Anna A.Osmirko 4 3 2 1 Vladimir А. Usoltsev * Corresponding author:[email protected] Ulaanbaatar, Krasnoyarsk, Mongolian Aboveground Region Ledeb additivity, models, allometric dummy variables, tablesofbiomass. with thehelpofheightand diameter data. models allow oflarch the determination biomassindifferent ecoregions ofEurasia stands, is bydetermined which, both in climatic turn, and edaphic factors. The obtained role. in the morphological structureof is the variation factor Also,important important of theirstems, remains open.Undoubtedly, thedifferences intree agehere play an regional differences in the structure of biomass of with thesame linear dimensions weighing,of componentbiomass, theirdrying, etc.). The questionwhatexplains the of age, diameter, heightofatree, ofsupposedlyrepresentative theselection samples and unaccounted anderrors ofmeasurements factors inalltheirphases(assessment by otherscientists. Biomassrelations regions between may beexplainedby unknown the growing conditionsoflarch are in Mongolia notastough asitwassuggested earlier the background of the most ecoregions of Eurasia. Based on our results, we conclude that stands. The aboveground biomassoflarch trees doesnotstandout against inMongolia ofstem woodwith ahighbasicdensity and relatively highdeveloped crown inopen that may beexplainedby permafrost conditions, by tree growth inlow-yielding stands component structure. A significant excess of biomass in the forest-tundra is found into accountregional (andcorrespondingly species-specific)differences oftrees inits ofthecomponentandtotal biomassequations,contradictions secondly, ittook tree aboveground ways: biomasswasharmonized intwo first,iteliminated the internal combined problem andregionality ofthemodel. ofadditivity Ouradditive modelof developing anadditive modelofbiomasstree components. Ourapproach solved the of Key Abstract Institute of General and Experimental Biology, andExperimental ofGeneral Institute Academy Mongolian ofSciences, V. ofForest, Institute N.Sukachev RussianAcademy Branch, ofSciences, Siberian Garden, Botanical RussianAcademy ofSciences, UralBranch, Yekaterinburg, Russia UralState Forest Engineering University, Yekaterinburg, Russia Larix from different ecoregions ofEurasia, involving 61trees from for Mongolia words . We usedourdatabaseoftree biomasswithanumberof433sampletrees : genusLarix spp., aboveground tree biomass, regional differences, equations .) 1 , Ivan S. Tsepordey 1, 2 , IgorМ.Danilin Larch Biomass In 2 The , Viktor P. Chasovskikh rix Sibiric (La 3* , Zaandrabalyn Tsogt, Eurasian Of

1

4 , a

117 GES 03|2019 118 GES 03|2019 models have either regional application of tree biomass. As aresult, thepublished data developed onlocal (regional) harvest These and othermodelsare usually estimators. andresultingprinciple inmore effective fitting isused, accounting theadditivity incasewhenthesimultaneous observed butthisisnot biomass isnotsatisfactory, for biomasscomponentsandtotal coefficients of fitting additivity) (without to Sanquettaetal. (2015),independent the equationfor total biomass. According equal to the value of biomass obtained by obtained by thecomponentequationsis components (stems, branches, needles) and meansthatthetotal biomassof of themodelby biomasscomponents composition assumestheconsistency al. 2016). ofthecomponent The additivity Parresol 2001;Zhengetal. 2015;Zhanget 1970; component composition(Kozak 2012; Z andCunia 1980;Fua model(Jacobs etal. either by involving dummy into variables models oftree biomass, thatisfulfilled related ofallometric to theharmonization cover oftheplanet,there isatrend increasing biosphere of forest function conditions of continuously 1983). In height) isthecurrent standard (Marklund of stem diameters (and sometimes of tree trees, wasrepresented inthewholerange using the results of sampling of model method ofestimatingforest biomasswhen at a tree level. The regression (allometric) models ofbiomassare needed, developed inmany casesallometric productivity estimation of theirbiologicalcorrect represented by mixed forests, for the of forestsignificant part cover has been Since a carbon. as sinks of atmospheric Forest role ecosystems play an important GEOGRAPHY, DOI-10.24057/2071-9388-2018-70 Vol.12, No3,p. 117-132 Sibirica Ledeb.) Forests In The Eurasian Region. Geography, Environment, Sustainability, S. Tsepordey, Viktor P. (2019)Aboveground Larch Mongolian (Larix BiomassOf Chasovskikh Citation INTRODUCTION INTRODUCTION eng 2015),orby providing additive : Vladimir А.Usoltsev, IgorМ.Danilin,Z ENVIRONMENT, SUSTAINABILITY of diameter classes varies fromof diameter classes varies 0.5 to 32 breast height (DBH) more 0.5 cm. Range height more 1.5 m and stem diameter at an area involving 200-400 trees with the measured all trees. Each plotconsiders We established 5 sample plots and 700-1500 ma.s.l.) (Fig. 1). (49°10'N,110°0'E, village ofMungun-Mort Mountains, nearthe ofKhentii part Eastern the Mongolia, Ledeb.) ofnorth-eastern forests larch oftheSiberian ( biomass of 43trees were obtainedinthe dataonthe 2016).Harvest level (Usoltsev developments onthetrans-continental methodologicalthe calledmodern of Eurasia provides thedatato develop created for theforest-forming species biomass The databaseofsingle-tree (Draper andSmith1966)into amodel. of Eurasia by involving dummy variables with thetree biomassofotherregions larch interfaced the Siberian inMongolia, aboveground biomassontheexampleof a regional additive model(AM) oftree mentioned approaches, andto develop The goalofthispaperto becombinethe composition ofthebiomass. ensure ofthecomponent theadditivity arethey ofregional nature anddonot Zhao et al. 2011; Battulga et al. 2013), but etal.Bjarnadottir 2007; Novák etal. 2011; already beenpublished(Shietal. 2002; from stem diameter andheightdatahave estimating theaboveground biomass genus any regionalization 2017).For (Usoltsev component compositionbutwithout composition, orprovide of additivity of component additivity the principle dummy andwithoutproviding variables withthehelpof without harmonization M aandrabalyn Tsogt, Ivan AnnaA.Osmirko, aterials Larix spp. a number of models for AND METHODS 03 (12)2019 Vladimir А.Usoltsev, IgorМ.Danilinetal. 10. The calculation of the stem volume forof sections eachtree wasnoless than the heightof the tree stem. The number dependingon wasdetermined of sections heights (according to sections). The length samples cutoutofthestem atdifferent hectare. wood ratiowasfound Bark from andreducedof stem thickness, to 1 analytically, summedaccording to classes state wasequalized an absolutely dry ±0.1 g. The weight ofcomponentsin to at electronicscalewithanaccuracy three(+105°C during days) andweighed state up to an absolutely dry in a dryer dried moisturedetermine content were they to thelaboratory, wheretransported to into polyethylene and packets,marked, from eachcomponentwere packed at aweight greater than10 kg. Samples to ±100g and withanaccuracy to 10kg to ±1gataweight up with anaccuracy weighed inwet state balance atlever arid divided into biomasscomponentsand DBH class. trees Model were completely trees – one mean value exemplar of each model cm. Oneachplotwe harvested Fig. 1. The positionofthesamplearea onthemapofKhentii Mountains (marked by 1:6000000 theblackpoints). Scale years in the upper river(62 years intheupperKolyma thousand trees per1ha at theage of 30 river (64 in the basin of the Nizhnyaya Tunguska thousand perhaattheage of 86-87years sample plots, modelsof andallometric larchof Siberian trees wasstudiedon 18 Altai, the biomass structure Mongolian western border in the mountains of the another region onits In ofMongolia, (64 at theageof40years near Arkhangelsk E) (Pozdnyakov 1975),4.8thousandperha (62 per 1haattheageof14years in Yakutia particular, larch has112thousandtrees ofRussia.In but alsointhenorthern the harshmountainousareas ofMongolia, forthat istypical larch, growing notonlyin the relatively of larch highdensity stands, shown in Table 1. We draw attention to the tree biomassonthesampleplotsare Tsogt 2014,2015). The results ofestimating Huber’s formula (Danilin2009;Danilinand according to thecomplex was performed 0 0 N, 130 N, 40 0 N, 100 0 E) (Molchanov 1971),2.7-2.9 E)(Molchanov 0 Aboveground E) (Pozdnyakov 1975),55 0 E) (Abaimov etal. 1997). Biomass 0 N, 147 Of

