|
<br />
<br />11l68a: Data on M()untain EnVUOIllfleDts U: Franl Range, Col-
<br />orado, Fuur Climax Regions, 1953-11158. Univ~rsiry of Co/-
<br />orado SI"diu Sulu In Biology, 28. 16\l pp.
<br />Man. J. w.. Clark, J. M.. Osburn. W. S., and Paddock. M. W.,
<br />IIl68b: Data on M()untain EnvironmcnUl W: Front Row$-e, Col-
<br />oradG, FQur Climax Regioru. I 11511-1964. Uni~rsiry of Col-
<br />oraJoS/4d,esSeriulnBiology.21l.18lpp.
<br />Pcpin. N. C., Benham, D.. and Taylor, K., 1999: Modeling lapliC
<br />rates In the maritime uplamls of northern Enahmd: Implica.
<br />lions for climate change. Arc/ic. Antarclic, ll1Id Alpine Re.
<br />uurch. 31. 130-143.
<br />Wilhoms. M. W.. L()slebcn, M.. Caine, N., and Grcenland, D..
<br />1996: Changcsin climille and hydro<:hemicalresponliCs ina
<br />hich.elevation catchmenl in the Rcx:!(y mountains, U.S.A..
<br />Limnolo8Y and Oceanography. 41: 939--946.
<br />Williams, M. W., Bardsley. T, and Rikkers. M., 19911; Ovcres-
<br />Ilrnallon of snow depth and inorganic nilrOgen welfall using
<br />NADP data. NlWOl Ridge. Colorado. AtmOSpheric En...iron.
<br />ment, 32:3827-3833.
<br />
<br />Table 11
<br />
<br />Arclic, N1torcfic, and Alp'ne Reurm:h. Vol. J2. Nu. 2. 2000. pp. 147-/54
<br />
<br />Subsf(Jnlial gaps {>J moJ in Ille records
<br />
<br />Simulation of Carbon and Nitrogen Cycling in an Alpine Tundra
<br />
<br />""'.
<br />_ N_
<br />AI '':nt_n No. t~7~. ~~ Au. 1~77-1 Jun 1978. II scp 1~7ll-n
<br />Apr 1979. K N". 1979_!3JuI1911O
<br />lit Hln-II Doc 197~. ~ Dc, 1971-1 Jun InK. ;!j j~n 19110-Jul
<br />3119llO.
<br />CI 290<e 11>SS--30Nn I~S6
<br />DI II~ntl>~lliJunII>S6.gi'l<'vll>77_IIJunI97g
<br />
<br />Alexis H. Conley, '"
<br />Elisabeth A. Holland, t
<br />T. R. Seastedt,t. and
<br />W. J, ParlOn~
<br />
<br />,
<br />J
<br />
<br />Abslrlld
<br />
<br />SimulD.tions of an D.lpine tundra ecosystem using the CENTURY ecosystelll model
<br />were condllcted to test model descriptiolls of <:arbon D.rI.d nilrogen cyclinil and to
<br />cxplore the alpine ecosystem response to physical and chemical components of
<br />global chD.llge. The pnramc:terizD.tion of the alpine lundra for CENTURy was
<br />updated to reflect currenl knowledge of the site. and sensitivity analyses were
<br />conducled. Verilic:uion of resull~ from a 6-yr fertilization experiment in the alpine
<br />tested Ihe predictive capabiliti.cs of the paramelerization. Simulations With tn-
<br />creased winler precipitation and wilh the chmale predicted under doubled atmo-
<br />spheric carbon dioxide concenl11llions were tll.en conducted.
<br />Modifications 10 the pacD.meterization necessary to uescribe carbon D.Ild ni-
<br />trogen cycling included decreasing the C;N ratios ofplD.llt tissues. increasing the
<br />amount of mtrogen relranslocated at Ihe end of the growing scason, elltending
<br />the length of the growing season, and lowering the rate of decomposition. The
<br />updated parameterization requires 30% greater thllP observed inputs of nel pri-
<br />mary productivity to simulate observed level$ of 10lal soil carbon suggesling Ihal
<br />soil carbon sequestralion is. not well represented in the model. Carbon and nitro"
<br />gen cycling showed greatest sensitivity to the length of the growing season D.Ild
<br />10 the temperD.ture regulation of de<:omposition. Simula.lion of the nitrogen fertil-
<br />ization experiment resulted in II % greater productivity thllP observed empirically.
