UBC Faculty Research and Publications

Ecosystem carbon dioxide fluxes after disturbance in forests of North America Davis, K. J. 2011

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

Item Metadata


Black_AGU_2010JG001390.pdf [ 391.46kB ]
JSON: 1.0041934.json
JSON-LD: 1.0041934+ld.json
RDF/XML (Pretty): 1.0041934.xml
RDF/JSON: 1.0041934+rdf.json
Turtle: 1.0041934+rdf-turtle.txt
N-Triples: 1.0041934+rdf-ntriples.txt

Full Text

Ecosystem carbon dioxide fluxes after disturbance in forests of North America B. D. Amiro,1 A. G. Barr,2 J. G. Barr,3 T. A. Black,4 R. Bracho,5 M. Brown,4 J. Chen,6 K. L. Clark,7 K. J. Davis,8 A. R. Desai,9 S. Dore,10 V. Engel,3 J. D. Fuentes,8 A. H. Goldstein,11 M. L. Goulden,12 T. E. Kolb,10 M. B. Lavigne,13 B. E. Law,14 H. A. Margolis,15 T. Martin,5 J. H. McCaughey,16 L. Misson,11,17 M. Montes‐Helu,10 A. Noormets,18 J. T. Randerson,12 G. Starr,19 and J. Xiao20 Received 15 April 2010; revised 9 June 2010; accepted 18 June 2010; published 27 October 2010. [1] Disturbances are important for renewal of North American forests. Here we summarize more than 180 site years of eddy covariance measurements of carbon dioxide flux made at forest chronosequences in North America. The disturbances included stand‐ replacing fire (Alaska, Arizona, Manitoba, and Saskatchewan) and harvest (British Columbia, Florida, New Brunswick, Oregon, Quebec, Saskatchewan, and Wisconsin) events, insect infestations (gypsy moth, forest tent caterpillar, and mountain pine beetle), Hurricane Wilma, and silvicultural thinning (Arizona, California, and New Brunswick). Net ecosystem production (NEP) showed a carbon loss from all ecosystems following a stand‐replacing disturbance, becoming a carbon sink by 20 years for all ecosystems and by 10 years for most. Maximum carbon losses following disturbance (g C m−2y−1) ranged from 1270 in Florida to 200 in boreal ecosystems. Similarly, for forests less than 100 years old, maximum uptake (g C m−2y−1) was 1180 in Florida mangroves and 210 in boreal ecosystems. More temperate forests had intermediate fluxes. Boreal ecosystems were relatively time invariant after 20 years, whereas western ecosystems tended to increase in carbon gain over time. This was driven mostly by gross photosynthetic production (GPP) because total ecosystem respiration (ER) and heterotrophic respiration were relatively invariant with age. GPP/ER was as low as 0.2 immediately following stand‐replacing disturbance reaching a constant value of 1.2 after 20 years. NEP following insect defoliations and silvicultural thinning showed lesser changes than stand‐replacing events, with decreases in the year of disturbance followed by rapid recovery. NEP decreased in a mangrove ecosystem following Hurricane Wilma because of a decrease in GPP and an increase in ER. Citation: Amiro, B. D., et al. (2010), Ecosystem carbon dioxide fluxes after disturbance in forests of North America, J. Geophys. Res., 115, G00K02, doi:10.1029/2010JG001390. 1Department of Soil Science, University of Manitoba, Winnipeg, Manitoba, Canada. 2Environment Canada, Saskatoon, Saskatchewan, Canada. 3Everglades National Park, Homestead, Florida, USA. 4Faculty of Land and Food Systems, University of British Columbia, Vancouver, British Columbia, Canada. 5School of Forest Resources and Conservation, University of Florida, Gainesville, Florida, USA. 6Department of Earth, Ecological, and Environmental Sciences, University of Toledo, Toledo, Ohio, USA. 7USDA Forest Service, New Lisbon, New Jersey, USA. 8Department of Meteorology, Pennsylvania State University, University Park, Pennsylvania, USA. 9Department of Atmospheric and Oceanic Sciences, University of Wisconsin‐Madison, Madison, Wisconsin, USA. 10College of Engineering, Forestry, and Natural Sciences, Northern Arizona University, Flagstaff, Arizona, USA. 11Department of Environmental Science, Policy, and Management, University of California, Berkeley, California, USA. 12Department of Earth System Science, University of California, Irvine, California, USA. 13Canadian Forest Service, Fredericton, New Brunswick, Canada. 14College of Forestry, Oregon State University, Corvallis, Oregon, USA. 15Faculty of Forestry and Geomatics, Université Laval, Quebec, Quebec, Canada. 16Department of Geography, Queen’s University at Kingston, Kingston, Ontario, Canada. 17Deceased 5 March 2010. 18Department of Forestry and Environmental Resources, North Carolina State University at Raleigh, Raleigh, North Carolina, USA. 19Department of Biological Sciences, University of Alabama, Tuscaloosa, Alabama, USA. 20Complex Systems Research Center, University of New Hampshire, Durham, New Hampshire, USA. Copyright 2010 by the American Geophysical Union. 0148‐0227/10/2010JG001390 JOURNAL OF GEOPHYSICAL RESEARCH, VOL. 115, G00K02, doi:10.1029/2010JG001390, 2010 G00K02 1 of 13 1. Introduction [2] Disturbances are an important feature of North Amer- ican forests, renewing forest stands or changing the vegeta- tion dynamics following less severe disturbances. Wildland fires burn an average of about 40,000 km2 in Canada and the U.S. annually [Stocks et al., 2003; U.S. National Interagency Forest Fire Centre, http://www.nifc.gov/fire_info/fires_acres. htm, accessed 25 March 2010] with large interannual vari- ability. About 50,000 km2 of forested land is harvested annually in Canada and the U.S., with about half of this as clear‐cut harvest [Kurz and Apps, 1999; Birdsey et al., 2006]. Moderate to severe insect infestations are highly variable among years but can affect a greater area than either fire or harvesting [Kurz and Apps, 1999; Birdsey et al., 2006]. Hurricanes and tornados are estimated to affect about 17,000 km2 annually in the U.S. [Dale et al., 2001] and there are many additional less severe storms that cause windthrow in forests. These disturbances have a large effect on the carbon (C) balance of North American forests. For example, analyses for Canadian forests showed clearly that insects and fire have caused a net forest C loss that fluctuates annually, but has a lasting legacy [Kurz and Apps, 1999; Kurz et al., 2008a]. Disease and storms also affect the C balance, but their impacts are often difficult to quantify because they result in decreased growth without initial cat- astrophic removals of C. The net C effects of disturbances have been modeled extensively with some of the model outputs being compared to direct eddy covariance mea- surements of C flux over daily to annual scales [e.g., Grant et al., 2007]. These direct determinations of net ecosystem exchange (NEE) of carbon dioxide (CO2) have helped us understand the processes controlling C exchange at indi- vidual sites following fire [e.g., Goulden et al., 2006; Welp et al., 2007; Dore et al., 2008; Mkhabela et al., 2009], harvesting [e.g., Chen et al., 2002; Schwarz et al., 2004; Misson et al., 2005; Giasson et al., 2006; Noormets et al., 2007; Krishnan et al., 2009; Zha et al., 2009], insect infesta- tions [Cook et al., 2008; Clark et al., 2010; Brown et al., 2010], and storms [Barr et al., 2010]. [3] In North America, there has been collaboration among researchers making eddy covariance measurements [e.g., Baldocchi et al., 2001] through the AmeriFlux and Cana- dian Carbon Program (Fluxnet‐Canada Research Network) networks. In particular, the Canadian network had a dis- turbance focus [Margolis et al., 2006]. The North American Carbon Program (NACP) [Wofsy and Harriss, 2002] pro- vides a framework for bringing some of these measurements together to investigate some general relationships between disturbance and forest C exchange across North America. Eddy covariance flux towers have not yet been established along disturbance chronosequences in Mexico. Our goal in the present paper is to synthesize the measurements for lo- cations where tower clusters have been used to measure NEE of forest chronosequences after different types of disturbances. [4] One option to investigate the effects of disturbance is to measure an ecosystem for several years prior to the distur- bance, and then follow with long‐term continuous observa- tions to understand the ecosystem response and recovery following the disturbance. Such experiments are not always possible, especially considering the very long times needed to measure ecosystem development over decades or centu- ries. Hence investigators often study sites of different ages in parallel to infer the status of an ecosystem as a function of age (time since disturbance). This chronosequence approach is a practical design to acquire age‐related data in a short period of time. This is important when employing eddy covariance because we do not have an archive of compar- ative measurements taken over several decades to study the effects of forest development on the net CO2 flux. However, there are now some relatively long‐term eddy covariance sequences that have run well over a decade at mature forest sites [Barr et al., 2007; Dunn et al., 2007; Urbanski et al., 2007]. [5] A conceptual trajectory of C flux following a distur- bance, based on vegetation development [Odum, 1969] and dynamics of decomposition, is that the predisturbance eco- system is fixing C at some rate, and then there is an immediate C loss as a result of the disturbance, with sub- sequent recovery over some period [e.g., Kashian et al., 