3. MetHodS to CAlCulAte CARBon StoCKS

Transcription

1 3. Methos to lulte ron stoks Two groups of t were use to evelop moel of ron stoks nmely: 1) site-speifi mesurements otine through fiel survey; n 2) sptil t lyers erive from GIS, remote sensing n environmentl moelling. Site-speifi t inlue inventory plots with mesurements of the numer n imensions of plnts (usully stem imeter n height) within efine re. Allometri equtions were use to estimte iomss from the mesurements of plnt imensions. To extrpolte ron stoks lulte from site-speifi mesurements to other lotions requires sptil t lyers tht esrie the sptil vrition in environmentl ftors tht influene iomss. These t lyers (see Tle A1) esrie the vegettion, limte, geology n ln-use history ftors of mining explortion, timer utting n wilfire ourrene. Correltions in vrition of the sptil environmentl ftors n iomss t the fiel sites form the lgorithms use in the extrpoltion moel of ron stoks. 3.1 Site-speifi t Fiel t ANU reserhers otine site inventory t of trees n shrus uring three fiel surveys from 2005 to Dt were otine for 21 sites of 250 m 250 m in re (6.25 h). The lrge sites were use: 1) to otin representtive mesurements of tree n shru size istriutions in heterogeneous ommunities; n 2) to orrespon to the pixel size of remote sensing imgery, n thus reue the errors ssoite with up-sling from smll plots to pixels. Sites were mrke out suh tht the vegettion within the ounries ws s homogeneous s possile. Site lotions re shown in Figure 3.1. Figure 3.1 Mp of the GWW showing the lotion of the 21 fiel survey sites (inite y ). Ros (see Tle A1) re shown s lk lines. This figure inorportes ro lotion t tht re opyright Stte of Western Austrli

2 32 At eh site, the mjor nopy lyers were ientifie nmely: 1) monopoil or single-stemme trees with up to three nopy lyers; 2) mllee or multi-stemme lowtree lyer; 3) tll shru lyer; 4) sltush/lueush low shru lyer; 5) mllee shru lyer; n 6) slerophyllous (heth) shru lyer. A Point Centre Qurter (PCQ) metho ws use for smpling (see Bonhm 1989; Tongwy n Hinley 2004). Within sites, 16 point entres were lote using gri, s shown in Figure 3.2. Mesurements were me within 25 m rius of eh point entre. Within eh of the four qurters surrouning point entre, the istne ws mesure to the nerest stem within eh nopy lyer (Figure 3.3). Plnt imensions of stem imeter t 1.3 m height (orresponing to imeter t rest height, h) for the tree nopy lyer n plnt height for ll nopy lyers were lso mesure. Asolute ensity of plnts within eh nopy lyer (units: stems h -1 ) ws lulte from the numer of oservtions n istne (see Mithell 2007; Wre n Petrnk 1981). The fiel t n lultions re summrise in the Appenix n Tle A5. 50 m 250 m 250 m N r1 r2 r3 r4 Point entre oorintes: r23 50 m 50 m 250 m 250 m N r1 r2 r3 r4 Point entre oorintes: r23 50 m Figure 3.2 Lotion of the 16 point entres within stuy site.

3 Figure 53. The Figure PCQ 53. metho The PCQ of mesuring metho of sptil mesuring istriution sptil of istriution Vegettion, of showing Vegettion, mesureme showin single point. single point. Figure 53. The Figure Dvi PCQ metho 53. The Tongwy, Dvi of PCQ mesuring metho Tongwy, sptil LFA istriution mnul LFA of Vegettion, mnul of mesuring sptil istriution of showing Vegettion, mesu single Figure point. 53. single The PCQ point. metho of mesuring sptil istriution of Vegettion, showing Figure 53. The PCQ Dvi metho of Tongwy, mesuring Dvi single point. sptil Tongwy, LFA istriution mnul of LFA Vegettion, mnul showing me single point. Dvi Tongwy, LFA mnul Figure 53. The PCQ metho of mesuring sptil istriution of Veget Dvi Tongwy, Dvi LFA Tongwy, mnul LFA mnul Figure 53. The Figure PCQ single 53. metho The point. PCQ of mesuring metho of sptil mesuring istriution sptil of istriution Vegettion, of showing Vegettio m Figure Figure 53. The 53. The Figure PCQ 53. metho The PCQ of mesuring metho of sptil Dvi mesuring istriution sptil Tongwy, of istriution Vegettion, of LFA showing Vegettion, mnu mesurem show single PCQ metho point. single of mesuring point. sptil istriution of Vegettion, showing mesurements single single point. Dvi Figure point. 53. single Tongwy, The point. Figure 53. PCQ The Dvi metho PCQ LFA metho of mesuring Tongwy, Dvi mnul of mesuring sptil Tongwy, istriution sptil LFA istriution mnul of LFA Vegettion, of Vegettion, mnul showing sho m Figure 53. The PCQ metho Dvi of mesuring Tongwy, sptil istriution LFA of mnul Vegettion, showing mesureme Figure single Dvi 53. point. Tongwy, Dvi LFA Tongwy, mnul LFA mnul single The point. PCQ metho of mesuring sptil istriution of Vegettion, single point. Dvi Tongwy, showing m Dvi Tongwy, LFA mnul single point. Dvi Tongwy, LFA mnul LFA mnul Figure 53. The PCQ Figure metho 53. of The mesuring PCQ metho sptil of istriution mesuring of sptil Vegettion, istriution showing of Vegettion, mesurem Figure 53. Dvi The single PCQ Tongwy, point. Dvi metho Tongwy, LFA of mesuring mnul LFA sptil mnul single point. istriutio Dvi Tongwy, Dvi LFA Tongwy, mnul LFA mnul single point. Figure iomss (AGB) 53. for The eulypt PCQ trees metho using generlise of mesuring equtions. We teste sptil these equtions istriution of using our inventory Figure t tht hve 53. The tree imeter PCQ t metho rest height of (h) mesuring rnging from sptil 2 single to 118 m. point. The llometri equtions n e suivie into two groups: those tht require inputs of 1) stem single imeter point. or sl re, n 2) stem imeter n stem height. An Dvi Tongwy, Dvi Tongwy, LFA m Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Trnset Figure 3.3 The PCQ metho for mesurement within eh qurter ( ) is shown s the istne from the point entre to the nerest stem of eh nopy lyer. Two nopy Figure lyers 53. re The shown PCQ in metho this exmple. of mesuring sptil istriution of Vegettion, showing mesurements t single point Allometri equtions Figure 53. The PCQ metho of mesuring sptil istriution of Vegettion, showing mesurements t Allometri equtions relte tree imensions to iomss n re erive using t single point. from trees tht were estrutively smple n the iomss omponents weighe. Allometri reltionships esrie plnt morphology n vry in reltion to site environmentl onitions n plnt genetis. Allometri equtions o not exist for the Figure 53. The PCQ metho of mesuring sptil istriution of Vegettion, showing mesurements t speies within the GWW. Hene, we ompre llometri equtions evelope for single point. speies elsewhere in Austrlin woolns to ssess their utility for iomss estimtion for tree n shru nopy lyers in the GWW. Trnset Dvi Tongwy, LFA Figure 53. height. The following PCQ re metho Figure the min llometri of 53. mesuring The equtions PCQ for wooln sptil metho trees istriution of in Austrli. mesuring of Veget spti single point. Trnset single point. Figure 53. The PCQ metho of mesuring sptil istr Dvi Tongwy, Dvi LFA Tongwy mnu single point. Trnset A ompiltion of llometri equtions erive for wooln speies in Austrli is tthe in the Appenix (Tle A6). Aitionlly, we h ess to more reently erive equtions for some wooln trees n mllees in south-western Western Austrli (J. Jonson unpulishe; see Tle A7). Here, we present omprisons of preite ove-groun ssumption of (1) is tht there is onsistent reltionship etween stem imeter n stem Burrows et l. (2000) mesure iomss of stns of Eulyptus rer, E. melnophloi n E. populne in entrl Queensln rngelns. They erive the generl reltionship (Eqution 3.1). Dvi Tongwy, LF 33