... 0

119 GES 03|2019 120 GES 03|2019 GEOGRAPHY, No. 18 17 16 15 14 13 12 11 10 26 25 24 23 22 21 20 19 8 7 6 5 4 3 2 1 9 Table 1.Aboveground biomassoflarch oftheKhentii trees inthe eastern part years Tree Tree age, 14 15 17 18 17 18 18 18 15 33 15 21 25 27 27 36 35 32 27 24 29 35 31 30 32 34 Stem Stem DBH, DBH, 12.5 13.2 15.4 17.6 10.1 11.2 сm 1.0 1.9 2.9 4.0 4.9 6.3 7.5 8.9 0.5 0.8 1.8 3.9 7.3 9.8 3.2 4.1 5.6 7.1 8.0 8.8 ENVIRONMENT, height, height, Tree Tree 10.9 10.7 11.1 11.3 10.0 10.4 2.4 3.3 3.6 4.3 4.9 5.2 5.9 6.5 1.5 2.3 3.0 5.8 7.3 8.3 6.2 7.1 7.9 8.8 9.2 9.5 m 100.6 125.4 volume, dm Total 11.2 16.9 22.6 73.8 18.3 39.9 75.8 10.5 20.8 27.1 33.3 43.9 54.4 0.2 1.1 2.2 4.3 6.4 0.1 0.2 0.8 4.3 2.9 5.3 Stem Stem SUSTAINABILITY Stem Stem bark 0.05 28.1 37.9 0.05 20.8 18.4 10.0 13.4 16.8 3.5 5.3 7.1 0.5 0.9 1.4 2.0 0.1 0.3 1.2 5.4 9.5 0.9 0.9 2.2 6.6 8.3 3 Total Biomass in absolutely dry condition,kg Biomass inabsolutely dry 0.12 40.2 51.4 0.05 0.15 18.4 29.7 29.0 11.5 14.2 19.0 23.5 5.1 7.4 9.6 0.4 1.0 2.0 3.1 0.4 1.9 8.5 1.4 2.7 4.8 8.7 Mountains Stem only Bark Bark 0.05 11.2 0.02 0.05 1.3 1.8 2.3 0.1 0.3 0.6 0.8 8.8 4.8 6.7 0.1 0.6 2.3 6.4 2.9 0.4 0.7 1.2 2.4 3.3 3.6 3.7 0.032 0.021 Bran- ches 0.21 0.42 0.84 13.8 19.8 0.04 0.21 1.66 0.07 0.08 0.23 0.95 2.7 4.6 6.2 1.2 8.1 3.7 0.5 3.3 8.0 2.1 2.8 3.4 Foliage 0.03 0.01 0.01 0.04 0.44 0.03 0.03 0.06 0.40 1.1 1.9 2.6 0.1 0.3 0.5 0.7 5.2 7.5 2.7 1.4 0.2 1.3 2.9 0.5 0.7 0.9 0.182 0.081 10.05 Sum Sum total 13.9 18.4 0.62 1.72 3.34 59.2 78.7 0.20 29.2 34.8 0.65 13.1 39.9 13.6 2.81 5.09 16.8 22.5 27.8 8.9 5.0 2.6 1.5 numbers numbers per hа 56200 19800 5700 Tree Tree 03 (12)2019 diameter, Mean Mean cm 1.6 5.9 7.2 P a = e xp a 0 + Vladimir А.Usoltsev, IgorМ.Danilinetal. different regions have differences notonly equations of thatallometric isknown It addition to the stem diameter and height. model asan independentvariable, in result, thetree agewas involved into the the influenceofageon comparison from 55to 173years. order to In exclude to 75 years Altai and on the Mongolian from Mountains 14 ranges: ontheKhentii regions were obtainedindifferent age tree biomassdataofthetwo The harvest generalized equation. separately for regions, two orto give one equationseither can calculate allometric and dependingontheresult obtainedwe mountainregionsthe two ofMongolia, tree biomassdataof in theabove-ground to findoutwhetherthere are differences the stem (Battulgaetal. 2013). We decided according to the diameter and height of biomasswereabove-ground calculated a 43 42 41 40 39 38 37 36 35 34 33 32 31 30 29 27 28 1 I nA + 60 67 70 70 72 73 74 73 75 33 40 41 43 42 39 44 43 a 2 I nD 12.2 15.4 18.7 22.0 24.5 27.2 29.7 32.3 12.0 15.9 19.8 23.5 31.0 27.4 8.8 4.5 8.1 + a P 3 a I nH 10.5 13.7 16.8 17.5 18.2 18.8 19.3 19.9 20.4 12.6 16.2 16.8 17.3 17.8 17.6 = 6.6 9.0 e + xp a 4 ( 151.9 256.2 360.5 436.0 511.5 587.0 662.4 102.4 178.8 288.4 398.0 508.4 618.7 33.7 92.8 15.8 25.9 a I nD 0 + ) a ( 1 I 104.4 111.6 118.7 114.3 144.0 10.3 26.2 42.2 66.1 90.0 97.2 21.6 36.5 60.5 84.5 I nH 4.1 6.7 nA ) + + a 112.6 169.8 195.6 221.6 247.5 273.2 128.5 185.7 228.9 272.0 a 13.5 34.4 55.3 10.3 40.8 71.2 3.1 5 2 X I nD + a 6 + 18.6 28.3 35.2 42.3 49.3 56.2 13.6 20.7 27.7 33.2 38.7 X 1.9 5.4 8.9 0.7 2.5 8.1 a of thisresult, theconstanta According interpretation to ageometric andХ=0forfor Altai. Mongolian Khentii mountainous regions Х=1 of Mongolia: encodingdatabelongingvariable to two where structure: calculated theequationoffollowing and variable regionstwo by thebinary dataofthe said, we codedtheharvest exponent (Battulga et al. 2013). With that slope oftheregression line, i.e. allometric in thevalueofintercept, butalsointhe сm; D –stem diameter atbreast height, А–tree condition,kg; age, yrs; in dry ( P 3 I I nD a nH = ) 15.0 20.3 25.8 31.2 13.8 19.5 36.4 53.2 Н – tree height, m; e 1.3 1.6 2.0 5.9 9.6 1.0 1.4 4.7 7.9 + + xp a a P 7 4 X a ( –aboveground tree biomass I a ( nD 0 I nH Aboveground 0.6 0.7 0.8 1.7 2.5 3.7 4.8 5.9 7.0 0.5 0.6 1.3 2.2 3.4 4.6 6.2 7.8 + ) a ( I 1 ) nH I nA 120.2 181.9 214.3 246.7 279.2 311.4 145.7 209.8 271.5 333.0 , 15.4 36.7 58.1 12.3 46.8 81.3 4.6 ) + + a a 5 2 X I nD + 2100 2900 a – the binary Х – the binary Biomass 6 5 + characterizes characterizes X a ( 3 I I nD nH ) + + 18.4 7.5 a a Of 7 4 X ( I

(1) ... nD ( I nH ) ( I ) nH

, 121 GES 03|2019 ) +

a

5 X +

a 6 X ( I nD ) + a 7 X ( I nH ) , P a = e xp 1.487 + 1.658 I nD + 0.16 P a = e xp ( I nD 1.487 ) ( I nH 122 GES 03|2019 ) + 1.658 , and thet for values oftheintercept orontheslopes aboveground tree biomass, noronthe regressions regions ofthetwo for the significant differences the between This meansthat there are no statistically a variables fromvariables X designated accordingly by eight dummy trees, representing eightecoregions, ofthe data forquantity thisstudyis433 Mayr.,entire andL.leptolepis Gord.). The sibirica L., L.gmeliniiRupr., L. cajanderi L. decidua Mill., L. sukaczewii L. N. Dyl., were (respectively additionallytaken of sixspeciesthegenus Larix. spp mentioned database390sampletrees fromin geographical aspect, theabove analysis oflarch biomassofMongolia For ofcomparative thepurpose common dummy variable. isallocated bybiomass ofMongolia one transcontinental analysis, Larix tree comparative And inthefurther environments isobtained: ofMongolia equation, commonto themountain showed thatonlytheconstantsa the standard Statgraphics software Calculating regression equationsusing etal.(Usoltsev 2019). values ofH ofD to theshift owing the violation of allometry corrects ordinates, a regionsvalues ofthetwo ontheaxisof the difference ofthe regression intercept GEOGRAPHY, and a trees were andprocessed harvested The constanta respectively alongtheaxes D axes ofabscissas,via theorthogonal the regression regions slopesofthetwo of Student's criterion t of Student'scriterion respectively, thatismore thant respectively, whichislessthant R P 4 a were statisticallysignificant (thevalue 2 = D = andH e 0.92 I 7 xp nD are 0.85,0.54,0.75,1.80and0.80, act + 1.487 6 for theconstantsa regression lines. The resulting , 0.16 and , whichwasshown earlier ENVIRONMENT, RM 4 of the variable (lnD ofthevariable a 0 S to X to upthestem for small 7 + ( E show thedifference in I 1.658 nD = 1.8 7 ) (Table 2).Sample (Table ( act I nH is8.2and7.8, I , nD ) SUSTAINABILITY + , 1 0.16 R , a andH. 05 2 05 3 , a =2), = 2 )(lnH =2. and 0.92 5 , a ( (2) I nD 6 )