2006]. This loss is caused by death of photosynthesizing vegetation, biomass combustion in the case of fire, harvest removals, or insect herbivory. In cases where insects kill trees (e.g., bark beetles) or a storm causes tree defoliation and mortality, there is usually minimal initial export of C from the ecosystem. However, following most disturbances, changes to the ecosystem have the potential to alter NEE. This could be driven by changes to heterotrophic respiration when forest C pools change, coupled with decreased pho- tosynthesis because of less leaf area. The ecosystem can change quite quickly as new growth follows the disturbance [Chen et al., 2002; Gough et al., 2007]. These general- izations also depend on the severity of the disturbance, with stand‐replacing disturbances having different trajectories than less severe partial disturbances (e.g., stand thinning). A fundamental difference between harvesting and fire is that harvesting removes the coarse woody material, whereas fire removes the fine materials (forest floor and fine fuels). Such a difference likely impacts the postdisturbance respiration. Fire may also mineralize nutrients more rapidly, making them available for vegetation uptake. Insect infestations usually do not totally replace forest stands although there are exceptions in single‐species forests [e.g., Kurz et al., 2008b]. Similarly, some storms can be catastrophic caus- ing high immediate losses of C [Lindroth et al., 2009]. Typically, shorter disturbance intervals result in lesser C stocks [Gough et al., 2008]. [6] Here, we synthesize eddy covariance tower‐based CO2 flux data from forest disturbance chronosequence studies across Canada and the U.S. We have included eco- systems affected by fire, harvesting, and major insect and storm (hurricane) events. There are many more chronose- quence studies that have measured C fluxes using chambers [e.g., Bond‐Lamberty et al., 2003; Irvine et al., 2007] or biometric sampling [e.g., Bond‐Lamberty et al., 2004; Campbell et al., 2004, 2009; Gough et al., 2008]. However, our synthesis of only tower‐based eddy covariance data allows us to compare whole ecosystem fluxes using similar methodologies and footprints. Further, we have only included studies where annual Net Ecosystem Production (NEP) has been measured. This excludes many studies where data are only available from short (usually summer) field campaigns or partial years. We have used data from AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 2 of 13 forested sites that are less than 100 years old to concentrate on younger ecosystems. We recognize that this period is less than the disturbance cycle in some areas (e.g., the boreal fire cycle tends to be slightly longer than this on average) so that we are not integrating to estimate the net C budget for a full disturbance cycle. Also, there have been several previous syntheses that have concentrated on the C dynamics of old‐ growth forests [Carey et al., 2001; Luyssaert et al., 2008]. We have not included the C removals during the disturbance event, such as biomass combustion, harvest removals, or insect migration, and only measure the postdisturbance effects. In this paper we focus on the following broad hypotheses to reconcile the responses of a broad range of forests to disturbance using the existing eddy covariance tower flux data: (1) Disturbances decrease NEP because of decreased Gross Primary Production (GPP) with little effect on Ecosystem Respiration (ER); (2) NEP recovers similarly following stand‐replacing fire and harvesting (null hypoth- esis); (3) Non‐stand‐replacing disturbances such as insects, storms, and silvicultural thinning have very short‐term effects on NEP. 2. Methods 2.1. Site Descriptions 2.1.1. Fire Chronosequences [7] There are four major fire chronosequences included in this synthesis (Table 1). Three are located in the boreal forest in Alaska, Saskatchewan and Manitoba. The Alaska and Manitoba sites represent more northerly parts of the boreal forest dominated by black spruce (Picea mariana), whereas the Saskatchewan sites are at the southern fringe and have components of black spruce, jack pine (Pinus banksiana) and trembling aspen (Populus tremuloides). The Manitoba and Saskatchewan locations correspond to the northern and southern study areas, respectively, and were investigated during the BOREAS experiment [Sellers et al., 1997]. However, the fire chronosequences were established in the late 1990s following the completion of BOREAS. At both of these chronosequences, long‐term flux towers at mature sites anchor the longer‐term flux record [Dunn et al., 2007; Kljun et al., 2006]. At all the boreal fire chron- osequences, different forest stand ages were selected in a reasonably close geographic area, typically within 100 km. However, despite attempts to match site conditions, some of the variability among sites within a chronosequence will be caused by local environmental factors such as differences in soil type and hydrology. The fire chronosequence in Arizona captured the effects of a stand‐replacing wildfire in a high‐ elevation semiarid ecosystem dominated by ponderosa pine (Pinus ponderosa). 2.1.2. Harvest Chronosequences [8] Most of our harvest chronosequences were stand replacing, involving harvest with most of the larger trees removed and perhaps some protective vegetation left for regeneration [e.g., Giasson et al., 2006]. However, slash may also have been left at the site, depending on local harvesting practices (e.g., New Brunswick site). In the boreal forest, harvest chronosequences were measured in Saskatchewan and Quebec. In both cases, mature (control) forests were fire generated because harvesting is a relatively new activity. The Saskatchewan site represents a southern boreal condition with jack pine, whereas the Quebec site is a northern boreal black spruce site. These sites also differ by longitudinal gradient with the more eastern Quebec site receiving greater annual precipitation. [9] In temperate forests, we have harvest chronosequences from several sites in the Wisconsin‐Michigan northern hardwoods forest area. These sites are part of the larger ChEAS experiment [Chen et al., 2008]. Essentially all stands less than 100 years old in this area regenerated fol- lowing harvesting. There are several different forest stand types, ranging from jack pine to aspen and maple‐basswood‐ ash (Acer‐Tilia‐Fraxinus spp.) deciduous species mixes. There are two temperate forest chronosequences on the west coast, located on Vancouver Island and in Oregon, representing Douglas fir (Pseudotsuga menzesii) and pon- derosa pine forests, respectively. The clear‐cut site in New Brunswick is a balsam fir (Abies balsamea) forest. The Florida chronosequence is a fast‐growing slash pine (Pinus elliottii var. elliottii) plantation managed on a rotation of 20–25 years. [10] Thinning treatments were conducted in a balsam fir forest in New Brunswick and ponderosa pine forests in Arizona and California. Measurements were made prior to thinning and then followed after the treatment. These thin- nings are part of local forest management and represent treatments that are employed routinely. In Arizona, the slash from thinning was piled and burned the first year following thinning; this was undertaken to reduce forest fuels and the risk of an intense fire. 2.1.3. Insect Chronosequences [11] There are three insect chronosequences where infesta- tions have been sufficiently severe to cause major defolia- tion or tree death. The mountain pine beetle (Dendroctonus ponderosae) site in British Columbia was established spe- cifically to measure the effects of the beetle, which has been killing large areas of lodgepole pine (Pinus contorta) forest [Kurz et al., 2008b]. The affected areas have almost 100% tree mortality, but there is new growth by early successional species. The site in New Jersey experienced a severe defo- liation by gypsy moth (Lymantria dispar) in 2007. Several forest types were measured that represented various mix- tures of pine (Pinus rigida, P. echinata) and oak (Quercus velutina, Q. prinus, Q. alba, Q. marlandica, Q. ilicifolia). The Willow Creek site in Wisconsin, although established as part of a harvest chronosequence, experienced a severe defoliation by forest tent caterpillar (Malacosoma disstria) in 2001. This allowed an opportunity to study the effects of defoliation of hardwoods (Tilia americana, Fraxinus penn- sylvanica, Quercus rubra), although there was a second growth of leaves following defoliation (i.e., most trees were not killed). 2.1.4. Storm Chronosequence [12] Although severe storms take a toll on North American forests through tree windthrow, there have been few oppor- tunities to measure the effects using flux towers. A study in a mangrove ecosystem in southwest Everglades National Park, Florida quantified the impact of Hurricane Wilma (October 2005) on mangrove forest NEE. Species included red (Rhizophora mangle), black (Avicennia germinans), and white (Laguncularia racemosa) mangroves. The storm was severe enough to defoliate the forest crown and destroyed about 30% of the mangrove trees. Flux tower instruments AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 3 of 13 T ab le 1. D es cr ip tio n of S ite sa C hr on os eq ue nc e L oc at io n (L at itu de , L on gi tu de ) V eg et at io n T yp eb Y ea r of M os t R ec en t D is tu rb an ce Y ea rs M ea su re d A nn ua l M ea n A ir T em pe ra tu re (° C ) A nn ua l T ot al P re ci pi ta tio n (m m ) R ef er en ce fo r D at a S ite N am e U se d in R ef er en ce F ir e C hr on os eq ue nc es A la sk a 63 .9 N , 14 5. 