4 Eqution 3.1 AGB = 6.51B 6.65 (t h -1 ) In Eqution 3.1, B is the stn sl re (m 2 h -1 ) (erive from imeter t 1.3 m height; h rnge m). This eqution yiels negtive vlues of AGB for sites where the stn sl re is less thn m 2 h -1. This prolem n e irumvente y foring the regression through zero using the simplifie Eqution 3.2. Eqution 3.2 AGB = 6.10B (t h -1 ) Estimte site iomss of the tree nopy lyer, se on Equtions 3.1 n 3.2, is presente in Tle 3.1. Using these equtions, it is ssume tht AGB is linerly relte to stem sl re tht is, tht stem height woo si ensity hs onstnt vlue (6.1 t h -1 ). It is, however, iffiult to ount for tree mge tht is ommon in nturl stns for exmple, ue to fire n win using simplifie liner reltionship. For iniviul Eulyptus melnophloi (silver-leve ironrk) smple t two sites (380 km istnt) in entrl Queensln, Burrows et l. (2001) erive Eqution 3.3. Eqution 3.3 ln(agb) = 2.700lnC (kg tree -1 ) In Eqution 3.3, C is the stem irumferene t 30 m ove groun (h rnge m). The si ensity 6 of this speies is ~1080 kg m -3 (Boln et l. 1984:532) vlue similr to the si ensity expete for GWW eulypts. For mllee eulypts (Eulyptus soilis/e. umos omine), t the lotion 33º53 S, 146º30 E (h rnge m), Burrows et l. (see Emus et l n Tle A6.1) erive Eqution 3.4. Eqution 3.4 ln(agb) = 2.262lnC (kg tree -1 ) The generl eqution evelope y Jonson (unpulishe) for eulypt wooln speies (h rnge m) in the whetelt of south-western Western Austrli (see Tle A7) hs the sme form s Equtions 3.3 n 3.4. The eqution of Jonson, however, preits AGB up to ~20 per ent smller thn Equtions 3.3 n 3.4 (Figure 3.4) for the mximum stem h mesure t our fiel sites. Willims et l. (2005) hve erive generl llometri eqution for estimting AGB of trees (mostly eulypts) in tropil n sutropil eulypt woolns ross the Northern Territory, Queensln n New South Wles (Eqution 3.5). Eqution 3.5 ln(agb) = ln(D) (ln(H)) 2 (kg tree -1 ) In Eqution 3.5, D is the h t 1.3 m (h rnge 3 86 m) n H is the tree height (m). Willims et l. foun tht there ws goo (1:1) greement etween oserve n preite vlues of AGB following the k-trnsformtion of the logrithm. This implies tht no orretion ftor nees to e pplie to just for is. Although the Jonson eqution oes not inlue H, there is goo greement etween the preitions of 6 The si ensity of woo is the rtio of ry mss to fresh volume. 34

5 AGB y the Jonson n Willims equtions (Figure 3.5). In Figure 3.5, there is some stter of points roun the regression line euse the reltionship etween h n height is not onsistent in the fiel t set. Tht is to e expete euse our fiel t inlue wie rnge of speies n ge lsses n mny stems hve suffere mge from fire, timer utting, win n termites. The GWW inventory t inlue mesurements of height n h, so Eqution 3.5 ws use to estimte AGB of trees hving h of 10 m or more t our fiel sites AGB Jonson (tonnes) AGB Burrows (tonnes; Eqn 3.3 & 3.4) Figure 3.4 Comprison of AGB preitions from stem imeter (h) using the generl llometri eqution of Jonson n the omine equtions (Eqution 3.3 n 3.4) of Burrows et l. The t points re lulte AGB vlues (tonnes per tree) from fiel mesurements of h of iniviul trees. The re line inites 1:1 reltionship AGB Jonson = 0.96 AGB Willims R² = AGB Jonson (tonnes) AGB Willims (tonnes; Eqn 3.5) Figure 3.5 Comprison of AGB (tonnes per tree) preite for GWW trees using Eqution 3.5 (Willims et l.) n the generl eqution erive y Jonson. The regression line is rwn in lk. The 1:1 line is rwn in re. 35

6 Tle 3.1 Estimtes of ove-groun iomss (AGB, t h-1) for GWW fiel sites. The mixe moel AGB omprises the omine estimtes of tree n shru AGB from Equtions 3.5 n 3.7. Site no Survey ID Fiel vegettion esription Stn (tree) sl re (m 2 h -1 ) AGB Burrows Eqution 3.1 (t h -1 ) AGB Burrows Eqution 3.2 (t h -1 ) Mixe moel AGB (t h -1 ) _01 Eulypt wooln _02 Open eulypt wooln _03 Low shruln no t _04 Low shruln no t _05 Eulypt wooln _06 Eulypt wooln _07 Eulypt wooln _08 Tll shruln no t _09 Eulypt wooln _10 Tll shruln no t _01 Low eulyptus wooln _02 Eulypt wooln _03 Eulypt wooln _04 Eulypt wooln _05 Eulypt wooln _06 Tll i shruln _07 Tll i shruln _01 Mllee shruln _02 Eulypt wooln _03 Eulypt wooln _04 Heth/mllee shruln no t