) , ( I RM eliminate thedisplacements causedby been introduced intheequationsto hasdata, acorresponding correction been calculated onthe log-transformed have model designed the in coefficients et al.(Dong 2015). Since the regression of the biomass of all components the regression simultaneously,of thatensures theadditivity stepsequations ofalltwo are evaluated coefficients The 3). (Table estimated by corresponding equations is dividedinto biomasscomponents P et al. 2015),theaboveground biomass step additive equation system (Zheng the disaggregation modelofatwo- (Fig. 3).According to thestructureof equations for aboveground biomass disaggregating additive systems of 2015) andparallel(Zhangetal. 2016) versions: asasequential(Zhengetal. has beenimplemented intwo It fittingapproach.(without additivity) independent the above-mentioned is developed asanalternative to generalto particular" "from principle weighing basedontheproportional The disaggregation method oftwo-step 2018).(Usoltsev underestimated from to two five times biomass oftrees andstandsmay be fine root biomass, the total underground ofmethodsto estimate imperfection tree biomassshowed thatdueto the dataofundergroundanalysis oftheworld the methodoftheirestimation. The sample plots, withoutspecifying often bydetermined researchers notat all there were islackofsuchdata,they consider thebiomassofroots, because is shown inFig. thiscasewe donot 2.In sample plotsintheecoregions ofEurasia for thewhole tree. ofThe distribution 80-100 thesamplesattemperature of drying andafter thebasicdensity) determine (and for ofstems wood alsoto andbark matter content to thedry determine fromtaken eachbiomasscomponent each sample plot. Then samples were in a numberfrom 5 to 10copies on nH a , estimated by theinitialequation, S ) E , = 0 R C,theresults were recalculated 1.8 2 = 0.92 , , RM S E = 1.8 03 (12)2019 , Vladimir А.Usoltsev, IgorМ.Danilinetal. Ecoregions , the the Siberia, Mountains Mountains flood plain flood plain of thePur European Mongolia Japanese Japanese northern northern Western Western Far East, Russian Russian Eastern Eastern Europe part of of part Central Central Siberia, Siberia, steppe Islands Turgay tundra forest- Russia taiga river and and of of Table 2.Aboveground biomassoflarch oftheKhentii trees inthe eastern part

64°03'- 67°00' 36°50' 36°30' 57°50' 55°20' 53°30' 70°00' 49°10' 49°19' 47°00' 60°30' – North Latitude, - -

101°10' 139°40' 138°10' 135°49' - 16°40' 110°00' 148°00'

78°00' 78°00' 48°10' 37°00' 64°30' 09°00' 09°00' – East Longitude, – L. sibirica L. sukaczewii sukaczewii sukaczewii sukaczewii the genus the genus Species of Species of L. decidua L. decidua Larix spp. leptolepis cajanderi cajanderi cajanderi L. sibirica L. sibirica gmelinii (Rupr.) N.Dyl. N.Dyl. Gord. Mayr. Mayr. Rupr. Mill. L.; L. L.; L. L. L. L. L. L. L. L. L. L. L.

Х 0 0 1 0 0 0 0 0 1 Block ofdummy variables Х 0 0 0 1 0 0 0 0 2 Mountains Х 1 0 0 0 0 0 0 0 3 Х 0 0 0 0 1 0 0 0 4 Х 0 0 0 0 0 0 0 1 5 Х 0 0 0 0 0 1 0 0 6 Х 0 1 0 0 0 0 0 0 7 (1.0÷35.0) (2.1÷38.0) (0.3÷22.7) (4.0÷35.9) (6.2÷28.0) (7.1÷47.8) (3.9÷52.8) 17.6±11.2 18.3±11.7 (0.5÷32.3 brackets) (cm) and (cm) and DBH ±SE DBH ±SE its range its range 11.5±7.1 15.2±6.6 16.1±6.2 12.7±7.7 8.2±7.1 7.6±4.9 (in the (in the Mean Mean Aboveground (2.9÷24.8) (4.3÷26.7) (2.3÷28.0) (7.9÷17.8) (1.4÷14.8) (1.5÷24.3) (9.8÷34.0) (2.9÷30.0) brackets) 11.1±5.0 13.2±5.2 14.6±2.4 11.3±5.0 17.7±5.5 13.9±7.0 ±SE (m) ±SE (m) 8.9±5.7 6.5±2.8 and its and its height height (in the (in the range range Mean Mean tree tree

116 73 trees of Number 25 28 66 61 19 45 Biomass Shchepashchenko, Shchepashchenko, 1962; Karaseva, 1962; Karaseva, Karizumi, 1974 Karizumi, Usoltsev, 2016 Usoltsev, 2016 Group..., 1964; Dylis, Nosova, Nosova, Dylis, Battulga et al. Battulga etal. Burger, 1945; Danilin et al., Danilin etal., Danilin et al., Danilin etal., Vyskot, 1982 Vyskot, Moskalyuk, Moskalyuk, Polikarpov, Research Research Sources 1977; 1977; 2015; 2013; 2015 2003 2015 2015 Of

...