7W gr ou nd ve ge ta tio n 19 99 19 99 , 20 02 −1 .9 30 0 R an de rs on et al . [2 00 6] ; W el p et al . [2 00 6, 20 07 ] B n3 63 .9 N , 14 5. 4W bs , ta , w ill ow 19 87 20 02 –2 00 4 B n2 63 .9 N , 14 5. 7W bs 19 20 20 02 –2 00 4 B n1 S as ka tc he w an 53 .9 N , 10 6. 1W jp , bs , ta 19 98 20 01 –2 00 6 0. 4 47 0 A m ir o et al . [2 00 6] ; M kh ab el a et al . [2 00 9] F 98 54 .3 N , 10 5. 9W jp , ta 19 89 20 02 –2 00 5 F 89 54 .5 N , 10 5. 8W jp 19 77 20 04 –2 00 6 F 77 53 .9 N , 10 4. 7W jp 19 19 20 00 –2 00 7 K lju n et al . [2 00 6] O JP M an ito ba 56 .6 N , 98 .2 W ta , w ill ow , bs 20 03 20 04 –2 00 5 −3 .2 52 0 G ou ld en et al . [2 00 6] ; M cM ill an et al . [2 00 8] N S A 20 03 56 .6 N , 98 .2 W ta , w ill ow , bs 19 98 20 03 –2 00 5 N S A 19 98 55 .9 N , 99 .0 W ta , w ill ow , bs 19 89 20 03 –2 00 5 N S A 19 89 55 .8 N , 98 .4 W ta , w ill ow , bs 19 81 20 02 –2 00 5 N S A 19 81 55 .8 N , 98 .4 W bs , jp , ta 19 64 20 02 –2 00 5 N S A 19 64 55 .8 N , 98 .5 W bs 19 30 20 02 –2 00 4 N S A 19 30 A ri zo na 35 .1 N , 11 1. 8W pp 19 96 20 06 –2 00 8 7. 9 58 0 D or e et al . [2 01 0] F w f 35 .1 N , 11 1. 8W pp 19 19 20 06 –2 00 8 D or e et al . [2 00 8, 20 10 ] F uf H ar ve st C hr on os eq ue nc es S as ka tc he w an 53 .9 N , 10 4. 6W gr ou nd ve ge ta tio n, jp 20 02 20 04 –2 00 5 0. 4 47 0 K lju n et al . [2 00 6] ; M kh ab el a et al . [2 00 9] ; Z ha et al . [2 00 9] H JP 02 53 .9 N , 10 4. 7W jp 19 94 20 04 –2 00 5 H JP 94 53 .9 N , 10 4. 6W jp 19 75 20 04 –2 00 5 H JP 75 53 .9 N , 10 4. 7W jp 19 19 20 00 –2 00 7 O JP Q ue be c 49 .3 N , 74 .0 W gr ou nd ve ge ta tio n, bs 20 00 ha rv es t, 20 03 sc ar if ie d 20 02 –2 00 8 0 96 0 G ia ss on et al . [2 00 6] ; B er ge ro n et al . [2 00 8] H B S 00 49 .8 N , 74 .6 W bs 19 75 20 08 H . A . M ar go lis , pe rs on al co m m un ic at io n, 20 09 H B S 75 49 .7 N , 74 .3 W bs 19 15 20 04 –2 00 8 B er ge ro n et al . [2 00 7, 20 08 ] E O B S V an co uv er Is la nd 49 .9 N , 12 5. 2W df 20 00 20 01 –2 00 8 8. 6 14 50 H um ph re ys et al . [2 00 6] ; K ri sh na n et al . [2 00 9] H D F 00 49 .5 N , 12 4. 8W df 19 88 20 02 –2 00 8 H D F 88 49 .4 N , 12 5. 3 df 19 49 19 98 –2 00 8 D F 49 N ew B ru ns w ic k 46 .5 N , 67 .1 W bf 20 04 20 05 –2 00 8 3. 4 11 90 M . B . L av ig ne , pe rs on al co m m un ic at io n, 20 09 N L ‐C C 46 .5 N , 67 .1 W bf 19 75 , th in ne d in 19 91 an d 20 05 20 04 –2 00 8 N L ‐C T AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 4 of 13 T ab le 1. (c on tin ue d) C hr on os eq ue nc e L oc at io n (L at itu de , L on gi tu de ) V eg et at io n T yp eb Y ea r of M os t R ec en t D is tu rb an ce Y ea rs M ea su re d A nn ua l M ea n A ir T em pe ra tu re (° C ) A nn ua l T ot al P re ci pi ta tio n (m m ) R ef er en ce fo r D at a S ite N am e U se d in R ef er en ce W is co ns in 45 .9 N , 90 .1 W ta , ha rd w oo ds 20 01 20 07 5. 2 76 5 A . R . D es ai , pe rs on al co m m un ic at io n, 20 09 R ile y C re ek 45 .8 N , 90 .1 W ha rd w oo ds 19 30 20 00 –2 00 6 C oo k et al . [2 00 4, 20 08 ] W ill ow C re ek 45 .5 N , 84 .7 W ha rd w oo ds 19 20 19 99 –2 00 3 G ou gh et al . [2 00 8] U M B S 46 .6 N , 91 .1 W rp 19 39 20 02 –2 00 3 N oo rm et s et al . [2 00 7, 20 09 ] M R P 46 .6 N , 91 .1 W ha rd w oo ds 19 99 20 02 Y H W 46 .6 N , 91 .1 W jp , rp 19 95 20 02 Y R P 46 .6 N , 91 .1 W ha rd w oo ds 19 92 20 03 IH W 46 .6 N , 91 .1 W rp 19 82 20 03 IR P 41 .6 N , 83 .8 W oa k 19 47 20 04 –2 00 5 9. 2 84 0 N oo rm et s et al . [2 00 8] T O L O re go n 44 .4 N , 12 1. 5W pp 19 84 20 04 –2 00 5 7. 6 46 0 L aw et al . [2 00 3] ; Sc hw ar z et al . [2 00 4] ; T ho m as et al . [2 00 9] M e3 44 .4 N , 12 1. 6W pp 19 78 20 01 –2 00 2 7. 6 60 0 M e5 44 .5 N , 12 1. 6W pp 19 12 20 02 –2 00 8 7. 1 60 0 M e2 F lo ri da 29 .8 N , 82 .2 W sp 19 72 , 19 98 19 96 –2 00 7 20 .6 12 30 C la rk et al . [2 00 4] ; B in fo rd et al . [2 00 6] S P 2 29 .8 N , 82 .2 W sp 19 90 19 99 –2 00 5 S P 3 A ri zo na 35 .4 N , 11 1. 8W pp 20 06 th in ne d 20 06 –2 00 8 7. 9 58 0 D or e et al . [2 01 0] F M F C al if or ni a 38 .9 N , 12 0. 6W pp 19 90 , 20 00 th in ne d 19 99 –2 00 2 11 .9 12 90 M is so n et al . [2 00 5] B lo dg et t In se ct C hr on os eq ue nc es N ew Je rs ey G yp sy M ot h 39 .9 N , 74 .6 W oa k, pi ne 19 14 , in se ct s 20 07 20 05 –2 00 8 12 .3 11 55 C la rk et al . [2 01 0] S L T 39 .8 N , 74 .4 W pi ne , oa k 19 95 , in se ct s 20 07 –2 00 8 20 05 –2 00 8 C E D B ri tis h C ol um bi a M ou nt ai n P in e B ee tle 54 .5 N , 12 2. 7W lp 20 03 20 07 –2 00 8 2. 3 65 5 B ro w n et al . [2 01 0] M P B 03 55 .1 N , 12 2. 8W lp 20 06 20 07 –2 00 8 M P B 06 W is co ns in F or es t T en t C at er pi lla r 45 .8 N , 90 .1 W ha rd w oo ds 19 30 , in se ct s 20 01 20 00 –2 00 6 5. 2 76 5 C oo k et al . [2 00 8] W ill ow C re ek H ur ri ca ne F lo ri da 25 .4 N , 81 .1 W m an gr ov es 19 92 , hu rr ic an e 20 05 20 04 , 20 07 –2 00 9 23 .8 13 90 B ar r et al . [2 01 0] S ha rk R iv er a O nl y si te s le ss th an 10 0 ye ar s ol d w er e us ed in th is an al ys is .A nn ua lm ea n ai r te m pe ra tu re an d to ta lp re ci pi ta tio n (r ou nd ed to ne ar es t5 m m ) da ta ar e 30 ye ar no rm al s ta ke n fr om th e ne ar es tm et eo ro lo gi ca ls ta tio n. b S pe ci es ar e as fo llo w s: jp , P in us ba nk si an a; bs , P ic ea m ar ia na ; ta , P op ul us tr em ul oi de s; pp , P in us po nd er os a; df , P se ud ot su ga m en ze si i; sp , P in us el lio tti i; rp , P in us re si no sa ; lp , P in us co nt or ta . AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 5 of 13 and data acquisition were also destroyed by the storm. Researchers reestablished the study field site in October 2006. Hence, measurements here are a true sequence with predisturbance and postdisturbance measurements. 2.2. Field Measurements and Data Processing [13] The details of each individual field measurement and annual data processing protocol are given in the references listed in Table 1. Note that for this synthesis, each investi- gator supplied their best estimates of their annual totals of C fluxes for each site. This makes these measurements con- sistent with previous publications from each site. NEE was calculated every 30min using the covariance of a fast‐response infrared gas analyzer for CO2 and a three‐dimensional sonic anemometer for wind velocity. The gas analyzers were either closed‐path or open‐path systems. Although there was some variability among researchers, typical data pro- cessing included corrections for nonzero mean vertical wind velocity, tubing losses (closed‐path analyzers) and density effects (especially for open‐path analyzers). Known pro- blems with sensor heating of open‐path analyzers [Burba et al., 2008] were either corrected, or were compensated by excluding cold weather data [Welp et al., 2007] or filling cold‐weather periods with modeled respiration [Mkhabela et al., 2009]. [14] NEP was calculated as the gap‐filled annual sum of NEE, defined so that a downward flux is positive (i.e., terrestrial ecosystem gain is positive). ER (positive flux is upward) was calculated based on nighttime NEE measure- ments combined with regression and modeling. GPP (downward flux is positive) was calculated as the sum of NEP and ER, sometimes combined with light‐response modeling to fill gaps [e.g., Moffat et al., 2007]. [15] In a multiple‐site synthesis, variability among indi- vidual field techniques, annual data processing, and algo- rithms for GPP and ER contribute additional uncertainty [Desai et al., 2008]. We appreciate the general need for a greater degree of homogeneity in postprocessing flux data, however the site investigators are contributors to either the AmeriFlux or Canadian Carbon Program networks, and some standardization has been achieved through these col- laborations. Annual NEP is likely to be estimated within about ± 25 g C m−2 y−1 based on uncertainty in gap‐filling techniques [Moffat et al., 2007]. However, there is some additional uncertainty caused by random error of about ± 20 g C m−2 y−1 [Richardson and Hollinger, 2005]. Typ- ically, we would expect annual NEP to be estimated to better than ± 50 g C m−2 y−1 at most sites, with better estimates at certain sites. Baldocchi [2008] gives a good review of the potential issues that affect flux tower measurements. 