7 Bejn et l. (2008) hve propose tht tree stems n rnhes hve ner-onil shpe, relte to the mehnis of struturl support n wter flow through the stem n rnhes. In tht se, there oul e simple reltionship etween AGB n the volume of one or yliner (the volume of one is equl to one-thir of the volume of yliner). For our fiel t set of eulypt trees, AGB preite y Eqution 3.5 is pproximtely equivlent to Eqution 3.6. Eqution 3.6 AGB yl = πr 2 h0.56 (kg tree -1 ) In Eqution 3.6, r is the stem rius t 1.3 m, h is the height to the top of the nopy n the units re metres. The omprison etween AGB yl n AGB preite from Eqution 3.5 is shown in Figure 3.6. We use this simple reltionship to estimte the AGB of the shru lyer AGB Willims = 0.99AGB yl R² = 0.99 AGB Willims (tonnes; Eqn 3.5) AGB yl (tonnes) Figure 3.6 AGB (tonnes per tree) estimte from eulypt tree n mllee mesurements from our site t. The llometri eqution of Willims evelope for northern Austrlin eulypts yiels very similr estimtes to those erive from the formul for trunte one (0.56πr 2 h, in whih the onstnt ws erive from lirtion with Eqution 3.5). Estimtes of forest iomss ommonly ignore the ontriution of the shru lyer; however, out 6 million h of the GWW hs n upper nopy lyer of low trees or shrus (nopy height lsses L, S n Z in Tle 2.1). We mesure nopy imensions n height ut not h for mny trees with h <10 m n for lmost ll of the shrus reore in our fiel t. Consequently, none of the llometri equtions requiring stem imeter mesurement ws useful for estimting shru iomss. Allometri equtions, whih utilise nopy imensions to preit ri n semi-ri shru iomss, re onsiere inequte for sites istnt from the lotions from whih the equtions were erive (Den n Elrige 2008). (The equtions in tht stuy yiele iomss estimtes tht iffere y 85 per ent.) An eqution tht estimtes AGB from height t the top of the nopy ws require. We oul fin no suitle pulishe equtions (the equtions of Hrrington 1979 relte iomss to height ut the speies n form of the shrus re ifferent to those t the GWW). Inste, we erive n eqution relting AGB to height for 114 plnts (mostly eulypts) tht hve h <10 m for whih we reore h mesurement. First, we lulte AGB using: 1) the generl llometri eqution of Jonson; 2) Eqution 3.4; 3) Eqution 3.5; n 4) Eqution 3.6. Of these equtions, the est preitor of AGB from height (m) ws Eqution 3.6 (see Figure 3.7 n Eqution 3.7). 37

8 Eqution 3.7 AGB = h (kg plnt -1 ) This eqution ws susequently use to estimte the AGB of shrus t eh fiel site. The summe AGB for ll nopy lyers is presente in Tle 3.1. Clulte AGB (tonnes) Eqution 3.5 Jonson Eqution 3.6 Eqution 3.4 AGB = h R² = Height to top of nopy (m) Figure 3.7 Reltionship etween mesure nopy height n lulte AGB (tonnes per plnt) for 114 smll trees or shrus hving h <10 m. The eqution isplye on the grph (Eqution 3.7) yiels the est preition of AGB (se on Eqution 3.6) from height for these smll trees n shrus. Below-groun iomss (BGB) or root iomss is usully estimte s rtio of ovegroun iomss. The most pproprite rtios to use for the GWW re se on site t for similr speies n environmentl onitions; estimtes otine for rnge of Austrlin woolns re summrise in Tle 3.2. The vlue otine from the generi equtions of Jonson (n..) ws in lose greement with the vlues given in Tle 3.2 for Okvle n Howr Springs. Therefore, we use the rtio of BGB:AGB otine from the nlysis of Jonson to estimte BGB t our fiel survey sites. Tle 3.2 Summry of estimte rtio of elow-groun iomss to ove-groun iomss for rnge of Austrlin woolns. Speies Lotion Rinfll (mm yr -1 ) BGB:AGB Soure E. populne Okvle, NSW Zerihun et l. (2006) E. populne Rom, Ql Zerihun et l. (2006) E. minit, E. tetront Howr Springs, Humpty Doo, Willife Prk, NT 1200 > Chen et l. (2003) Severl eulypts, see Tle A3 Whetelt, WA ~ Jonson 38

9 We h no fiel mesurements of the three omponents of e ove-groun iomss (AGB e ) for the GWW: stning ewoo (SDW), orse wooy eris (CWD) on the groun n litter (Lit). In review of AGB e in Austrlin forests (Wolenorp n Keenn 2005), t from two wooln stuies gve AGB e equivlent to 26 per ent of the totl AGB totl (AGB living + AGB e ). For open forests, the perentge of AGB e :AGB totl (five stuies) ws 25 per ent. SDW ws ~7 per ent of AGB totl in the TRAPS stuy in Queensln woolns (Burrows et l. 2002). In ontrst, the men perentge of SDW:AGB totl reporte y Wolenorp n Keenn (2005) ws 15.6 per ent for the two wooln stuies n 1.9 per ent for the five open-forest stuies. Bse on these stuies, we ssume tht AGB e :AGB totl for woolns within the GWW is 26 per ent. Totl iomss ws lulte s the sum of AGB living, BGB n AGB e. The site-se estimtes of totl iomss re given in Tle 3.3. The sites hving the gretest totl iomss ( t h -1 ) re eulypt woolns. In ontrst, the totl iomss of the mllee, heth n other shruln sites rnges from 2 to 50 t h Dt lyers for sptil extrpoltion In this setion, we investigte how rnge of ftors for whih we hve sptil t oul influene iomss. Those ftors tht re shown to e importnt re onsequently tken into ount in the sptil moelling proess y whih we extrpolte from the 21 fiel sites to estimte the totl iomss n ron stok of the ~16 million h of the GWW Vegettion mpping The NVIS MVG lsses (DEWHA 2005) ompre resonly well with our oservtions t our fiel sites (Tle 3.4). In some ses, however (speifilly sites 3, 4, 10, 11 n 21), we i not oserve the upper tree or shru lyer, n in some ses (sites 1, 2 n 9) we oserve tree lyer t sites lssifie y NVIS s mllee wooln n shruln. One possile explntion for these isprities is tht the vegettion hs hnge in the perio sine the NVIS mpping t were ollete (whih ws t vrile times efore 2005 in ifferent regions of Austrli). As note, the GWW vegettion hs een sujete to severl types of isturne uring the pst entury, inluing fire, timer utting, mining n minerl explortion. The res known to e ffete y these isturnes re shown in Figure Some prts of the GWW hve lso een impte on y pstorlism. 39

10 Tle 3.3 Estimte ove-groun iomss, totl iomss n orresponing eril photogrphs of the 21 ANU fiel survey sites. Site esription Site _01 Eulypt wooln AGB estimte (t h -1 ) 46 Totl iomss (live+e) (t h -1 ) 102 Site esription Site _02 Open eulypt wooln AGB estimte (t h -1 ) 6 Totl iomss (live+e) (t h -1 ) 14 Site esription Site _03 Low shruln AGB estimte (t h -1 ) 0.6 Totl iomss (live+e) (t h -1 ) 1.3 Site esription Site _04 Low shruln AGB estimte (t h -1 ) 2 Totl iomss (live+e) (t h -1 )