123 GES 03|2019 124 GES 03|2019 GEOGRAPHY, Fig. ofEurasia, ofthesampleplotsonterritory 2.Distribution onwhichlarch trees Table additive ofthetwo-step 3.Structure model(AM), implemented according to have beenharvested. The samplearea intheKhentii Mountainsismarked withthe Step 2b Step 2а the principle of proportional weightingthe principleofproportional (Zhengetal. 2015;Dongetal., 2015). For sequential (left) andparallelsequential (right) (left) schemes. The schemesshow therelationship Fig. 3. The pattern ofthedisaggregating weighting proportional two-step AM of Step 1 disaggregation andfrom indicate right to summation left (Zhengetal. 2015; between eachbiomasscomponent, where to linesfrom right indicate left ENVIRONMENT, symbols here andbelow seeequations (3)and(4) P P P w c f = = = 1 1 1 + + + a a a a a a bk SUSTAINABILITY w c s f b D D D D D D 1 1 1 b b b b b b b w c s b f k H H H H H H Zhang etal., 2016) c c c c c c w b s c b f k black square P P P a c s P P P bk b s = = = 1 1 1 + + + a a a a a a bk f c b s w D D D D 1 1 D D 1 b b b b b b c b s b f w k H H H H H H c c c c c c b s c w f b k 03 (12)2019 P P P a c s I nP i = a i + b i ( I nD ) + c i to 2).Σg 7(Table index (code) ofdummy variables, from 0 of numerous j- independentvariables; where regressionInitial equationsare calculated: RESULTS 1972). (Baskerville variables transformationthe logarithmic ofthe Vladimir А.Usoltsev, IgorМ.Danilinetal. and stem bark (bk); and stem bark (s), foliage (f),branches b), stem wood (w) aboveground (a),crown (c),stem over bark kg; and stem diameter ofsampletrees given data oftree height the rangeofharvest a confidenceof0.95. The modelisvalidin for are variables significant numerical with (4) equations of coefficients regression All form (4)isgiven in Table A.1(Appendices). transformation anditsreductionto the for theircorrections after thelogarithmic are obtained,and theircharacteristics that coefficients of equations (3) are calculated regressionof multivariate analysis, the According to thestandard program procedure model(1)hastheform: of thej-thecoregion. After anti-log forvariables thei–thbiomasscomponent ( I P I nP height, i nH Tree = i istheindexofbiomasscomponent: i 10 m = e ) ai + a P D i i d isbiomassofi-thcomponent, + bi i ( H b I in Mongolia byin Mongolia tree height andstem diameter; unitvalues inkg Stem andbark Aboveground i nD components ( c Stem wood Tree crown I i Stem bark Table biomasstable of theSiberianlarch 4.Elaborated dry D Branches nD Biomass Foliage ) ( di ij I X ( ) nH I nH j +

is theblockofdummy a ) c e ) i i …d ( + gi I nH j X j g i are coefficients are i ) j X + 15.64 19.77 24.14 j d 4.12 3.29 1.09 4.38 10 , i ( I nD ) ( I nH 25.49 32.57 10.40 42.97 (4) (3) 7.08 8.05 2.35 14 ) + 4 for thegiven diameter andheightdata. corresponding fragment ofthefinal Fig. of larch trees biomassstructureby the ofsomeregionalthe comparison features we willlimitourselves to article, journal ecoregions exceeds theformat ofthe Since the volume of all the tables for eight in thelarch forests 4). (Table ofMongolia biomass structureandintended for itsuse normative tablewasobtained, additive in of thedummy X variable the given valuesofDand H andthevalue resulting A.3)according model(Table to tabulation ofthe A.3 (Appendices). By the final ofthatis form given in Table wasobtained,double harmonization composition oflarch tree biomasswith the transcontinentalAM ofcomponent weighing,step schemeofproportional (Appendices), according to the two- of theAM presented in Table A.2 A.1 (Appendices) inthestructure of theinitialequationsfrom the Table regressioncoefficients the substituting By A.1). (Table respectively are 0.992,0.903and0.852 biomass ofstem, branchesandfoliage for aboveground biomassis0.991,andfor in Table 2. The coefficient of determination g i j X 10.43 36.09 46.52 15.43 19.55 66.07 4.12 18 j , DBH, сm Aboveground 22 ------6 26 =1,aregional ------Biomass 30 ------Of

...

125 GES 03|2019 GEOGRAPHY, ENVIRONMENT, SUSTAINABILITY 03 (12) 2019

Aboveground 38.99 71.76 113.16 162.80 220.36 - Tree crown 2.16 5.67 11.60 20.43 32.59 -

03|2019 Foliage 0.55 1.31 2.49 4.12 6.22 - 18 Branches 1.60 4.36 9.11 16.32 26.36 - Stem and bark 36.83 66.08 101.56 142.37 187.77 -

126 GES Stem wood 30.83 54.72 83.39 116.08 152.18 - Stem bark 6.00 11.36 18.17 26.29 35.59 - Aboveground - - 156.30 227.40 310.70 405.94 Tree crown - - 7.53 13.66 22.39 34.12 Foliage - - 1.63 2.78 4.30 6.24 26 Branches - - 5.89 10.89 18.09 27.88 Stem and bark - - 148.77 213.74 288.31 371.81 Stem wood - - 125.80 179.51 240.73 308.86 Stem bark - - 22.97 34.23 47.58 62.95 Evaluation of the model The ratio of harvest biomass data and values obtained by the calculation of initial Since it has been established that the equations and AM of tree biomass, shows elimination of the internal inconsistency the degree of correlation of these values of biomass equations by ensuring their and the absence of visible differences additivity does not necessarily mean the in the structure of residual dispersions increase in the accuracy of its estimates obtained from the two models (Fig. 4). (Cunia and Briggs 1984; Reed and Green 1985), it is necessary to clarify whether The results of the comparison indicate that the obtained AM is adequate enough and there is no absolute superiority in terms of how its characteristics relate to the indices adequacy of either the initial or additive of the adequacy of initial equations. equations (Table 5).