3. Results and Discussion 3.1. Fire Chronosequences [16] NEP, GPP and ER showed similar trends with time since disturbance for all of the four fire chronosequences (Figure 1). All sites that are less than 10 years of age were C sources. The three boreal chronosequences (Saskatchewan, Manitoba, Alaska) became net C sinks after about 10 years. During this period of positive NEP, there was substantial interannual variability at any given site, often ranging about 100 g C m−2y−1. This magnitude of interannual variability seemed to be similar at any given location, even when there were only three site years. Considering this variability, the NEP response with time for sites older than 10 years of age was relatively invariant with time. The Saskatchewan chronosequence at about 70 years of age included data from the southern old jack pine site, which tended to have lower NEP than at the nearby old aspen site [Kljun et al., 2006], and likely does not fully represent the successional endpoint of the young postfire forests. At about 30 years of age, the Saskatchewan chronosequence showed three site years with consecutive negative NEP. This site had visibly decaying coarse woody debris lying on the surface contributing to ER. The Arizona site at about 10 years of age had similar NEP to the slightly younger boreal sites. Trees had still not regenerated at the Arizona site, and a sparse grassland had succeeded (seasonal maximum leaf area index = 0.6) with a large amount of coarse woody debris present [Dore et al., 2008]. We do not have intermediate‐aged forest data for Arizona, so the comparative trajectory is not clear even though the older Arizona site (about 90 years of age) had a similar positive NEP to the boreal sites. However, the postfire Figure 1. Annual (a) NEP, (b) GPP, and (c) ER for the fire chronosequences. AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 6 of 13 vegetation dynamics are very different at the Arizona site compared to the boreal sites, and it could take many years to reach positive NEP following fire at the Arizona site. [17] Separation of NEP into the GPP and ER components helps to identify the relative driving forces. GPP clearly increased with stand age for the first 20–30 years (Figure 1b). The Saskatchewan sites had the highest GPP in the 15– 30 year age range. These sites had a warmer climate than the Manitoba or Alaska sites and had a fast‐growing deciduous component of trembling aspen. After 30 years, the three boreal chronosequences tended to show similar GPP. The Arizona site had similar GPP to the boreal sites at about 10 years, but greater GPP at about 90 years. A possible explanation is that the Arizona chronosequence was more moisture limited for vegetation establishment following fire but the older stand benefited from the warmer climate, compared to the boreal sites. [18] Ecosystem respiration had a less well‐defined tra- jectory with age than either NEP or GPP, although it is slightly lower at very young ages (Figure 1c). With the exception of the Saskatchewan sites at 15 and 30 years, and the 90 year old Arizona site, ER did not vary much with age. The high ER at the Arizona site and at some Saskatchewan sites is consistent with the high GPP at these sites. However, the negative NEP at the 30 year old site indicates that het- erotrophic respiration is an important factor, likely because of decaying coarse woody debris. It is important to note that the Saskatchewan sites at 15 and 30 years used open‐path gas analyzers that have known problems with sensor heating [Burba et al., 2008]. Use of these analyzers without heating corrections will underestimate ER. In this study, data col- lected when temperatures were below 0°C were excluded and replaced using respiration estimated as a function of soil temperature. Although this increases the uncertainty in ER, the high ER was mostly caused by summertime losses at these sites, likely enhanced by decomposition of coarse woody debris [Mkhabela et al., 2009]. 3.2. Harvest Chronosequences [19] NEP trajectories following harvesting showed similar temporal changes to those after fire, although there was more variability caused by location (Figure 2a). In particular, sites in warmer climates had higher NEP, and the Florida trajectory clearly showed faster recovery following harvest. The greatest C loss in the early years following harvesting was also at warmer sites, which caused a greater dyna- mism in the C flux than at colder sites. The boreal sites (Saskatchewan, Quebec) had reduced amplitude throughout the trajectory. The scale of Figure 2a used to accommodate all ecosystems diminishes the relative interannual variability of the boreal sites. This comparison showed greater absolute interannual variability in C flux for warmer sites than colder sites in North America. However, the overall interannual variability effect needs to be integrated spatially over all North American forests to evaluate its importance. [20] Similarly to NEP, GPP showed clear ecosystem dif- ferences among geographic locations (Figure 2b). The much larger fluxes for Florida, Vancouver Island and Oregon dwarfed the boreal trajectories. However, for all chron- osequences, a clear GPP recovery occurred within about the first 20 years following harvest. The boreal, Wisconsin, and Florida GPP values were relatively flat after 20 years, whereas the west coast (Vancouver Island, Oregon) eco- systems showed an increase over time. This western forest increase is consistent with continued C accumulations observed in biometric and inventory data [Law et al., 2004; Hudiburg et al., 2009]. Lower photosynthetic assimilation fluxes were also documented by Buchmann and Schulze [1999] for forest stands less than 20 years of age. As in the case of fire, ER had a relatively flat trajectory, with some reduction in ER in the very first years following harvest at some sites (Figure 2c). 3.3. Respiration Components for Fire and Harvest Chronosequences [21] The large differences in C fluxes among geographic regions are largely climate driven. Process models can help explain these differences, but there is no simple climate scaling variable such as normalization by annual tempera- ture alone when we consider the chronosequence. For example, at very young (<4 years old) harvested sites, there were similar values of NEP and GPP for sites with different annual temperatures, although ER was much better sepa- Figure 2. Annual (a) NEP, (b) GPP, and (c) ER for the harvest chronosequences. AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 7 of 13 rated (Figure 2, Table 1). However, we would expect C allocation to scale in some fundamental ways. For example, the ratio of GPP to ER clearly showed low values at early years following disturbance (Figure 3) for both postfire and postharvest sites. As discussed for NEP, GPP/ER became greater than unity (i.e., NEP > 0) by 20 years for all chronosequences and by 10 years for most. More importantly, GPP/ER did not vary much with age following this initial increase for sites generated by either harvest or fire, and had an asymptote of 1.23. This is consistent with a global analysis of flux data where most of the forested sites were mature, which gave an average ratio of GEP/ER = 1.2 [Law et al., 2002]. [22] ER can be separated into autotrophic (Ra) and heterotrophic (Rh) respiration components. Although we usually do not have independent measures for each of these, it is instructive to examine the chronosequence data by assuming some relationship between Ra and GPP. We rec- ognize that the Ra/GPP ratio can vary [DeLucia et al., 2007; Piao et al., 2010], especially among our large range of ecosystem types and climates. However, a range of studies estimate Ra to be about 0.55·GPP [e.g., Landsberg and Waring, 1997; Jassal et al., 2007], which would yield GPP/ER = 1.82 if Rh = 0 (horizontal line in Figure 3). All points were below this maximum because Rh > 0, and the relative effect of Rh was clearly greatest in the first 10–20 years following disturbance (i.e., Ra is low). When GPP = ER (i.e., NEP = 0), Rh = 0.45·ER. At the lowest measured value of GPP/ER of about 0.2 in Figure 3, Rh was about 90% of ER. This corresponded to the sites where GPP was low immediately following harvest (Figure 2b), with most of this GPP likely being from ground vegetation contributions before new trees were established. [23] We can carry the respiration analysis further, again assuming a constant fraction of Ra to GPP (0.55), recog- nizing that this may change with forest age and species [DeLucia et al., 2007]. In Figure 4, we plot the derived Rh as a function of age. Here we have normalized the Rh for each site by the mean value for all sites in the local chronose- quence (e.g., each Saskatchewan fire site year Rh is divided by the mean value of all Saskatchewan fire site years) to allow comparison on the same scale. Although there were some higher Rh values at young sites, there was no signif- icant regressional relationship with age (y = 0.98 + 0.0004 x; r2 = 0.003). Splitting the sites into groups of > 10 and ≤ 10 years of age also showed no statistical difference (Student’s t test P = 0.5). Hence, it appears that Rh is largely invariant with age for the full data set, even though there are likely trends in some of the individual chronosequences. This is evident for some of the youngest harvested sites and for the 30 year old Saskatchewan fire site, which had high Rh. [24] The data support the hypothesis that disturbances decrease NEP mostly because of decreased GPP. Fire and harvesting slightly decreased ER soon after the disturbance, but ER then either increased over time or did not vary. Figure 3 also supports the hypothesis that NEP recovery is similar following stand‐replacing fire and harvesting, at least in a broad sense. However, there are differences at any given location, as has been demonstrated for individual chronosequences [Mkhabela et al., 2009]. 