11 Tle 3.3 Continue Site esription Site _05 Eulypt wooln AGB estimte (t h -1 ) 36 Totl iomss (live+e) (t h -1 ) 79 Site esription Site _06 Eulypt wooln AGB estimte (t h -1 ) 48 Totl iomss (live+e) (t h -1 ) 105 Site esription Site _07 Eulypt wooln AGB estimte (t h -1 ) 49 Totl iomss (live+e) (t h -1 ) 108 Site esription Site _08 Tll shruln AGB estimte (t h -1 ) 2 Totl iomss (live+e) (t h -1 )

12 Tle 3.3 Continue Site esription Site _09 Eulypt wooln AGB estimte (t h -1 ) 32 Totl iomss (live+e) (t h -1 ) 70 Site esription Site _10 Tll shruln AGB estimte (t h -1 ) 5 Totl iomss (live+e) (t h -1 ) 11 Site esription Site _01 Low eulypt wooln AGB estimte (t h -1 ) 11 Totl iomss (live+e) (t h -1 ) 24 Site esription Site _02 Eulypt wooln AGB estimte (t h -1 ) 58 Totl iomss (live+e) (t h -1 )

13 Tle 3.3 Continue Site esription Site _03 Eulypt wooln AGB estimte (t h -1 ) 47 Totl iomss (live+e) (t h -1 ) 104 Site esription Site _04 Eulypt wooln AGB estimte (t h -1 ) 49 Totl iomss (live+e) (t h -1 ) 108 Site esription Site _05 Eulypt wooln AGB estimte (t h -1 ) 38 Totl iomss (live+e) (t h -1 ) 84 Site esription Site _06 Tll i shruln AGB estimte (t h -1 ) 4 Totl iomss (live+e) (t h -1 )

14 Tle 3.3 Continue Site esription Site _07 Tll i shruln AGB estimte (t h -1 ) 6 Totl iomss (live+e) (t h -1 ) 13 Site esription Site _01 Tll mllee shru over low mrlok, melleu unerstorey AGB estimte (t h -1 ) 23 Totl iomss (live+e) (t h -1 ) 50 Site esription Site _02 Eulypt/melleu wooln AGB estimte (t h -1 ) 38 Totl iomss (live+e) (t h -1 ) 84 Site esription Site _03 Eulypt wooln AGB estimte (t h -1 ) 90 Totl iomss (live+e) (t h -1 )

15 Tle 3.3 Continue Site esription Site _04 Mllee/heth AGB estimte (t h -1 ) 2 Totl iomss (live+e) (t h -1 ) 4.7 The eril photogrphs re reproue y permission of WA Ln Informtion Authority, CL34/2010 (< Tle 3.4 Comprison of vegettion esription se on fiel oservtion n NVIS Mjor Vegettion Group (MVG) lss. Site no. Survey ID Fiel vegettion esription NVIS-MVG extnt lssifition oe NVIS-MVG extnt vegettion esription _01 Eulypt wooln 14 Mllee woolns n shrulns _02 Open eulypt wooln 14 Mllee woolns n shrulns _03 Low shruln 14 Mllee woolns n shrulns _04 Low shruln 14 Mllee woolns n shrulns _05 Eulypt wooln 5 Eulypt wooln _06 Eulypt wooln 5 Eulypt wooln _07 Eulypt wooln 5 Eulypt wooln _08 Tll shruln 17 Other shruln _09 Eulypt wooln 14 Mllee woolns n shrulns _10 Tll shruln 14 Mllee woolns n shrulns _01 Low eulypt wooln 5 Eulypt wooln _02 Eulypt wooln 5 Eulypt wooln _03 Eulypt wooln 5 Eulypt wooln _04 Eulypt wooln 5 Eulypt wooln _05 Eulypt wooln 5 Eulypt wooln _06 Tll i shruln 15 Low lose forest n tll shruln _07 Tll i shruln 15 Low lose forest n tll shruln _01 Mllee shruln 14 Mllee woolns n shrulns _02 Eulypt wooln 5 Eulypt wooln _03 Eulypt wooln 5 Eulypt wooln _04 Heth/mllee shruln 5 Eulypt wooln Soure: NVIS MVG DEWHA (2005). 45

16 3.2.2 Reent fire history The ourrene of wilfire n e etete through the nlysis of time-series of stellite imgery. We use remotely sense t from two stellite-se soures nmely: Lnst imges 7 t 30 m pixel resolution (1972, 1977, 1980, 1985, 1988, 1989, 1991, 1992, 1995, 1998, 2000 n 2002) n MODIS (or Moerte Resolution Imging Spetroriometer) imges t 250 m pixel resolution eginning in 2000 (see Setion for etils). While hving orser pixel resolution, the MODIS t were proesse to proue n essentilly lou-free time-series on one-month time step (see Berry et l. 2007). Using these t, fire footprints over intensely urnt res were highlighte through ifferene imges, whih show hnges tht hve ourre etween susequent imges. We use this nlytil pproh to mp res tht were severely urnt y lrge lnspe fires eginning ir The extent n frequeny of these fires re shown in Figure 3.8. Figure 3.8 Are ientifie s eing urnt y intense wilfires etween 1972 n The se mp is the Lnst 2005 lyer. The fire frequeny inites the numer of times n re within the GWW hs een urnt y intense fires over the perio. The se mp is visile in res within the GWW hving fire frequeny of zero etween 1972 n Fire frequenies re not shown for res outsie the GWW ounry. The slt lkes re oloure lue. Mjor ros re shown in lue n + signs inite the lotion of ANU fiel survey sites. Sptil t soures: see Tle A1 Fire mpping, MODIS, WA Lnover, Ros. This figure inorportes Lnst 2005 t n ro lotion t tht re Copyright Stte of Western Austrli 2007, n Lnst Continentl Mosi (AGO) t tht re Copyright Commonwelth of Austrli We foun tht ll five sites for whih our vegettion struture oservtions iffere from the NVIS MVG mpping were urnt sine 1972, n mostly sine This suggests tht either: 1) t some time fter the vegettion survey for mpping, fire hs ltere the vegettion struture; or 2) the vegettion mp use to inform NVIS epite the limx 7 See Tle A1 for etils. 46