Fig. 4. The ratio of the harvest biomass and its values obtained by calculating the initial (a) and additive (b) models of the larch tree biomass determine its distribution by thediameter itsdistribution determine rate of larchcycling substances will largely biomass component and, accordingly, the ofoneoranother share ofparticipation tree, acandidate for dying. Therefore, the and attheageof35years itisadepressed the ageof15years suchtree istheleader, cenotic positionofequal-sized trees: at (Usoltsev, ofthe 1972)due to theageshift times, andfor theaspen–by 2.4-4.4times at theageof35years atthebirch by 1.5-2.0 12 cmattheageof15years exceeds that crown mass of the tree with a diameter of Forstands iswell known. example, the of equal-sized trees andtheirageinforest relationship thecrown between biomass under influenceoftree age. Anegative increase of tree height by 2.5-2.6 times same stem diameter decreases with an But foliage andbranchesbiomassofthe increase oftree heightfrom 10to 26m. of 18cmincreases by 3.2times, withan biomass oftrees having thediameter regardless oftree height. The stem increases asthestem diameter increases, of alltree componentssignificantly significantly.and tree heightvary Biomass dependence ofthevaluestem diameter in thestructureoftree biomassin As shown in Table ofchanges 4,patterns Vladimir А.Usoltsev, IgorМ.Danilinetal. Note. *Symbolsdesignations seeequation(3). DISCUSSION Adequacy Adequacy indices RMSE RMSE Bias Bias R R 2 2 Table 5.Comparison indices for ofadequacy theinitialand 35,61 35,61 0,973 0,973 -1,91 -1,91 P a additive equations oflarch tree biomass 20,63 19,85 0,988 0,989 -2,09 -2,20 P s 32,48 32,42 0,974 0,974 -1,90 4,11 P Biomass components* w Initial equations Initial and heightoftrees. diameter of14cm. Battulgaet al. conclude regions was made for trees with a stem ofabovegroundcomparison biomassby are believed by Battulga etal. (2013).Our other regions and not astough asthey approximately of thesameas in majority conditions of larch are inMongolia When following ourresults, thegrowing 2001). ofstem andSperry wooddensity (Hacke in low-yielding stands with a high basic permafrost conditions, by tree growth the forest-tundra may beexplainedby significant excess oflarch biomassin of trees in40-year-old plantations. The developeddue to crown theabnormally by thebiomassoflarch inCentral Europe isoccupied intermediate position(64kg) (58kg). the mountainsofMongolia The kg). Within thisrangethere are trees in value of biomass is much lower (53-60 andintherest(77 kg), ofecoregions the communities growing onpermafrost diameter andheight occur in open larch tree aboveground biomassofequalboth According to Fig. 5thegreatest valuesof Ecological considerations trees from different Eurasian ecoregions. Comparison between biomassoflarch 11,92 0,910 0,956 -8,89 -1,79 8,38 AM P bk Aboveground -19,83 -19,68 18,53 18,08 0,594 0,613 P cr -20,91 -20,90 16,55 16,35 0,575 0,585 P b Biomass -21,98 -21,89 0,595 0,624 2,45 2,36 P Of f

...

127 GES 03|2019 GEOGRAPHY, ENVIRONMENT, SUSTAINABILITY 03 (12) 2019 03|2019 128 GES

Fig. 5. Changes in the structure of the estimated biomass of trees with 14 cm

diameter and 13 m heightin different ecoregions. Designations: Х0 - Western and Central Europe, Х1 - European part of Russia, Х2 - Turgay steppe, Х3 - Western Siberia, the flood plain of the Pur river, 4Х - Eastern Siberia, forest-tundra, Х5 - Russian Far East, northern taiga, Х6 - mountains of Mongolia, Х7 - Japanese Islands that there is more aboveground biomass and unaccounted ecological traits, and in large trees of Mongolia compared allometric biomass models including air with equal-sized trees of Iceland, and temperature and precipitation, are known explain this phenomenon by more as models sensitive to climate variables stringent growth conditions in the steppe (Zeng et al. 2017). compared to the conditions of Iceland. They compare the calculated indices of Nevertheless, we cannot know what above-ground biomass. But if we compare proportion of the variability of the the actual values of biomass of the trees residual variance may be explained by having the diameter of 14 cm (see: involving climatic (regional) variables Battulga et al., Fig. 2), then the differences in the model, because this residual are virtually absent. It is possible that the dispersion accumulates both unknown underestimation of biomass in Iceland and unaccounted factors and errors due to the reduced basic density (climatic of measurements in all their phases factor) is compensated by overestimation (assessment of age, diameter, height due to the more developed tree crown of a tree, the selection of supposedly in plantations (cenotic factor) compared representative samples of component to the dense natural stands of Mongolia. biomass, their drying, weighing, etc.). The Thus, some uncertainties in comparison question of what explains the regional biomass between ecoregions can be result differences in the structure of biomass of unaccounted and unknown regional of trees with the same linear dimensions features of age and morphological of their stems, remains open. Here, the structures of forest stands. differences in tree age play an undoubted role, as well as the variation in the The main part of the variability in the morphological structure of stands, which, model (3) is assumed by the stem in turn, is determined by both climatic and diameter and height, and the influence of edaphic factors. This will be the subject of unaccounted factors falls on the residual further research. variance. The dummy variables confined to particular ecoregions extracted some The reasons for the relatively large bias portion of the variability from this residual values in both types of equations, especially variance. Belonging of the sample plots to for crown mass, can be seen in Fig. 5, these ecoregions mediates both climatic showing the lack of homoscedasticity of