3.4. Insects, Storms, and Thinning [25] The studies of insects, storms, and thinning followed a true sequence of years that encompass the disturbance at a given site. The strength of these measurements is that there is no spatial variability when a true chronosequence is fol- lowed. However, interannual variability caused by climate fluctuations needs to be assessed using a control site [e.g., Dore et al., 2010]. Although the time period (i.e., age effect) was shorter, the continuous measurements demonstrated the nature of the disturbance effect. NEP following gypsy moth and forest tent caterpillar infestations showed a decrease in the year of infestation, with a clear recovery in subsequent Figure 3. The ratio of annual GPP/ER with stand age for fire and harvest sites. The horizontal line at GPP/ER = 1.82 corresponds to a constant autotrophic respiration value of 0.55·GPP and no heterotrophic respiration. The dashed line is the regression for all data points where y = 1.23 (1 − e−0.224x), r2 = 0.60, and n = 162. Figure 4. Estimated heterotrophic respiration with stand age. We assume that heterotrophic respiration = ER − 0.55·GPP. Heterotrophic respiration values for each site are normalized by the mean value for the local chronose- quence for comparison on the same scale. There is no trend with age (r2 = 0.003). AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 8 of 13 years (Figure 5a). The gypsy moth‐affected sites were slightly more complicated because of 2 years of infestation in the case of one site, and the use of prescribed fire in some stands. The oak‐pine site was not defoliated in 2005 and 2006, but was completely defoliated in 2007. The pitch pine‐scrub oak site was not defoliated in 2005 and 2006, and understory oaks and some shrubs were defoliated in 2007. In Figure 5a, we have included data from the oak‐pine and pitch pine‐scrub oak stands [Clark et al., 2010]. We have not included the data in the year following infestation at the pitch pine‐scrub oak site, which experienced pre- scribed fire in March 2008 and could have complicated the response due to insect disturbance. [26] The mountain pine beetle sequence included two sites: one site had data for the first 2 years following infestation, whereas the other was measured 4 and 5 years after. For all four site years, NEP was slightly negative becoming neutral by year five, whereas we would expect these 80 to 100 year old lodgepole pine stands to have been net C sinks (there is no pre‐effect measurement site). During the year of gypsy moth and forest tent caterpillar infesta- tions, both GPP (Figure 5b) and ER (Figure 5c) decreased. GPP increased following the infestation year, however ER continued at a lesser rate than it was prior to the forest tent caterpillar infestation. We recognize that differences can also be caused by interannual variability in climate, which is difficult to assess with a single sequence. [27] The effect of Hurricane Wilma in Florida is shown where a complete year (2004) and a partial year (2005, with November and December gapfilled) were measured before the October 2005 event [Barr et al., 2010], and then mea- surements resumed in 2007. A decrease in NEP was mostly attributed to an increase in ER, presumably because of more woody debris available for decomposition. Also, an increase in soil temperature of 1 to 3°C resulting from increased radiation penetration to the surface may have contributed to higher heterotrophic respiration following the storm. The high NEP values at this Florida mangrove site were driven by a high GPP without a correspondingly high ER. This was likely caused by a net lateral efflux of particulate and dis- solved organic C that was not available for local respiration. In addition, a portion of the respired CO2 was exported with the outgoing tide as dissolved inorganic C and released to the atmosphere outside of the tower flux footprint. [28] The thinned site in Arizona showed a decrease in NEP during the treatment year, recovering in the following year. ER at this site increased through the 3 years of mea- surement and GPP decreased in the year of thinning and then increased in the following year. The California site, which was also ponderosa pine, showed a similar NEP and GPP trajectory to the Arizona site, being carbon neutral in the thinning year. ER increased 2 years after the thinning treatment. This site was quite young, having been planted in 1990. The New Brunswick site had some thinning con- ducted 14 years prior to the commercial thinning. Following the commercial thinning, NEP, GPP and ER all decreased. The short period of record is insufficient to know the longer‐ term effect. More measurements are needed on the effects of forest operation treatments, especially over longer periods [e.g., Gough et al., 2008]. In summary, the results on the effects of insects and thinning support the hypothesis that the C impact is relatively short‐term, with the greatest decrease in the year of disturbance. However, the hurricane effects appear to be longer lasting. 3.5. Implications for Modeling and Further Measurements [29] The observed fluxes exhibited a trajectory charac- terized initially by C losses occurring immediately after the disturbance, followed by a recovery phase of positive NEP. However, the speed of the recovery was important, and all trajectories showed a positive NEP at 20 years following disturbance and most were positive at 10 years. Once the young forests became net C sinks, NEP quickly reached a fairly stable value that remained relatively constant with age, at least for the 100 year limit of the trajectories pre- sented here. The exceptions to this trend were the west coast sites (Vancouver Island, Oregon) which tended to have an increase in NEP over the period of our data sets. This was caused by steadily increasing GPP with relatively little Figure 5. Annual (a) NEP, (b) GPP, and (c) ER for insects, storms, and thinning. The zero year is the year of disturbance. AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 9 of 13 change in ER. The relatively time‐invariant trajectories of NEP in boreal, southern pine and deciduous forests after 20 years were caused by little change in either ER or GPP. For many chronosequences, leaf area index did not change much for forests greater than 20 years of age [e.g.,McMillan et al., 2008]. The GPP/ER stayed approximately constant at between 1 and 1.5 once NEP > 0 for both fire and harvest chronosequences (Figure 3). Heterotrophic respiration was invariant with age (Figure 4), suggesting that this process can be modeled as close to being constant for a chronose- quence at a given geographical location, similar to that reported by Law et al. [2003] and Luyssaert et al. [2008]. [30] Integration of the NEP trajectories over time gives the net C sequestration. When considering the effects of dis- turbance, this integration should be over the period between disturbance events. However, there was sufficient inter- annual variability and differences among sites to make a simple integration of any of the NEP figures difficult. For example, a regression of NEP versus age for the boreal fire chronosequences (Figure 1a) yielded a best fit of NEP = −99.5 + 42.27·ln(age) with r2 = 0.34. We also included three site years at about 160 years of age at the Manitoba site in this regression. Removing the three negative NEP points at about 30 years (the Saskatchewan F77 site) only improved the regression slightly to r2 = 0.42. Using this regression of all the boreal fire sites, the NEP crossover time to a positive value was 10.5 years. However, the potential to have a secondary period of negative NEP caused by woody debris decomposition was evident and should be considered in process models. This is because postfire decay requires the woody material to be wet and support microbial decom- posers, typically through contact with the ground, which lags the disturbance event. Such a lag may not occur with whole‐tree harvest when the finer slash materials left on the postharvest site would have the opportunity to start decomposing without lag. It was difficult to gauge this process in the flux trajectories, but the boreal fire and harvest sites did not appear much different in the very early years, and the most negative NEP was seen at the warmer sites in Florida, New Brunswick, Vancouver Island and Wisconsin (Figure 2a). However, a closer comparison of harvest and fire on an expanded scale show some differences when closely located sites were compared. For example, the fire‐ generated sites had higher GPP and ER than harvest‐ generated sites in Saskatchewan [Mkhabela et al., 2009], whereas in Arizona, the thinned sites had higher GPP and ER than the postfire sites [Dore et al., 2010]. [31] In this synthesis, we examined all the disturbance data on a common scale. This approach helped to show some large generalities, but it is clear that more complete trajectories would help for any given location. The Florida harvest sites were likely the most complete because of their continuity and the relatively quick vegetation growth fol- lowing a disturbance. The Manitoba sites were the most complete fire chronosequence, but given the relatively long fire cycle (typically over 100 years) [Stocks et al., 2003], missing ages still pose questions about whether the trajec- tories are relatively smooth. Measurements at more sites in a given geographic