17 vegettion inste of n intermeite suessionl stte. Vegettion struture oul hnge if fire use the eth of ove-groun plnt tissues so tht susequent regenertion is from elow-groun tissues (lignotuers or roots) or from see. The reltive unne of speies n hnge ue to iffering rtes of regenertion. If the fire return intervl is too short then it is possile tht lignotuer resoures will e eplete n no re-sprouting will our. Also, there will e insuffiient time for proution n umultion of see. This oul le to lol loss of speies n possily hnge in the speies omposition of the ominnt nopy lyer (for exmple, eulyptominte nopy oul e reple with n i-ominte nopy). Hopkins n Roinson (1981), in their stuy of the effets of single fire in the GWW, onlue tht the mllee-heth vegettion ws pyri islimx vegettion following fire in eulypt wooln. Ber (1968:251), in his report on mpping of vegettion of the Lke Johnston n Boorin res within the GWW, note tht the eulypt trees tht ominte slerophyll wooln re kille y fires n regenerte from see. Ber lso sttes tht [m]llee is sujet to frequent fires whih estroy the top growth, regenertion tking ple from oppie. It is not ler to wht extent the mllee hit is genetilly ontrolle or ue to fire. Ber notes tht very mny mllee speies n e foun in the form of moertely size or smll trees. In his esription of the sru heth vegettion formtion, Ber (1968:249) oserve tht this vegettion is urnt so frequently tht mture struture hs little hne to evelop. The oservtions of Hopkins n Roinson, n Ber, suggest tht fire hs plye pivotl role in moifying the vegettion struture in the GWW. Oservtions uring fiel surveys etween 2005 n 2007 y ANU reserhers support those of Hopkins n Roinson (1981) n Ber (1968) (Figures 3.9 n 3.10). 47

18 Figure 3.9 Figure 3.9 Figure 3.9 Figure 3.9 Figures 3.9 Erly stges of regenertion of eulypts fter fire. Fire hs kille the tree strtum. For the purpose of ron ounting, the re-growth vegettion is lssifie s low lose forests/tll lose shrulns (NVIS MVG 15). Photos: S. Berry. 48

19 Figure 3.10 Figure 3.10 Figures 3.10 This surviving fire-mge tree () testifies to the pre-fire wooln stte of the vegettion (NVIS MVG 5), whih woul urrently e lssifie s mllee wooln or shruln (MVG 14) or low lose forest/tll lose shruln (NVIS MVG 15). If fire is exlue for severl ees, suession of the vegettion in () will le to slerophyll wooln (NVIS MVG 5) with similr struture to tht shown in (). 49

20 If the vegettion struture mpping in the GWW inlues pyri suessionl vegettion (suh s tht shown in Figure 3.9), we woul expet ongruene etween ounries of vegettion mpping units n fire footprints in ition to ongruene etween the vegettion n surfiil geology ounries 8. We present eviene for ongruene etween fire footprints n vegettion ounries in Figures 3.11 n The mppe (MVG 5) eulypt wooln ounries orer mny footprints of fires in the perio (Figure 3.11). Post-1990 fire footprints our over lrge res tht re mppe s eulypt wooln (Figure 3.11), s woul e expete if the vegettion mp lyer ws erive from pre-1990 informtion. The presene of fire footprints over vegettion mppe s wooln (Figure 3.11), long with our fiel oservtions (Figures 3.9 n 3.10), emonstrtes tht the eulypt wooln hs een extensively urne in the pst two ees. This n e ontrste with the oservtion of Ber (1968:232) tht woolns, when he oserve them, urne to only minor extent. Figure 3.11 Figure 3.11 Figures 3.11 Lotion of fire footprints ientifie from () Lnst stellite imgery for the perio , n () oth Lnst n MODIS imgery for the perio The istriution of eulypt wooln (NVIS MVG 5) is shown s n overly. These figures inorporte NVIS Version 3.0 t n Lnst Continentl Mosi (AGO) t tht re Copyright Commonwelth of Austrli 2005 n As note in Setion 2.3 n Tle A2, however, the erivtion of the vegettion n surfe geology lyers for some mp sheets relie on interprettion from eril photogrphs, the surfiil geology ounries eing inferre from vegettion over. 50

21 The exmple presente in Figure 3.12 n re tht ws isture y timer utting etween 50 n 100 yers go lso provies some eviene to support the proposition tht eulypt wooln n e reple y shruln ominte y single-stemme eulypts (mrlok), mllee or i in the ees fter intense fire. Eulypt woolns (MVG5) re mppe s ourring over wie rnge of surfe geology lsses, with geology hving onsierle influene on the speies represente (Tle A3). In Figure 3.12, low lose forest/tll lose shruln (MVG 15) vegettion mpping oinies with the mpping of the Czs (Cinozoi sns) geology lss, while joining wooln vegettion is mppe on the Czg (Cinozoi grvel) geology. This orresponene etween vegettion lsses n geology lsses is, however, not onsistent ross ll of the mp sheets. On some sheets, eulypt wooln is mppe over Czs. Figure 12 Figure 3.12 Figures 12 Congruene etween fire footprint n eulypt wooln ounries in prt of the GWW. () The vegettion n fire lyers re superimpose on the 1: surfe geology mp for Boorin. The entire re is mppe s eing ut over for mining timer n/or fuel woo. The yer in whih timer utting egn within eh setion (outline in lk) is shown. () The extnt vegettion over the fire footprint is mppe s shruln (NVIS MVG 15 n 16). Dt soure: see Tle A1 Fire Mpping, Mp Sheet, Geology, Timer utting, MVG. This figure inorportes geology t tht re Copyright Stte of Western Austrli 2007, n NVIS Version 3.0 t n Lnst Continentl Mosi (AGO) t tht re Copyright Commonwelth of Austrli 2005 n

22 Given the ove, n s in the semi-ri GWW, the eulypt seelings n splings mking up the post-fire regenertion of wooln form tll shruln (MVG 15; Figure 3.9) for severl ees following fire, we hve upte the mppe istriution of extnt eulypt wooln to ount for the hnges wrought y intense fires in the pst 35 yers y relssifying the 1.46 million h of MVG 5 (eulypt forests n woolns) vegettion tht hs een severely urnt y fires etween 1972 n 2007 into MVG 15 (low lose forests n tll lose shrulns) (Figure 3.13). Figure 3.13 This mp shows the likely urrent istriution of vegettion struturl formtions (NVIS MVGs) in the GWW in It ws me y relssifying the 1.46 million h of NVIS MVGs 4 n 5 (eulypt forests n woolns) vegettion tht hve een severely urnt y fires etween 1972 n 2007 into MVG 15 (low lose forests n tll lose shrulns) to etter esrie the struture of the post-fire regenertion of seelings n root sprouts. Dt soure: see Tle A1 MVG. This figure inorportes NVIS Version 3.0 t tht re Copyright Commonwelth of Austrli The mpping of fire footprints (Figure 3.11) is restrite to fires fter Footprints of fires preting tht time re, however, lso evient in the Lnst stellite imgery. If we ssume tht the NVIS MVG lsses tht inlue Ai n mllee woolns n shrulns (MVGs 13, 14, 15 n 16) rise following intense fire in eulypt woolns n forests in the GWW, we n preit the extent of the ltter in no-fire onition, n this is shown in Figure In this senrio, the eulypt woolns oupy 13 million h of the GWW oule the urrent extent. This estimte is likely greter thn the tul extent of eulypt wooln (MVG 5) struture pre-1750, s our senrio ssumes no fire n no geologil ifferenes unerlying the istriutions of eulypt woolns n mllee formtions. 52