similar dataofotherecoregions ofEurasia, with andincomparison in Mongolia larchbiomass dataoftheSiberian trees 1. Onthebasisofobtainedharvest CONCLUSIONS 2018), themodelwillgetmore consistent. etal.and 96trees inPoland (Jagodzinski 600 larch trees inChina(Z (butnotyet published)dataon harvested trees, may be supplemented with database, containing433larch sample our data.If a limited amountofharvest approximation, becauseitisbasedon solution to theproblem onlyinthefirst stock oflarch forests. However, thisisa more accurate assessment of the carbon ecoregions ofEurasia, isdesigned for a The proposed AM, adapted for use in 8 ofsmallesttrees.community value, not peculiar to the biomass of the the regression thebias linedetermines of somesinglelargest trees relative to underestimation ofthecrown biomass As aresult, aslightoverestimation or thelargedetermines valuesofthebiases. remaining trees, andthisphenomenon greater than the total residual bias of the four largest trees to contributes thebias of the three tothe biomass variability the crown biomassvalue. Accordingly, increasesit extremely with increase of (notlog) coordinates:is fulfilledontrivial residual whenthecomparison variance Vladimir А.Usoltsev, IgorМ.Danilinetal. References Icelandic Agricultural Sciences, 20,pp. 125-135. forand volume functions young larch Siberian Iceland. ineastern trees sibirica) (Larix M.-M.,Sigurdsson B. A.C., Brinker B., D.Bjarnadottir Inghammar (2007).Singletree biomass Forestry Research, 24(3), pp. 431−437.doi:10.1007/s11676-013-0375-4. the aboveground intheforest-steppe sibirica biomassofLarix of Journal ofMongolia. Battulga P., Tsogtbaatar J., Dulamsuren C.,and HauckM.(2013).Equations for estimating ofForestCanadian Journal Research, 2(1),pp. 49-53.doi:10.1139/x72-009. regression G.L.(1972).Useoflogarithmic intheestimationofplantbiomass. Baskerville oflarchand functioning forests onpermafrost soils. Lesovedenie, 5,pp. Russian). 13-23.(In Abaimov A.P., Prokushkin S.G.,Z eng etal. 2017) yryanova O. L.N.(1997).Featuresyryanova A.,Kaverzina offormation Academy of Sciences. Forestry BranchofRussian ofSiberian Branch and V.N. of Institute Sukachev University, Garden Botanical oftheUral research oftheUralForest Engineering with theprograms ofcurrent scientific incompliance This studywasconducted height tree measurement. of Eurasia whenusingthediameter and larch forest biomassindifferent ecoregions 4. The results obtainedallow to determine the Far North. of Eurasia, withexcept for forest-tundra in of the mostecoregionsas ofthemajority isapproximatelyof Mongolia thesame having the same sizes in the mountains 3. The aboveground biomassoftrees on itscomponentstructure. species) differences ofequal-sized trees account regional (andcorrespondingly equations, into andinaddition, it takes and of the aboveground biomass ofthecomponent contradictions internal ways:harmonized in two it eliminated the 2. The AM oftree biomassofLarix is regionality ofthemodelwassolved. the combinedproblemand ofadditivity spp. biomasswasdeveloped, andthus a Trans-Eurasian AM of the genus A cknowledgements Aboveground Biomass Of Larix

...

129 GES 03|2019 130 GES 03|2019 GEOGRAPHY, the Government Forest Station,259,pp. Experiment 1-99. of investigationproduction. (I): Methods and estimation of the root biomass. Bulletin of N.(1974). Karizumi oftreeThe mechanismandfunction root in the process offorest Technical University. 376 pp. Russian). (In Povolzhye. ontheMiddle State sibirica M.A.(2003). Larix Mari Karaseva Ioshkar-Ola, doi:10.3390/f9100587. Chronosequence. deciduaMill. stand-level biomassestimationinaLarix Forests, 9,587; P. andHorodecki K., Gesikiewicz M.K., A.M.,Dyderski (2018). Jagodzinski Tree- and tables. ofForest Canadian Journal Research, 10(4), pp. 483-490.doi:10.1139/x80-079. M.W.,Jacobs andCunia T. (1980).Useofdummy to harmonize variables tree biomass Ecology, Evolution andSystematics, 4,pp. 97-115. J.S. U.G., (2001).FunctionalHacke Sperry anatomy. andecological xylem Perspectives in Forest Science, 58(3),pp. 101-115.doi:10.17221/69/2011-JFS. equations atdifferent of Journal . scales–Acasestudy Southern Masson pinein for biomass and dummy modelapproaches compatiblesingle-tree variable to construct Fu L.Y., Z Moscow, Publishing. Nauka 143pp. Russian). (In N.V.,Dylis andNosova L.M.(1977).Biomassofforest region. biogeocenosesunderMoscow 392 pp. Translated underthetitle regressionnyi“Prikladnoi analiz”. Moscow: “Statistika” Publ., 1973. Draper N.R.andSmithH.