area would help account for spatial vari- ability. However, measurements at the Wisconsin group of sites showed that interannual variability at a given site is of the same magnitude as intersite variability [Noormets et al., 2009]. For recently disturbed sites where NEP and GPP recovery was rapid (e.g., boreal fire), both interannual and spatial variability were relatively less important during this early period compared to later in the trajectory, at least on a 100 year timescale. Although we need to consider the length of each disturbance cycle, these results suggest that good modeling of the period beyond about 10 years is important to get the net rotational C balance because of greater vari- ability during this period. The first 10–20 years is also very important where NEP recovery is slower, such as the Arizona fire or Vancouver Island harvest sites, or where the rotational period is quite short (e.g., Florida harvest). [32] We were fortunate to have some data on the effects of insects and storms, but clearly these were insufficient to understand fully the impacts on C exchange. The continuous sequence data at these sites, and the thinned sites, helped to bracket the nature of the impacts. However, we would expect some of the effects to have a temporal lag. For example, changes in GPP in a given year could affect ER in the following year because of changes to leaf‐litter inputs. This is further complicated by potentially different weather conditions in subsequent years. Despite this complication, a short‐term decrease in NEP was observed in the year of defoliation (Figure 5a). For insect infestations causing tree death, such as for the mountain pine beetle sites, new veg- etation growth appears to be compensating for decomposi- tion of killed trees with the ecosystem being C neutral about 5 years following the attack [Brown et al., 2010]. However, as in the case of fire, it is likely that there will be a second period of enhanced ER when these dead trees fall over and decompose more quickly. Such processes could be docu- mented by selectively measuring annual fluxes in stands where this event has happened. [33] The thinning sites in Arizona, California and New Brunswick illustrate the potential to investigate C dynamics as a result of silvicultural practices. Given that there are many possible ecosystem management options, it is likely that future eddy covariance measurements need to be employed in a diverse range of practices and sites to capture the range of responses [e.g., Vesala et al., 2005]. Although silviculture usually focuses on tree growth performance or fuel reduc- tion, the ability to determine annual NEP for different forest management options would help establish net C benefits ranging from years to decades. The effect of Hurricane Wilma reduced NEP for at least 4 years following the event, and continued data are needed to evaluate the recovery dynamics of severe storms. 4. Conclusions and Recommendations [34] Eddy covariance studies following forest disturbance illustrate that C flux trajectories are consistent with a con- ceptual model of net loss in early years, followed by C gain as the new forest becomes established. The data showed that recovery to a net C sink is relatively rapid in most ecosys- tems investigated, usually occurring within 20 years. The postfire Arizona site is an exception because of succession to grassland without tree establishment. We had fewer measurements following less severe ecosystem changes, such as insect defoliation or forest silvicultural thinning, and these effects need to be assessed more fully and compared to interannual variability caused by climate. We also recognize AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 10 of 13 that our chronosequence measurements are essentially a “snapshot” of recent CO2 fluxes, with older forests having developed under past conditions that had lower ambient CO2 concentrations and a different climate. [35] Although the boreal fire and some of the harvest chronosequence sites have been decommissioned, there is a need to learn more about developmental dynamics, such as the decomposition of coarse woody debris following dis- turbance. It is possible that this is a short‐term respiration flux that occurs in a pulse of a few years in favorable cli- matic conditions, or could be a more continuous process that affects the net flux for several decades as in the semiarid ponderosa pine region [Sun et al., 2004]. We also have poor spatial representation, especially in nonboreal fire‐dominated ecosystems. For example, there are many ecosystems where needleleaf evergreen forests regenerate directly after fire without a broadleaf deciduous successional component. We could then expect a more delayed NEP recovery. Alterna- tively, tree generation can be greatly delayed or absent, as in the Arizona fire chronosequence (Figure 1a). Understory vegetation has shown a compensatory effect on NEP with the thinning of overstory trees [Campbell et al., 2009], so ecosystem studies should include understory measurements to quantify the trajectory of response of ecosystem compo- nents after disturbance. Postharvest trajectories only have a small spatial sample. Even in a single geographical area, there are many possible trajectories. For example, the data from Wisconsin showed considerable spatial variability at times less than 20 years, even though we had few points (Figure 2a). Continuity of data at these sites and addition of nearby sites would help address the issue of spatial and temporal variability. [36] Research opportunities provided by major dis- turbances, such as Hurricane Wilma, provide valuable data for tracking recovery of ecosystems, as changes in pools and fluxes after such events are difficult to model. More observations in disturbed systems are essential for calibrat- ing models that are used to map C balances across regions. At sites that have been shut down, there is an opportunity to revisit these sites again in the future to fill in gaps in the trajectories. The difficulty is that we will likely miss sur- prises, either caused by vegetation dynamics (e.g., changes in species mix), climate/weather, or other forcings (storms, insects, disease). An example is the forest tent caterpillar event in Wisconsin [Cook et al., 2008]. Even with a good model, it would be difficult to plan the best period to revisit sites, or to determine how long we would need to make measurements. Given site logistics, a campaign of at least 5 years would likely be needed at a given site. Hence, revisiting sites will likely be opportunistic, and new dis- turbances (e.g., storms) [Chambers et al., 2007] will also create additional opportunities. In addition to revisiting sites, there are many additional scientific questions that need to be answered. In particular, those tied to a management decision would be a high priority. This involves spending resources for a silvicultural practice, insect and disease control, or suppression or addition of fire. Many of these management options are linked to values other than C, but C is becoming an increasingly valued commodity, and needs to be con- sidered. In addition, experiments should be conducted for each disturbance type to investigate the impacts of climate. This could be done with paired chronosequences with similar disturbances under different climates, which would provide a partial analog for future climate change and allow us to identify differences in disturbance recovery caused by climate. Whole ecosystem studies of the C consequences of these management options using eddy covariance will con- tinue to provide valuable information to assess C sources and sinks in a biosphere undergoing rapid changes resulting from climatic variations and disturbances. [37] Acknowledgments. Data collected at each site were supported by a host of collaborators as outlined in papers describing each site. The AmeriFlux network was supported by the U.S. Department of Energy (e.g., B.E. Law’s AmeriFlux synthesis work was supported by the Office of Science (BER), U.S. DOE grant DE‐FG02‐04ER63911) and National Science Foundation as well as many local funders for individual sites. Similarly, the Fluxnet‐Canada Research Network and Canadian Carbon Program were supported by the Canadian Foundation for Climate and Atmospheric Sciences, NSERC, and the BIOCAP Canada Foundation as well as several local and national organizations (Environment Canada, Canadian Forest Service, and Parks Canada). We thank Eric Kasischke for leading the NACP disturbance synthesis initiative. References Amiro, B. D., et al. (2006), Carbon, energy and water fluxes at mature and disturbed forest sites, Saskatchewan, Canada, Agric. For. Meteorol., 136, 237–251, doi:10.1016/j.agrformet.2004.11.012. Baldocchi, D. (2008), ‘Breathing’ of the terrestrial biosphere: Lessons learned from a global network of carbon dioxide flux measurement sys- tems, Aust. J. Bot., 56, 1–26, doi:10.1071/BT07151. Baldocchi, D. D., et al. (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, doi:10.1175/1520-0477(2001)082<2415:FANTTS>2.3.CO;2. Barr, A. G., T. A. Black, E. H. Hogg, T. J. Griffis, K. Morgenstern, N. Kljun, A. Theede, and Z. Nesic (2007), Climatic controls on the carbon and water balances of a boreal aspen forest, 1994–2003, Global Change Biol., 13, 561–576, doi:10.1111/j.1365-2486.2006.01220.x. Barr, J. G., V. Engel, J. D. Fuentes, J. C. Zieman, T. L. O’Halloran, T. J. Smith III, and G. H. Anderson (2010), Controls on mangrove forest‐ atmosphere carbon dioxide exchanges in western Everglades National Park, J. Geophys. Res., 115, G02020, doi:10.1029/2009JG001186. Bergeron, O., H. A. Margolis, T. A. Black, C. Coursolle, A. L. Dunn, A. G. Barr, and S. C. Wofsy (2007), Comparison of carbon dioxide fluxes over three boreal black spruce forests in Canada, Global Change Biol., 13, 89–107, doi:10.1111/j.1365-2486.2006.01281.x. Bergeron, O., H. A. Margolis, C. Coursolle, and M.