23 Figure 3.14 Preite extent of eulypt forest n wooln n other vegettion groups ssuming fire ws sent from the GWW for mny ees to enturies n NVIS MVGs our t present in lotions tht were previously (tht is, efore pst fire event) eulypt wooln (MVG 5). We susequently refer to this hypothetil vegettion over s the no-isturne senrio. Dt soure: see Tle A1 MVG. This figure inorportes NVIS Version 3.0 t tht re Copyright Commonwelth of Austrli Timer utting The pproximte lotion of timer-utting tivities in the GWW uring the perio (igitise from the mp shown in Figure 2.12) is shown superimpose over the NVIS MVG lyer in Figure Timer utting in this perio ffete pproximtely 4.4 million h of vegettion in the GWW (Tle 3.5). Historil reors inite tht timer utting in the GWW utilise mostly the slmon gum n gimlet woolns (see Setion 2.5). Following timer utting, severl speies of eulypt re le to regenerte from oppie (the uthors hve oserve this t their fiel survey sites). The oppie growth rises from the top of the stump of the ut tree so tht the originl monopoil tree retins monopoil se, whih supports multiple stem sprouts. In ontrst, mllee sprouts rise from the se of the stem to form multi-stemme shru or tree. Aoring to our mp (Figure 3.15) n Tle 3.5, eulypt woolns (NVIS extnt MVG 5 n 11) mke up only 2.5 million h of the timer-utting re. However, 1.1 million h of the re ut over for timer hs mppe vegettion over lss of mllee wooln n shruln (MVG 14), tll lose shruln (MVG 15) or i shruln (MVG 16) (Tle 3.5 n Figure 3.15). Hlf of the 1.1 million h of MVGs 14, 15 n 16 hs susequently een urne y intense fires. In Setion 3.2.2, we propose tht MVGs 14, 15 n 16 oul represent pyri suessionl vegettion. If fire hs ltere the vegettion struture in this wy following timer utting then prior to timer utting the vegettion 53

24 over oul hve een similr to the moelle vegettion over shown in Figure We show this moelle vegettion over, eneth the timer-utting overly, in Figure Uner the no-isturne senrio, 3.6 million h of eulypt wooln woul hve een ville for timer utting ir Figure 3.15 Extent of timer utting in the GWW etween 1900 n 1975 is shown s n overly on the NVIS MVG extnt vegettion lyer. Figure 3.16 Extent of timer utting in the GWW etween 1900 n 1975 is shown s n overly on the no-isturne senrio moelle vegettion lyer (see Figure 3.14). Dt soure: see Tle A1 Timer utting, MVG. These figures inorporte NVIS Version 3.0 t tht re Copyright Commonwelth of Austrli

25 Tle 3.5 Are within eh NVIS MVG lss for extnt (Figure 3.15) n no-isturne (Figure 3.16) senrio vegettion, within the re mppe s ffete y timer utting etween 1900 n The re of eh vegettion type ffete y oth timer utting n fire ( ) is lso shown. MVG lss Are ffete y timer utting (h) Are ffete y timer utting n fire (h) Are ffete y timer utting (h) NVIS extnt vegettion NVIS extnt vegettion Hypothetil no-isturne senrio 4. Eulypt low open forest Eulypt woolns Ai forests n woolns Cllitris forests n woolns Csurin forests n woolns Eulypt open woolns Ai open woolns Mllee woolns n shrulns Low lose forests n tll lose shrulns Ai shrulns Other shrulns Hethlns Hummok grsslns Forlns Slt lkes Clere, non-ntive vegettion Nturlly re sn, rok, lypn Totl

26 3.2.4 Mining n minerl explortion Most of the mining n minerl explortion in the GWW is lote within the ounries of the greenstone lithologil unit (Figure 2.6). The re impte y open-ut mining oul e reltively smll. Minerl explortion hs, however, more profoun impts on the vegettion inluing the use of vehiles off-ro to ess remote ountry, onstrution of ess trks, exvtions to expose eeper soil horizons, onstrution of seismi lines n the utting own n removl of woo for fuel, mine-props, n so on (see Pltes 1 5). A mp showing the lotion of the mjor lithologil units n the ln tenure type is presente in Figure A lrge portion of the grnite-greenstone lithologil unit is uner lesehol tenure n is use for mining n minerl explortion. Plte 1 Professor Brenn Mkey, ANU, stning in typil trenh rete uring minerl explortion, GWW.. Plte 2 Aeril photogrph of n open-ut mine site, GWW. Reproue y permission of WA Ln Informtion Authority, CL34/2010 (< 56

27 Plte 3 Seismi trks ner sites 13 n 14, GWW. Reproue y permission of WA Ln Informtion Authority, CL34/2010 (< 57

28 Plte 4 Aeril photogrph showing the environmentl footprint of nikel mine, GWW. Reproue y permission of WA Ln Informtion Authority, CL34/2010 (< 58

29 Plte 5 Stump n felle tree, Site 14. Severl trees h een felle t this remote site lote on the greenstone elt Ln tenure Four types of ln tenure preominte in the GWW: vnt rown ln (55 per ent), nture onservtion reserve (10 per ent), Aoriginl lesehol (1.5 per ent) n other lesehol (20 per ent) (see Figure 3.17). The re of overge of eh ln tenure type, the ln tenure of the extnt eulypt wooln n the ln tenure of the mppe fire footprint (Figure 3.8) re given in Tle 3.6. Thirty-two per ent of the extnt wooln (s mppe in Figure 3.13) ours on lesehol ln, while 55 per ent is on vnt rown ln n 9 per ent is within nture onservtion reserves. Some ln tenure types pper to e more prone to fire n pproximtely 36 per ent of the vnt rown ln, nture onservtion reserve n Aoriginl reserve lsses er fire footprint ( fire events) in ontrst with 6 per ent of lesehol ln. The lesser fire footprint on lesehol ln oul e onsequene of: 1) less fuel ue to ln lering n grzing; 2) fire protetion n suppression to protet eonomi ssets; n/or 3) fewer ignition events ue to restrite ess. 59