(1966)AppliedRegression analysis. New York, Wiley Publ. biomass equations. Forest Science, 61(1),pp. 35-45. L.,ZhangLiF.Dong weighting system ofnonlinear proportional (2015).Athree-step 1 (9),pp. 33-36.http://elar.usfeu.ru/bitstream/123456789/4051/1/Danilin.pdf. and birch intheforests Mountains. Eco-Potencial, Khentii andtheEastern ofCentral Siberia Danilin I.М., Tsogt Z.,andUsoltsev V.A. (2015).Aboveground phytomass oftrees oflarch 158-169. doi:10.1134/S1995425514020036. fireMongolia. Problems larchContemporary of Ecology,forests intheNorthern 7(2),pp. Danilin I.M.,and Tsogt ofpost- ofstructureandbiological productivity Z.(2014).Dynamics Ecology, 2(1),pp. 55-65.doi:10.1134/S199542550901010 X. fire progressive of ProblemsContemporary Central Siberia. of successioninthenorth organization of and larchfunctional Danilin I.M. (2009). Structural phytocenosis in a post- SJFS20150508. ofForestJournal doi:10.15372/ Science, 5,pp. 96-104(inRussianwithEnglishsummary). Siberian Mongolia). Ledeb. sibirica Larix (Northern trees Khentii intheEastern larch Danilin I.M.,and Tsogt parameters andphytomass Z.(2015).Morphometric oftheSiberian ofForestCanadian Journal Research, 14(3),pp. 376-384.doi:10.1139/x84-067. Cunia T., R.D. andBriggs (1984)Forcing results. ofbiomasstables:someempirical additivity - Schweizerische AnstaltfürdasForstliche Versuchswesen, 24(1),pp. 7-103. Burger H. (1945). Holz, Blattmenge und eng W.S., Tang R.P., S.Z., Sharma (2012). Usinglinear mixed model and LiH.K. ENVIRONMENT, SUSTAINABILITY Z uwachs. Die Lärche.VII. Mitteilung: Mitteilungen 03 (12)2019 Vladimir А.Usoltsev, IgorМ.Danilinetal. University. Russian). (http://e.lanbook.com/view/journal/91094/). 336рр.(In geography,climate-related weight tables. Yekaterinburg: Ural State Forest Engineering Usoltsev V.A. biomassofforest-forming (2016). Single-tree speciesinEurasia: database, 4,pp. Russian). 77-80 (In Kazakhstan), (BulletinofAgricultural of Science Kazakhstana Nauki Vestnik Selskokhozyaistvennoi Usoltsev V.A. (1972).Birch andaspencrown Kazakhstan. biomass in forests ofNorthern 5, pp. 23-32. gmelinii ofForest ofLarix forestsChina.Eurasian Journal productivity innortheast Research, Shi F., QuL., Wang W., Matsuura Y., Koike T., (2002).Aboveground SasaK. biomassand bitstream/123456789/4062/1/Schepaschenko.pdf. of the North-East Yakutia. Ecо-Potencial, 1(9), pp. Russian).http://elar.usfeu.ru/ 37-40(In D.G.Shchepashchenko (2015). Aboveground phytomass trees Mayr cajanderi in of Larix of Forest Science, 72,pp. 1099–1112. Simultaneous estimation as alternative to independent modeling of tree biomass. Annals A.P.D., A.,Corte1 Sanquetta C.R.,Behling Netto S.P., (2015). A.B., doAmaralM.K. Schikowski GijutsuKyokai, Ringyo Tokyo. 61pp. of theForest. Part II.Larch leptolepis Gord.) (Larix forests (1964).Nippon ofShinshuDistrict Research Group ofForest Productivity oftheFour Universities: StudiesontheProductivity doi:10.1139/x85-193 . using nonlinearmodels. ofForest Canadian Journal Research, 15(6),pp. 1184-1187. D.D.,Reed andGreen E.J. (1985).Amethodofforcing ofbiomasstableswhen additivity Pozdnyakov larch. (1975).Dahurian Moscow, 312pp. Russian). L.K. (In Nauka, Academy USSR,171pp. Russian). ofSci. (In N.P.Polikarpov pine young (1962).Scots forest dynamicsonclearcuttings. Moscow, Research, 31(5),pp. 865-878.doi:10.1139/x00-202. Parresol B.R. ofnonlinearbiomass equations. (2001).Additivity ofForest Canadian Journal Science, 57(1),pp. 8-15. stand withrespect to thinning –European ofForest Journal larch deciduaMill.). (Larix Novák J., SlodičákM.,DušekD. (2011).Aboveground biomassofsubstitute tree species elar.usfeu.ru/bitstream/123456789/4300/1/Moskaliuk.pdf. region. of the Magadan productivity Eco-Potencial., 2 (10), pp. Russian). http:// 17-25 (In Moskalyuk T.A. (2015).Biomassofsampletrees asabasisfor studyingtheforest biological 275pp. Russian). (In Nauka, A.(1971).ProductivityMolchanov oforganic massinforests ofdifferent zones.Moscow, (France),Colloques del’INRA 19,pp. 37-43. Mesures desbiomassesetaccroissements forestiers. In: aspects) and practical Les (weighingmethodological tree aspects components, sampling, statistical, economical datafor L.G.(1983).Collecting biomassequationdevelopment: some Marklund analysis. The Forestry Chronicle, 46(5),pp. 402-405.doi:10.5558/tfc46402-5. ofbiomasscomponentsby for regression A.(1970).Methods additivity Kozak ensuring Aboveground Biomass Of