‐A. Giasson (2008), How does forest harvest influence carbon dioxide fluxes of black spruce ecosystems in eastern North America?, Agric. For. Meteorol., 148, 537–548. Binford, M. W., H. L. Gholz, G. Starr, and T. A. Martin (2006), Regional carbon dynamics in the southeastern U.S. coastal plain: Balancing land cover type, timber harvesting, fire, and environmental variation, J. Geophys. Res., 111, D24S92, doi:10.1029/2005JD006820. Birdsey, R., K. Pregitzer, and A. Lucier (2006), Forest carbon management in the United States: 1600–2100, J. Environ. Qual., 35, 1461–1469, doi:10.2134/jeq2005.0162. Bond‐Lamberty, B., C. Wang, and S. T. Gower (2003), Annual carbon flux from woody debris for a boreal black spruce fire chronosequence, J. Geophys. Res., 108(D3), 8220, doi:10.1029/2001JD000839. Bond‐Lamberty, B., C. Wang, and S. T. Gower (2004), Net primary pro- duction and net ecosystem production of a boreal black spruce wildfire chronosequence, Global Change Biol., 10, 473–487, doi:10.1111/ j.1529-8817.2003.0742.x. Brown, M., T. A. Black, Z. Nesic, V. N. Foord, D. L. Spittlehouse, A. L. Fredeen, N. J. Grant, P. J. Burton, and J. A. Trofymow (2010), Impact of mountain pine beetle on the net ecosystem production of lodgepole pine stands in British Columbia, Agric. For. Meteorol., 150, 254–264, doi:10.1016/j.agrformet.2009.11.008. Buchmann, N., and E.‐D. Schulze (1999), Net CO2 and H2O fluxes of terrestrial ecosystems, Global Biogeochem. Cycles, 13, 751–760, doi:10.1029/1999GB900016. Burba, G. G., D. K. McDermitt, A. Grelle, D. J. Anderson, and L. K. Xu (2008), Addressing the influence of instrument surface heat exchange on the measurements of CO2 flux from open‐path gas analyzers, Global Change Biol., 14, 1854–1876, doi:10.1111/j.1365-2486.2008.01606.x. AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 11 of 13 Campbell, J. L., O. Sun, and B. E. Law (2004), Disturbance and net eco- system production across three climatically distinct forest landscapes, Global Biogeochem. Cycles, 18, GB4017, doi:10.1029/2004GB002236. Campbell, J., G. Alberti, J. Martin, and B. E. Law (2009), Carbon dynamics of a ponderosa pine plantation following a thinning treatment in the northern Sierra Nevada, For. Ecol. Manage. , 257 , 453–463, doi:10.1016/j.foreco.2008.09.021. Carey, E. V., A. Sala, R. Keane, and R. M. Callaway (2001), Are old forests underestimated as global carbon sinks?, Global Change Biol., 7, 339–344, doi:10.1046/j.1365-2486.2001.00418.x. Chambers, J. Q., J. I. Fisher, H. C. Zeng, E. L. Chapman, D. B. Baker, and G. C. Hurtt (2007), Hurricane Katrina’s carbon footprint on U.S. Gulf Coast forests, Science, 318, 1107, doi:10.1126/science.1148913. Chen, J., M. Falk, E. Euskirchen, K. T. Paw U, U T. H. Suchanek, S. L. Ustin, B. J. Bond, K. Brosofske, N. Phillips, and R. Bi (2002), Biophy- sical controls of carbon flows in three successional Douglas‐fir stands based on eddy‐covariance measurements, Tree Physiol., 22, 169–177. Chen, J., K. J. Davis, and T. P. Meyers (2008), Ecosystem‐atmosphere carbon and water cycling in the temperate northern forests of the Great Lakes Region, Agric. For. Meteorol., 148, 155–157, doi:10.1016/ j.agrformet.2007.08.016. Clark, K. L., H. L. Gholz, and M. S. Castro (2004), Carbon dynamics along a chronosequence of slash pine plantations in northern Florida, Ecol. Appl., 14, 1154–1171, doi:10.1890/02-5391. Clark, K. L., N. Skowronski, and J. Hom (2010), Invasive insects impact forest carbon dynamics, Global Change Biol . , 16 , 88–101, doi:10.1111/j.1365-2486.2009.01983.x. Cook, B. D., et al. (2004), Carbon exchange and venting anomalies in an upland deciduous forest in northern Wisconsin, USA, Agric. For. Meteorol., 126, 271–295, doi:10.1016/j.agrformet.2004.06.008. Cook, B. D., P. V. Bolstad, J. G. Martin, F. A. Heinsch, K. J. Davis, W. Wang, A. R. Desai, and R. M. Teclaw (2008), Using light‐use and production efficiency models to predict photosynthesis and net carbon exchange during forest canopy disturbance, Ecosystems, 11, 26–44, doi:10.1007/s10021-007-9105-0. Dale, V. H., et al. (2001), Climate change and forest disturbances, Bio- Science, 51, 723–734, doi:10.1641/0006-3568(2001)051[0723:CCAFD] 2.0.CO;2. DeLucia, E. H., J. E. Drake, R. B. Thomas, and M. Gonzalez‐Meler (2007), Forest carbon use efficiency: Is respiration a constant fraction of gross primary production?, Global Change Biol. , 13 , 1157–1167, doi:10.1111/j.1365-2486.2007.01365.x. Desai, A. R., et al. (2008), Cross site evaluation of eddy covariance GPP and RE decomposition techniques, Agric. For. Meteorol., 148, 821–838, doi:10.1016/j.agrformet.2007.11.012. Dore, S., T. E. Kolb, M. Montes‐Helu, B. W. Sullivan, W. D. Winslow, S. C. Hart, J. P. Kaye, G. W. Koch, and B. A. Hungate (2008), Long‐ term impact of a stand‐replacing fire on ecosystem CO2 exchange of a Ponderosa pine forest, Global Change Biol. , 14 , 1801–1820, doi:10.1111/j.1365-2486.2008.01613.x. Dore, S., T. E. Kolb, M. Montes‐Helu, S. E. Eckert, B. W. Sullivan, B. A. Hungate, J. P. Kaye, S. C. Hart, G. W. Koch, and A. Finkral (2010), Carbon and water fluxes from ponderosa pine forests disturbed by wildfire and thinning, Ecol. Appl., 20, 663–683, doi:10.1890/09- 0934.1. Dunn, A. L., C. C. Barford, S. C. Wofsy, M. L. Goulden, and B. C. Daube (2007), A long‐term record of carbon exchange in a boreal black spruce forest: Means, responses to interannual variability, and decadal trends, Global Change Biol., 13, 577–590, doi:10.1111/j.1365-2486.2006. 01221.x. Giasson, M.‐A., C. Coursolle, and H. A. Margolis (2006), Ecosystem‐level CO2 fluxes from a boreal cutover in eastern Canada before and after scarification, Agric. For. Meteorol., 140, 23–40, doi:10.1016/j.agrformet. 2006.08.001. Gough, C. M., C. S. Vogel, K. H. Harrold, K. George, and P. S. Curtis (2007), The legacy of harvest and fire on ecosystem storage in a north temperate forest, Global Change Biol., 13, 1935–1949, doi:10.1111/ j.1365-2486.2007.01406.x. Gough, C. M., C. S. Vogel, H. P. Schmid, and P. S. Curtis (2008), Controls on annual forest carbon storage: Lessons from the past and predictions for the future, BioScience, 58, 609–622, doi:10.1641/B580708. Goulden,M. L., G. C.Winston, A.M. S.McMillan,M. E. Litvak, E. L. Read, A. V. Rocha, and J. R. Elliot (2006), An eddy covariance mesonet to mea- sure the effect of forest age on land‐atmosphere exchange,Global Change Biol., 12, 2146–2162, doi:10.1111/j.1365-2486.2006.01251.x. Grant, R. F., T. A. Black, E. R. Humphreys, and K. Morgenstern (2007), Changes in net ecosystem productivity with forest age following clear- cutting of a coastal Douglas‐Fir forest: Testing a mathematical model with eddy covariance measurements along a forest chronosequence, Tree Physiol., 27, 115–131. Hudiburg, T., B. E. Law, D. P. Turner, J. Campbell, D. Donato, and M. Duane (2009), Carbon dynamics of Oregon and Northern California forests and potential land‐based carbon storage, Ecol. Appl., 19, 163–180, doi:10.1890/07-2006.1. Humphreys, E. R., T. A. Black, K. Morgenstern, T. Cai, G. B. Drewitt, and Z. Nesic (2006), Carbon dioxide fluxes in three coastal Douglas‐fir stands at different stages of development after harvesting, Agric. For. Meteorol., 140, 6–22, doi:10.1016/j.agrformet.2006.03.018. Irvine, J., B. E. Law, and K. A. Hibbard (2007), Postfire carbon pools and fluxes in semiarid Ponderosa Pine in central Oregon, Global Change Biol., 13, 1748–1760, doi:10.1111/j.1365-2486.2007.01368.x. Jassal, R. S., T. A. Black, T. Cai, K. Morgenstern, Z. Li, D. Gaumont‐ Guay, and Z. Nesic (2007), Components of ecosystem respiration and an estimate of net primary productivity of an intermediate‐aged Douglas‐fir stand, Agric. For. Meteorol., 144, 44–57, doi:10.1016/ j.agrformet.2007.01.011. Kashian, D. M., W. H. Romme, D. B. Tinker, M. G. Turner, and M. G. Ryan (2006), Carbon storage on landscapes with stand‐replacing fires, BioScience, 56, 598–606, doi:10.1641/0006-3568(2006)56[598: CSOLWS]2.0.CO;2. Kljun, N., T. A. Black, T. J. Griffis, A. G. Barr, D. Gaumont‐Guay, K. Morgenstern, J. H. McCaughey, and Z. Nesic (2006), Response of net ecosystem productivity of three boreal forest stands to drought, Ecosystems, 9, 1128–1144, doi:10.1007/s10021-005-0082-x. Krishnan, P., T. A. Black, R. S. Jassal, B. Z. Chen, and Z. Nesic (2009), Interannual variability of the carbon balance of three different‐aged Douglas‐fir stands in the Pacific Northwest, J. Geophys. Res., 114, G04011, doi:10.1029/2008JG000912. Kurz, W. A., and M. J. Apps (1999), A 70‐year retrospective analysis of carbon fluxes in the Canadian forest sector, Ecol. Appl., 9, 526–547, doi:10.1890/1051-0761(1999)009[0526:AYRAOC]2.0.CO;2. Kurz, W. A., G. Stinson, G. J. Rampley, C. C. Dymond, and E. T. Neilson (2008a), Risk of natural disturbances makes future contribution of Canada’s forest to the global carbon cycle highly uncertain, Proc. Natl. Acad. Sci. U. S. A., 105, 1551–1555, doi:10.1073/pnas.0708133105. Kurz, W. A., C. C. Dymond, G. Stinson, G. J. Rampley, E. T. Neilson, A. Carroll, T. Ebata, and L. Safranyik (2008b), Mountain pine beetle and forest carbon feedback to climate change, Nature, 452, 987–990, doi:10.1038/nature06777. Landsberg, J. L., and R. H. Waring (1997), A generalised model of forest productivity using simplified concepts of radiation‐use efficiency, carbon balance and partitioning, For. Ecol. Manage., 95, 209–228, doi:10.1016/ S0378-1127(97)00026-1. Law, B. E., et al. (2002), Environmental controls over carbon dioxide and water vapor exchange of terrestrial vegetation, Agric. For. Meteorol., 113, 97–120, doi:10.1016/S0168-1923(02)00104-1. Law, B. E., O. J. Sun, J. Campbell, S. Van Tuyl, and P. E. Thornton (2003), Changes in carbon storage and fluxes in a chronosequence of ponderosa pine, Global Change Biol., 9, 510–524, doi:10.1046/j.1365-2486. 