30 For the purpose of ron ounting, it ws neessry to etermine if the vegettion struture (n iomss) hs een moifie y ln lering on ifferent ln tenures in ition to the mppe timer utting (re of timer utting shown in reltion to ln tenure in Figure 3.17). As ll of our fiel stuy sites were lote on vnt rown ln or within nture onservtion reserves, we neee to utilise other t soures to ssess the vegettion over n iomss of lesehol ln. To o this, we utilise remotely sense imgery, s isusse in Setion 3.3 elow. Nonetheless, it is evient from Figure 3.17 tht most of the ln uner lesehol tenure overlies grnite-greenstone lithology (thus is likely to hve een impte on y minerl explortion) n/or hs een sujete to timer utting. Figure 3.17 Overly showing lotion of timer utting superimpose on ln tenure n lithology lyers (Figure 2.6). Dt soure: see Tle A1 Timer utting, Ln Tenure, Lithology, Ros, Towns. This figure inorportes geology n ro lotion t tht re Copyright Stte of Western Austrli 2007, n Austrlin ln tenure (1993) n popultion entre t tht re Copyright Commonwelth of Austrli 1993 n Tle 3.6 Arel extent of ln tenure types within the entire GWW, within the re mppe s wooln n within the re mppe s hving fire footprint etween 1972 n 2007 (see Figure 3.8). Ln tenure type Totl re (h 1000) Wooln re *(h 1000) Totl re urnt (h 1000) 1. Aoriginl reserve Aoriginl lesehol Forestry reserve Other freehol Other lesehol Mixe Nture onservtion reserve Other Vnt rown ln Wter reserve

31 3.2.6 Overview of isturne impts Our nlyses of the ville sptil t in this hpter hve revele tht most of the GWW hs experiene isturne from the mining inustry through minerls explortion n timer utting, from lering or thinning of the vegettion for ropping n grzing or from intense fire (Figure 3.18). All of these isturnes must e tken into ount when extrpolting from fiel-se mesurements of the iomss ron stok to the entire GWW. Figure 3.18 This mp shows the extent of iret (timer utting, minerls explortion n ln-use hnge sine 1900) n iniret (fire etween 1972 n 2007) isturnes in the GWW. The legen for the fire footprints is shown in Figure 3.8. Dt soure: see Tle A1 Timer utting, Ln Tenure, Lithology, Mp sheet. This figure inorportes geology n mp sheet Inex250 t tht re Copyright Stte of Western Austrli 2007, n Austrlin ln tenure (1993) t tht re Copyright Commonwelth of Austrli Reltionship etween ln use/isturne n AGB estimtes for fiel survey sites There is eviene tht fire hs h lrger effet thn timer utting on the ovegroun iomss (AGB) t the ANU fiel survey sites (see Figure 3.19). The AGB of ll fiel survey sites tht hve een urnt y intensive fire sine 1972 is low. AGB of the ANU wooln sites tht hve een ut over is, however, more omprle with tht of most of the other wooln sites. Further fiel surveys re neee to ientify whether these reltionships hol more roly ross the GWW, long with t on AGB vlues efore n post-timer utting. The fiel survey sites hving the highest AGB our in wooln where there hs een no timer utting, no fire sine 1972 n no mining tivity (tht is, not on grnite-greenstone lithology). 61

32 Aove Groun Biomss (tonnes h -1 ) Eu wooln - no fires post no timer utting 12. Eu wooln - no fires post no timer utting 14. Eu wooln - no fires post no timer utting - MinExpl-greenst. 6. Eu wooln - no fires post no eviene of timer utting - greenst. 13. Eu wooln - no fires post no timer utting - greenst. 15. Eu wooln - no fires post no timer utting 19. Eu wooln - no fires post no timer utting - greenst. 7. Eu wooln - no fires post timer utting pre greenst. 5. Eu wooln - no fires post timer utting pre Eu wooln - urnt no timer utting 21. Eu wooln - urnt no timer utting 1. mllee wooln - no fires post no timer utting - MinExpl-greenst. 9. mllee wooln - no fires post eviene of timer utting - greenst. 18. mllee wooln - no fires post no timer utting 4. mllee wooln - urnt no timer utting 3. mllee wooln - urnt no timer utting - greenst. 10. mllee wooln - urnt timer utting ir greenst. 2. mllee wooln - no fires post timer utting pre greenst. 17. Shruln - no fires post timer utting pre Shruln - no fires post timer utting pre Shruln - no fires post timer utting pre 1938 Figure 3.19 Reltionship etween NVIS MVG vegettion lss, fire history (post ~1970), timerutting history ( ), surfe lithology n AGB of ANU fiel survey sites (numere 1 21). Fire, timer-utting histories n lithology re inferre from the sptil t lyers n fiel oservtion. The lithology is inite only for sites ourring on the grnite-greenstone formtion, inite here y greenst. Our fiel notes reor impts of prospeting (vehile trks, exvtions, n so on) on or lose to mny sites lote on the grnite-greenstone lithology. Although our timer-utting sptil t lyer inite tht Site 6 h een ut over, we oul fin no eviene of this tivity hving ourre. There were lrge trees present n no ut stumps or oppie growth. 62

33 3.3 Extrpoltion of ron stoks from site t In the preeing setion, we investigte the extent of isturne y fire, timer utting n minerl explortion in the GWW n ientifie those fiel survey sites tht h een impte on y these ftors. Vriility in the vilility of wter n nutrients ross the GWW lso ffets folige over n must e tken into ount when extrpolting from fiel survey sites to other lotions Remote sensing of folige over Sptil informtion out the green folige over of wooy evergreen vegettion n e otine through the nlysis of time-series of stellite imgery (Berry n Roerik 2002; Roerik et l. 1999). We erive sptil lyer epiting the frtion of photosynthetilly tive rition interepte y the evergreen vegettion nopy (F E ) from MODIS 16-Dy L3 Glol 250m (MOD13Q1) stellite imgery for the perio from July 2000 to July 2005 (Berry et l. 2007) 9. The first step in the lultion of F E involve the lultion of monthly vlues of the normlise ifferene vegettion inex (NDVI). We then lulte the frtion of photosynthetilly tive rition interepte y the sunlit vegettion nopy (F PAR ) for ll months in the time-series using the following eqution: F PAR * = 1.118NDVI* in whih NDVI* 0.15 n F PAR * = 0 in whih NDVI* <0.15 in whih * inites vlue for speifi month n yer. We ssume tht when NDVI <0.15 the surfe over is re soil or rok. In ny month, F PAR will omprise persistent evergreen omponent (F E ) ue mostly to wooy vegettion (Roerik et l. 1999) n flututing sesonl omponent (F R ) the green-flush following rin tht is ue mostly to hereous vegettion. To seprte these omponents, we followe the generl metho of Donohue et l. (2009), ut pplie 12-month moving minimum winow n 15-month smoothing funtion. We then etermine the 12-month perio from Jnury 2001 to Deemer 2004 for whih F E ws t its mximum vlue n lulte 12-month verge F E. This pproh ws tken in orer to reue the noise from shortterm efolition events. The resulting sptil lyer of verge F E is shown in Figure Figure 3.20 Frtion of photosynthetilly tive rition interepte y the evergreen omponent of the vegettion nopy (F E ). Units: x 100. Ln over types hving low F E inlue slt lkes, rok outrops, griulturl ln n vegettion nopies tht hve een thinne y fire or ln lering. 9 For t soure for MODIS t, see Tle A1 MODIS. 63