...

131 GES 03|2019 132 GES 03|2019 z GEOGRAPHY, Received onNovember 29 E.G.,JiaL., zhengC.,Mason Wei S.,SunC.,Duan J. additive biomass (2015).Asingle-tree zhaoQ.,X. Liu Y., Z zhang C., Peng D.-L., Huang G.-S., European of Forest Journal Research, 136 (20), pp. 233–249. doi:10.1007/s10342-017-1024-9. biomass equationsandgrowth modelssensitive to for climate variables spp. Larix inChina. tree Individual W.T.(2017). Zou & Y.Pu X.J., X.Y.,Wang Chen X.D., Lei H.R., Duo W.-S.,Zeng 10.1007/s00468-014-1112-0. biomass equations. origin-based single-tree to construct Trees, 29(1), pp. 275-283. doi: aPrirodnich Matematických Rada Véd. Praha, 92(8),pp. 1-162. inbiomass. deciduaMill. Akademie Československé Rozpravy M.(1982).Larix Véd.Vyskot Eurasia. Ecological Questions, 30(2),pp. 57-67.doi:10.12775/EQ.2019.12. biomass modelsfor spp. sensitive Larix to temperature single-trees andprecipitation in Usoltsev V.A., Z production oftree roots? Eko-Potencial, 4(24),pp. 24-77.(inRussian). Usoltsev V. basementsofthebiosphere: А.(2018)In abouttheprimary What we know http://elar.usfeu.ru/handle/123456789/6550.summary). attempt of theiranalyticalreview. Eko-Potential, 2(18),pp. 23-46(inRussianwithEnglish Usoltsev V.A. and the (2017). On additive models of tree biomass: some uncertainties doi:10.1007/s00468-014-1148-1. model ofQuercus BlumeforestsChina. variabilis inNorth Trees, 29(3),pp. 705-716 olgensisplantations. Forestry ofLarix Journal Research, 22,pp.age-sequence 71-76. f7020032. trees inpoplarplantationsJiangsuProvince, China. Forests, 7(2),32.doi:10.3390/ equations bothcompatiblewithtree volume equationsandadditive systems for single- eng W.-S. (2015).Usingnonlinearmixed modelanddummy modelapproaches variable ENVIRONMENT, ukow ukow W., A.A., Osmirko Tsepordey V.P. I.S.,Chasovskikh (2019).Additive eng D. H.(2011).Aboveground allocationinan biomassandnutrient th , 2018 SUSTAINABILITY Z eng W.-S. aboveground (2016). Developing biomass Accepted onAugust 8 03 (12)2019 th , 2019