<br />a reasonable verification of the updaled parameterization. The major impact from
<br />increasing winler precipilation was a 30% increase in the D.mount of nitrogen in
<br />SlrCiIlU flow. Simulation with the climate predicted for a doubling of cunent car-
<br />bon dio:tide levels reduced production 10% while tOlal sot! carbon remained con-
<br />Stant. nus response was largely controlled by redUCed soil mOISture dunng the
<br />growing season.
<br />
<br />
<br />Ms submined MIJY 1999
<br />
<br />Appendix 1: Delllils of Dala Hallldling
<br />
<br />DiffelCnces in mean daily maxima and minima bclwccn the
<br />two instnJllIl:nts (hygrothermovaph and datapod) " sites A I and
<br />B I in turTl were calClllated usin& the overlapping d&IL "Ibcrc
<br />was 00 strong seasonal pllucm to the diffcrmea (Table 10) and
<br />the mean oorrcctiGllS listed were spplied 10 the dalapOd dal.l.
<br />Corrections arc broadly similar at Al and BI. Both maximum
<br />ilIld miuimullllempcratW"es arc underestimated by Ihc datapod
<br />in cGmpanson with the hygrothennopllph. altbouah differences
<br />in[llinimaarcs[llaU(<O.I"C).~lIIlIallyrcaetrela_
<br />lively ~lowly to rapid temperalUre chanacs and reJlslCr a teduccd
<br />diurnal temperature range (Marr. 1%7: Docskc:o. pen. comm..
<br />1998). Diffcn:nccbctwccn readings will abobcrelutdlOlhc klagcr
<br />sampling 41lerval of the daIapod. Daily rrw.imum aDd minimUIII
<br />~mpcratures frn- tilt datapod wen: cak:u!aIcd from mean tem-
<br />peratures taken every 10 min (see: SopbcrJJcullcr.eolondo.
<br />edu:70111l,NiwotRida:eDaIaJ.McteorologyandClimalolosY) mean.
<br />inglower maxima. AI big.h-e1evatiollS ttIIWma often (lCcur with
<br />a well-defined peak. due to stronS tnOllIing in4ola.tion whio::h. al
<br />
<br />TABLE 10
<br />
<br />DifferMC~t berwun dlJtlJpoa ond chart rerlliings/or each nwnth
<br />al 11.1 and BJ. The differences IJre blUed 011 W period of ollerlop
<br />(25 July J987-26 F"b /990 ind). All differences are gillen In
<br />degren Celsius. A patlllYe IIlJllle meO/l,f Ihm lhe dQlapod i.t cold.
<br />"J1wnllucJmn
<br />
<br /> ^' "
<br />M,m'h M.. Moo M~ Min
<br />, +o,n +OA~ +0.~2 -0.13
<br />> +0,21 +I.a -O.jot +0.29
<br />M +1l1~ -0,0' H),9l +0.33
<br />^ +fJJ +II.'N +0.17 +O.O!
<br />M +0,78 +0.66 +0,1& +G.O'
<br />, +0,~8 -OO~ +0.43 -0.09
<br />, ."~ -a,O +0..' '."
<br />^ +0.9-1 -a,ll +I.n -o.na
<br />, +0.36 -0.38 +1.11 +0.2S
<br />0 .,,, +0.69 +0.6' -,..
<br />N +l.J1 +1.111 +0.66 +0.l4
<br />D +06' +0.\17 ..~ +0.09
<br />,= +0.16 +0,2~ .... +0.06
<br />
<br />1461 ARcrJC. ANTARCTIC, AND ALPINB REsEARCH
<br />
<br />leut in summer. is oflen cut uff by conveclive cloud build_up
<br />around midday.
<br />1bere arc a few gaps in !he records. most noticeably al Al
<br />and Bl. and occasIonally at the higher siles (Table II). During
<br />the lale 1970sespecially, records were losldue 10 the failurc:of
<br />thc clock mcchanism on thc r<>>atinC drum. or failure of the pen
<br />to make contacl with the chan. In other cases severe we~ther
<br />mo:antU\a1 chans were leh On longer than their intended lifelimc
<br />c~usin, overlllppina: traces. However. mIlCh data could be eX'
<br />tracledthrough comparison with simultaneous cham al other
<br />locations IUId makina: the usumption that periods of anoma.