2003.00624.x. Law, B. E., D. Turner, J. Campbell, O. J. Sun, S. Van Tuyl, W. D. Ritts, and W. B. Cohen (2004), Disturbance and climate effects on carbon stocks and fluxes across western Oregon USA, Global Change Biol., 10, 1429–1444, doi:10.1111/j.1365-2486.2004.00822.x. Lindroth, A., F. Lagergren, A. Grelle, L. Klemedtsson, O. Langvall, P. Weslien, and J. Tuulik (2009), Storms can cause Europe‐wide reduction in forest carbon sink, Global Change Biol., 15, 346–355, doi:10.1111/ j.1365-2486.2008.01719.x. Luyssaert, S., E.‐D. Schulze, A. Borner, A. Knohl, D. Hessenmoller, B. E. Law, P. Ciais, and J. Grace (2008), Old‐growth forests as global carbon sinks, Nature, 455, 213–215, doi:10.1038/nature07276. Margolis, H. A., L. B. Flanagan, and B. D. Amiro (2006), The Fluxnet‐ Canada research network: Influence of climate and disturbance on carbon cycling in forests and peatlands, Agric. For. Meteorol., 140, 1–5, doi:10.1016/j.agrformet.2006.08.013. McMillan, A. M. S., G. C. Winston, and M. L. Goulden (2008), Age‐ dependent response of boreal forest to temperature and rainfall variability, Global Change Biol., 14, 1904–1916, doi:10.1111/j.1365-2486.2008. 01614.x. Misson, L., J. Tang, M. Xu, M. McKay, and A. H. Goldstein (2005), Influences of recovery from clear‐cut, climate variability, and thinning on the carbon balance of a young ponderosa pine plantation, Agric. For. Meteorol., 130, 207–222, doi:10.1016/j.agrformet.2005.04.001. Mkhabela, M. S., et al. (2009), Comparison of carbon dynamics and water use efficiency following fire and harvesting in Canadian boreal forests, Agric. For. Meteorol., 149, 783–794, doi:10.1016/j.agrformet.2008. 10.025. AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 12 of 13 Moffat, A. M., et al. (2007), Comprehensive comparison of gap‐filling techniques for eddy covariance net carbon fluxes, Agric. For. Meteorol., 147, 209–232, doi:10.1016/j.agrformet.2007.08.011. Noormets, A., J. Chen, and T. R. Crow (2007), Age‐dependent changes in ecosystem carbon fluxes in managed forests in northern Wisconsin, USA, Ecosystems, 10, 187–203, doi:10.1007/s10021-007-9018-y. Noormets, A., S. G. McNulty, J. L. DeForest, G. Sun, Q. Li, and J. Chen (2008), Drought during canopy development has lasting effect on annual carbon balance in a deciduous temperate forest, New Phytol., 179, 818–828, doi:10.1111/j.1469-8137.2008.02501.x. Noormets, A., J. Chen, L. Gu, and A. Desai (2009), The phenology of gross ecosystem productivity and ecosystem respiration in temperate hardwood and conifer chronosequences, in Phenology of Ecosystem Processes, edited by A. Noormets, pp. 59–85, doi:10.1007/978-1-4419-0026-5_3, Springer, New York. Odum, E. P. (1969), Strategy of ecosystem development, Science, 164, 262–270, doi:10.1126/science.164.3877.262. Piao, S., S. Luyssaert, P. Ciais, I. A. Janssens, A. Chen, C. Cao, J. Fang, P. Friedlingstein, Y. Luo, and S. Wang (2010), Forest annual carbon cost: A global‐scale analysis of autotrophic respiration, Ecology, 91, 652–661, doi:10.1890/08-2176.1. Randerson, J. T., et al. (2006), The impact of boreal forest fire on climate warming, Science, 314, 1130–1132, doi:10.1126/science.1132075. Richardson, A. D., and D. Y. Hollinger (2005), Statistical modeling of ecosystem respiration using eddy covariance data: Maximum likelihood parameter estimation, andMonte Carlo simulation of model and parameter uncertainty, applied to three simple models, Agric. For. Meteorol., 131, 191–208, doi:10.1016/j.agrformet.2005.05.008. Schwarz, P. A., B. E. Law,M.Williams, J. Irvine, M. Kurpius, and D.Moore (2004), Climatic versus biotic controls on carbon and water fluxes in sea- sonally drought‐affected ponderosa pine ecosystems,Global Biogeochem. Cycles, 18, GB4007, doi:10.1029/2004GB002234. Sellers, P. J., et al. (1997), BOREAS in 1997: Experiment overview, scien- tific results, and future directions, J. Geophys. Res., 102, 28,731–28,769, doi:10.1029/97JD03300. Stocks, B. J., et al. (2003), Large forest fires in Canada, 1959–1997, J. Geophys. Res., 108(D1), 8149, doi:10.1029/2001JD000484. Sun, O. J., J. Campbell, B. E. Law, and V. Wolf (2004), Dynamics of carbon storage in soils and detritus across chronosequences of different forest types in the Pacific Northwest, USA, Global Change Biol., 10, 1470–1481, doi:10.1111/j.1365-2486.2004.00829.x. Thomas, C. K., B. E. Law, J. Irvine, J. G. Martin, J. C. Pettijohn, and K. J. Davis (2009), Seasonal hydrology explains interannual and seasonal variation in carbon and water exchange in a semiarid mature ponderosa pine forest in central Oregon, J. Geophys. Res. , 114 , G04006, doi:10.1029/2009JG001010. Urbanski, S., C. Barford, S. Wofsy, C. Kucharik, E. Pyle, J. Budney, K. McKain, D. Fitzjarrald, M. Czikowsky, and J. W. Munger (2007), Factors controlling CO2 exchange on timescales from hourly to decadal at Harvard Forest, J. Geophys. Res., 112, G02020, doi:10.1029/ 2006JG000293. Vesala, T., et al. (2005), Effect of thinning on surface fluxes in a boreal forest, Global Biogeochem. Cycles, 19, GB2001, doi:10.1029/ 2004GB002316. Welp, L. R., J. T. Randerson, and H. P. Liu (2006), Seasonal exchange of CO2 and d 18O‐CO2 varies with postfire succession in boreal forest eco- systems, J. Geophys. Res., 111, G03007, doi:10.1029/2005JG000126. Welp, L. R., J. T. Randerson, and H. P. Liu (2007), The sensitivity of carbon fluxes to spring warming and summer drought depends on plant functional type in boreal forest ecosystems, Agric. For. Meteorol., 147, 172–185, doi:10.1016/j.agrformet.2007.07.010. Wofsy, S. C., and R. C. Harriss (2002), The North American Carbon Program, report of the NACP Committee of the U.S. Interagency Carbon Cycle Science Program, pp. 56, U.S. Global Change Res. Program, Washington, D. C. Zha, T., et al. (2009), Carbon sequestration in boreal jack pine stands following harvesting, Global Change Biol., 15, 1475–1487, doi:10.1111/ j.1365-2486.2008.01817.x. B. D. Amiro, Department of Soil Science, University of Manitoba, Winnipeg, MB R3T 2N2, Canada. (brian_amiro@umanitoba.ca) A. G. Barr, Environment Canada, Saskatoon, SK S7N 3H5, Canada. J. G. Barr and V. Engel, Everglades National Park, Homestead, FL 33030, USA. T. A. Black and M. Brown, Faculty of Land and Food Systems, University of British Columbia, Vancouver, BC V6T 1Z4, Canada. R. Bracho and T. Martin, School of Forest Resources and Conservation, University of Florida, Gainesville, FL 32611, USA. J. Chen, Department of Earth, Ecological, and Environmental Sciences, University of Toledo, Toledo, OH 43606, USA. K. L. Clark, USDA Forest Service, New Lisbon, NJ 08064, USA. K. J. Davis and J. D. Fuentes, Department of Meteorology, Pennsylvania State University, University Park, PA 16802, USA. A. R. Desai, Department of Atmospheric and Oceanic Sciences, University of Wisconsin‐Madison, Madison, WI 53706, USA. S. Dore, T. E. Kolb, and M. Montes‐Helu, College of Engineering, Forestry, and Natural Sciences, Northern Arizona University, Flagstaff, AZ 86011, USA. A. H. Goldstein, Department of Environmental Science, Policy, and Management, University of California, Berkeley, CA 94720, USA. M. L. Goulden and J. T. Randerson, Department of Earth System Science, University of California, Irvine, CA 92697, USA. M. B. Lavigne, Canadian Forest Service, Fredericton, NB E3B 5P7, Canada. B. E. Law, College of Forestry, Oregon State University, Corvallis, OR 97331, USA. H. A. Margolis, Faculty of Forestry and Geomatics, Université Laval, Quebec, QC G1V 0A6, Canada. J. H. McCaughey, Department of Geography, Queen’s University at Kingston, Kingston, ON K7L 3N6, Canada. A. Noormets, Department of Forestry and Environmental Resources, North Carolina State University at Raleigh, Raleigh, NC 27695, USA. G. Starr, Department of Biological Sciences, University of Alabama, Tuscaloosa, AL 35487, USA. J. Xiao, Complex Systems Research Center, University of New Hampshire, Durham, NH 03824, USA. AMIRO ET AL.: CARBON FLUXES AFTER FOREST DISTURBANCE G00K02G00K02 13 of 13


Citation Scheme:


Usage Statistics

Country Views Downloads
United States 5 0
Canada 5 0
China 3 0
City Views Downloads
Vancouver 5 0
Ashburn 3 0
Beijing 3 0
Mountain View 2 0

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


Share to:


Related Items