34 3.3.2 Reltionship etween folige over n AGB The vrile F E provies informtion out the ensity of green folige over ut not vegettion height, hene iret reltionship to iomss nnot e ssume. Over lrge environmentl grients n wie-rnging vegettion types, this reltionship is resonle ssumption euse the environmentl ftors tht influene nopy over lso influene height n iomss. Our fiel t for the GWW emonstrte, however, the lk of generl reltionship etween F E n iomss: F E of shruln sites rnge from 0.04 to 0.37 n eulypt wooln sites from 0.22 to Thus, there were shruln sites with nopies s ensely green s some wooln sites. Within these two ro vegettion groups n wooln isturne tegories, however, reltionships etween F E n AGB were evient. We foun tht lthough mny of our wooln sites h very similr F E vlues, sites tht h een isture y timer utting or minerl explortion h lower AGB thn unisture wooln sites. This fining inites tht the nopy (omprising leves supporte y smll rnhes) t isture sites hs h suffiient time to reover to the pre-isturne onition, while the mss (n volume) of wooy stems tht support the nopy remin elow the pre-isturne onition. We erive four equtions to preit AGB (t h -1 ) from F E (expresse s perentge) in the GWW (see Figure 3.21). Three of these equtions preit the AGB of eulypt woolns (NVIS MVG 4 n 5) for the three senrios i) no timer utting, no minerl explortion isturne (Eqution 3.8) Eqution AGB = F E ii) minerl explortion, no timer utting (Eqution 3.9) Eqution AGB = F E iii) timer utting (Eqution 3.10) Eqution 3.10 AGB = F E 2 The fourth eqution preits AGB for ll other wooy vegettion lsses of heth, mllee, mrlok n shru (Eqution 3.11). Eqution 3.11 AGB = 0.06F E All of the equtions ove hve een fore through (0, 0) lthough the line of est fit for senrio (i) woul pss through point where F E exees 0 when AGB equls 0 onition where smll wooy vegettion nopy exists with no mss. In the se of senrios (ii) n (iii), the line of est fit is plotte through group of three points. The stter etween points is likely ue to the vrile impts of minerl explortion n timer utting t these sites. These tivities hve ourre over long perio (75 yers in the se of timer utting). Both the time perio sine isturne n the intensity of isturne ffet AGB. It is not possile to evelop more speifi equtions without the vilility of etile informtion esriing these pst isturnes n greter numer of fiel survey sites within eh isturne lss. For our fiel sites, the effet of isturne on iomss ws greter thn the influene of other environmentl vriles. More fiel t woul e require to lirte the effet of other environmentl vriles within isturne lsses. 64

35 Eqution 3.8 Eqution 3.9 Eqution 3.10 Eqution 3.11 Eu. Wooln - long unisture Eu. Wooln - minerls explortion Eu. Wooln - timer utting Other Wooln n shruln AGB (t h -1 ) F E (%) Figure ,000 60,000 Eu. Wooln - long unisture Eu. Wooln - timer utting Eu. Wooln - minerls explortion Other wooln n shruln 50,000 Frequeny 40,000 30,000 20,000 10, F E (%) Figure 3.21 Figures 3.21 Reltionships etween ove-groun iomss (AGB, t h -1 ) n F E for the 21 fiel sites. F E vlues for the fiel sites were extrte from the sptil t lyer shown in Figure Equtions (see Vegettion Remote Sensing setion) re lso plotte. One eqution is suitle for preiting the AGB of ll vegettion strutures other thn eulypt wooln. () The frequeny histogrm shows the istriution of 250 m 250 m gri ells within F E lsses for eh of the AGB senrios. The open rs inite the mximum vlue of F E of fiel sites within the Other wooln n shruln lss. The lk, lue n re rs inite the mximum vlues of F E of the eulypt wooln lsses. 65

36 We teste the pity of Equtions to preit AGB y plotting the preite vlues ginst those erive from mesurements for our 21 fiel survey sites (Figure 3.22). Overll, there ws goo greement etween the tul n preite vlues (AGB preite = 0.99 AGB tul ; R 2 = 0.93) AGB Pre = 0.99 AGB Mes R 2 = 0.93 AGB Pre (t h -1 ) AGB Mes (t h -1 ) Figure 3.22 Comprison of site AGB lulte from fiel t (AGB Mes ) with vlues preite using the Equtions (AGB Pre ). The lines show the line of est fit using liner regression nlysis (lk) n the 1:1 reltionship (re). We note ove tht the rnge of F E for eulypt woolns in our fiel survey ws from 0.22 to The rnge of F E for res mppe s wooln in Figure 3.13, however, vries from 0 (no evergreen vegettion) to 0.5 (Figure 3.23). Those res hving very low vlues of F E re uner lesehol ln tenure. As note ove (Setion 3.2.5), most of the wooln uner lesehol tenure hs een sujete to minerl explortion n/ or timer utting isturnes ounte for y senrios (ii) n (iii). Low vlues of F E n rise if regenertion of wooln hs een suppresse y mngement tions tht might lso hve resulte in further ln lering. We utilise Google Erth n high-resolution eril photogrphs to investigte the ensity of wooln tree over on ln uner lesehol tenure. One suh re is shown in Plte 6. In this exmple, the wooln hs eviently een thinne or lere (in the south-western orner of the photo) for pstorlism. Plte 7 shows further exmple of wooln on ln tht ws formerly uner pstorl lesehol tenure ut is urrently within nture onservtion reserve. The wooln hs een thinne rther thn lere. The rnge in F E for woolns over the GWW (mx. 0.5) exees the F E vlues for the fiel sites (mx. 0.32); hene preitions of AGB will exee the lirtion t for the equtions. For other wooy vegettion types, the rnge in F E t the fiel sites is similr to the full rnge over the GWW. 66

37 Figure 3.23 F E of extnt eulypt wooln in 2008, overlye with ln tenure type (see Tle A1 Ln Tenure). Note the lower vlues of F E for wooln uner lesehol ln tenure. This figure inorportes Austrlin ln tenure (1993) t tht re Copyright Commonwelth of Austrli Plte 6 Aeril photogrph showing region mppe s wooln uner lesehol ln tenure in the GWW, north-est of Klgoorlie. The region hs low F E (F E = 0 in some MODIS pixels) n expnses of re groun re lerly visile. A m is visile in the south-western qurter. Reproue y permission of WA Ln Informtion Authority, CL34/2010 (< 67