<br />klUSly cold or warm weather were correlated between ~lle~. In
<br />litis way overlyr ofdala were added to the existin, Al and
<br />B I rec~. lhc mu:imum difference in readings that could re.
<br />suit from telldina: different poinls in the width of the chari line
<br />is aboul O.2S"C (Marr. 1%7).
<br />Anoltter subtle pmblem coneerns the clock mechanism.
<br />which often ran slow. Midnight lines on the chan would nGt
<br />com::sporut with mi<1nighl in reality. Errors of up to 6 h wete
<br />found. Because of the markedly diurnal lemperature re,ime in
<br />Colorado. al leut al lower elevations, Ibis is relatively easy to
<br />iOenlify. This problem Will affccl minimllm temperatures record.
<br />cd (from midnight to midnight), especially if ll1Inima on suc.
<br />~ssin nights arc "onsiderably differen!. By assuming daily
<br />R1inllna 10 occur around 6 am MDT in Slimmer ar>d slightly !aler
<br />in winler, the identification of true midniJltI times was possible.
<br />Masl cbans covered llIleasl a month and the cloo;:l<s arc uSllally
<br />consistendy slow (they do 1101 speed up and slow down im:g\l-
<br />larly). In many easc~, minima were re-rta<! using new midnight
<br />lines.
<br />Another concern is the vertical correclion 10 be applied to
<br />individual charts. Often a large-scale correclion is neccssary
<br />(usually a ffi\llliple of 10 deg F or C), as the venicaltemperatute
<br />seale is displaced. When a I!CW chan is assembled the "hart
<br />tempcralUrc is compared wilh the lelUal air temperature mea-
<br />iured by It standard Weather Bureau-type mercury in gl~ss ther-
<br />IllOffiCter. This leads to a eoxrection whiCh is wrillen on the chart
<br />When disassembling. a similar procedure is followed. Usually
<br />beginning and end corrections Were similar. but occasionally
<br />they were nolo The: hygrothc:nnograph is II remlUkably robusl
<br />inslrUmcnl (L05lebcn. pers. comm.. 1998) and any difftrence in
<br />the IwoC<UteClionswollldbcduc to differences in response time
<br />between the hygrothcnnograph and thcnnomeler. The hygro-
<br />1hcrmojrllph's ,reltter Ill' OOt:llsionaJly results in a rcadil'. as
<br />much as 20F (_I.IOC) differenl from the corresponding ther.
<br />mometer rcading(Marr. 11167), especially if the temperature is
<br />increasint; rapidly, ()flen lhe "uc al II am MST (the usual time
<br />of chan change). Difficult cases were complUed With subsequcnt
<br />"",d previOlls charu and thb oflen confinned that contr1lsting
<br />ins1t'WllCnlllllag was the problem. Unusual cOlTection v:alues
<br />were ignored iIrId in examples where lag was thought not 10 be
<br />wholdy responsible for contrasting bcginningandendcolTC\:tion
<br />values the cban was corrected by the mean of the beginnina and
<br />cnde=tions.
<br />
<br />"OcpanmcnlofEnvironmental.
<br />POJI'IlllIion. and OrgllJll.lOlC Biology.
<br />Unive,,"ityofColorado. Boulder.
<br />Colorado 80309-0334. USA
<br />PresentaddteSll; SSAI.NASA
<br />God<btdSpace F1ia:htCelller. Code
<br />923. Gteenbcll. Marylllnd:l.0771.
<br />U.S,A. Alcxjs.Conley@gsf~.nasa.gGv
<br />tAtmosphcric Chemi.try Division.
<br />NalionalCenter for AlIoo'phcric
<br />Re....""'h. Boulder, COIOlOldo 80303.
<br />U.S,A.
<br />tln.tjtuteofArcllcarniAlplne
<br />Ruewxh and Dcpanmcnl or
<br />Environmcntal,Populalion and
<br />Organismic Biology. Univtrsityof
<br />Coloritdo, Boulder. CoJoClldo 80309-
<br />0450, U.S.A.
<br />ONatllral Resour<:esEcol08Y
<br />Laboratory, Colorado Sllltt Universily.
