embed this item in your website.
                            <div id="ubcOpenCollectionsWidgetDisplay">
                            <script id="ubcOpenCollectionsWidget"
                            async >
IIIF logo Our image viewer uses the IIIF 2.0 standard. To load this item in other compatible viewers, use this url:


Related Items