<br />FGn Collins. Colorado 80523_1499.
<br />U.S.A.
<br />
<br />Introdudion
<br />
<br />Alleration of thc physical and chemical climMC Cllll impacl
<br />terrestrialecosysteml through altcration of nutrient slnrage and
<br />cycles. With strch chanacs, the ability of these ccosyslCrrts to
<br />sequcestercarbon may be altered re~lIlling in a positive feedback
<br />to global warmina: (Asner et al.. 1997). The impact of changes
<br />in the chemical climate On stream water quality is also 11 concern
<br />for sites thaI serve IS municipal waler sources. For Niwol Ridge,
<br />an 1I1pine tundra site inlhe Colorl<1o Rocky Mountain;, there is
<br />arcat inlCrest in dclermining the effect of documented climale
<br />chan,es including increased levc(s of winter precipitalion and
<br />nitrogen dcposilion as well as potential climate changes with
<br />inc~ased levels of almospheric carbon dioxide (Creenland.
<br />1989; Sievering el aI., 1992, 1996; Williams et aL. 19116). These
<br />types of allerationl may influence nUlrient cycling andslOrallC
<br />anduhilllll.tely impact no.Ionly this alpine ecosy.tem bUI also
<br />waler supply for the expanding down5lre;un communities of ~
<br />Colorado Front Range.
<br />To re.spond 10 qucslions thaI addrcss changcs in thc physical
<br />and chemical climale, an ecosystem model is useful for circum-
<br />vellW>$ !he: logistical, temporal. and spatial consuainlS thai an
<br />equivalenl empirical study imposes. Parlicularly effeclive ~
<br />ecosystem models in Which ccosystem funclion i~ regulated by
<br />physical (prc<:ipitation and lempcrature) and cllcmica! (N depo-
<br />sition) climates. Such models. including the CENTURY ecosys_
<br />tcm model developed by panon cI al. (1987) llIld the GEM mod_
<br />
<br />o 2000 Regenls of the Universily ofCoIora(\o
<br />IS2).043<WOS7.00
<br />
<br />el of Rastcllerel al. (1991). have beenempl()yed to explure the
<br />impli"ationsofcli~chanaclbrou,bthesitnulationofregl<,lll'
<br />al and global biogeochemical cyclin, (VEMAP Membe... 1995:
<br />McKlI1lc:: et al., 1991; Pul.,el al" 1998).
<br />The CENTURY mockl was used lO an:alyze long-term .im.
<br />ulatioll$ of increased nilrOllen dcwlilion on NiwOl Ridge (Baron
<br />etal" 1994). Since then. a number' of slIIdies thai deline prevt-
<br />ously unknown vegetatiOl:\ characttrislies IS well as .oiln1uogen
<br />dynanucs have been publisbcd. These studies document baseline
<br />value~ for nitrogen mineralization, microbial nitrogen immobi-
<br />lilatiunvalucl,planltissuecarbon:nitrogen(C:N)ratios.i(Jldni-
<br />trogen gas nuxes (Bowman. 19114; Neffct al.. 1994; Bowman el
<br />a1.. 1995. 1996; Fisk and Schmi<lt. 1995. 1996: Brooks ct aL.
<br />1996;Fiskelal.,19\l8l. A 6.yr ftrtilizahon Study lhalcan pro-
<br />vide inililllvelletationre~ponse to illcrca.seo Dltfogen supply was
<br />also condUCle<J on Niwot Ridge (Bowman et al.. 1993. 1995/.
<br />The"," studies provide a more extensive database to evaluate,
<br />improve. and verify the parameteril'3tion <>f Niwot Ridge for
<br />CENTURY.
<br />We dcvelopcd a thrte-pan study to evaluate thc contcols of
<br />carbon and nitrogen cycling for the alpine tundra on Niwot
<br />Ridge. First. we de!ermined whethcrour general understanding
<br />of carbon and nitrogencyclins. asreprestnted by thc CEl'nlJRY
<br />ecosystem model. is adcqllate for Niwol Ridge. Second. we test-
<br />ed tIu: model's abilily to simulate Ihe production rcaponse ob-
<br />se~vtd in the sh0l1-1erm nitrogen fel1iliz.ationexperirnent. Final.
<br />
<br />A. H. CONLEY ET AL. J 147
<br />
|