245

Trifolium thompsonii Stand Conditions Following a Wildfire Event in the Entiat Mountains of Central Washington

George Scberer1, Richard Everett• and Ben Zamora2

'USDA-Forest Science lab, 113:3 Western Ave., Wenatchee, WA 98801 Tel. (.509) 662-4:JJ.5; Far (.509) 664-2742

2Department of Natural Resource Sciences, Washington StaJe University, Pullman, WA

Abstract. TrifoUum thbmpsonii. (Thompson's clover) is listed as a threatened legume of the Eastern Cascades where wildfire is lhedootinantdisturbanceagenL Previous researchsuggesttdfiresupressionadverselyimpactspopo­lations of !his species. Recent fires in partS of Thompson's clover range indicate that Thompson• s clover responds favorable to fire distwbance. Four areas of Thompson's clover habitat were sampled for vegetacion composicion and structure. Thompson's clover stands we,e evaluated for species density and morphological expression. In areas ofrecentfire, Thompson'scloverindividualsw~signifi­cantly taller (means of range 32-25 cm) and had gn,arer pwnbe:s of flowerbeads (means of range 4 to 1 pee plant) than on sites wilh no =nt fue activity, suggesting a Thompson's clover response to fire caused resource release. This infonnation combined with a species density· ranging 0.6 to 4.8 plants m·2 suggests that Thompson's clover appea,:s to achieve opcimum stand conditions on sires wbtte perodic grass-shrub eliminaling fires occur. Re-establishment of historical fire effeas may be a key to improved management of landscapes in which fire.respon­sive ~-pecies such as Thompson's clover are components.

Introduction

Trifolium thompsollii (Thompson's clover) is one of 12 plant species listed as lhreaiened along the east slope of the Washington Cascades (WNHP 1994). Wilhin ics range it is adominantforb, fl!Slreported by J.W. Thomp­son in 1933 on the slopes of Swakane canyoo at the edge of Pinus ponderosa forests and into the Artemisia tridenllJla shrub sieppe adjacent to the western shores of the Columbia River (Hitchcock et. -al. 1969). Work by Canfield (1977) and Tiedemann ct.al (1977) expanded .knowledge of this species and docwnented 10 additional stands, the largest located within the 80 ha Thompson's Clover Research Natural Area. Survey wcrl< reported by Gamon and Sprague (1988) further extended the Chelan county occurrences of Tri[olium thompsonii IO canyons near lheEntiatRlverandadded information of a stand east

of the Columbia River, about 6 km southeast of the Thompson's Clover Research Narura! Area.

These new sightings snggesi even wider range poten­tials and pezhaps a different ecological amplitude than pre­viously described {Kennison and Taylor 1919). Toe pur­pose of this study was to clarify the habitat characieciscics of Tri[olium thompsollii and seek evidence which might indicaie the successiooal role of this species among the different plant communities where it is found. 1be ob­jectives of this.research were: 1. to docwnent the vegeia­tive communities and sue factors associated with Trifo­li.um thompsonii.; and 2. to compare moqihological fea­tures and plant density of Tri[olium thbmpsonii popula­tions among these vegetative communities.

Methods

Study Area

Climate of the area is typical of eastem flank of the Cascade range. Swnmers aie warm and <Icy and winlel's are modmtly cold. Precipitation occurs mostly as snow between November and April. The nearest climaie Sia· cion at Wenatchee, WA., has a mean January temperature of -3 °C and a mean Joly temperature of 22 °C. Annual precipitation is about 220 mm. This species occurs at el­evations of 345 to 1146 m (Canfield 1977). Using the normal lapse rate of 2 "C per 300 M lhe mean temperatw:e on the study sites would be -8 °C for January and 17 °C for July. At these elevations winter snowpaclc may be ex­pected to remain for op to 60 days longer in the spring; delaying bud emergence for Trifolium · thbmpsonii and other species (Gamon and Sprague 1988). . _

Discurbance History

In general, fire bas bcai an historical ~ agent in this region and was evident on a portion of the Trifo­lium thompsonii study area Figure 1 shows lhe ~ of wildfite within the known ra:ngeofTrifolium lhcmpsollll

246 Scherer, G., Everett, R. and Zamora, B.

since 1930 (WNF 1990, Canfield 1977). The most recent tireoccmrenc:ewasJuly 1983. I.andscapeswithln thisaiea were grazed by domestic sheep and caUle until 1965 (Tiedemannet.al.1977). NolmownfireshaveOCCUI7Cdon the Douglas =mty study sites but !he area is presently under seasonal grazing impact by domestic caale and portions of the area are cultivated for wheat crops (Gamon and Sprague 1988, E. Gotzwile. 1993, pm. comm.).

F~ld Sampling

Sampling sites were selected by reviewing maps of known Trifolium rhompsonii locations from the Wenatchee National Forest Thompson's Clove. Manage­ment Guideline (Gamon and Sprague 1988). Criteria for sample site selection was: The appearance of distinct patches of Trifoiium thompsonii in the overall mosaic of vegeiation coverage and patch size greater than 50 M di­ameter.

Vegetali.ve sampling was done by systematically plac­ing a 50cm x 20cm sampling frame in a4 X 5,20 frame matrix at three positions (0, 30, SOM) along SO mew tape transect running trough Trijolium thompsonii patches. The polygonal cover class of each species eocoonreted in the P,lot frame was recorded (Daubenmire 1959). Taxonomic nomenclature follows Hitchcoclc and Cronquist (1973). At the same time, soil surface condition was recorded at 10 _pointS on Ille plot frame as either litter, bate ground, plant basal area, crypwgamic crust, wood pieces, or rocks. At each sampled community !he density of Trifoliwn rhompsor.ii was measured by counting individuals within · Len ll'vf plot frames in !he same 4 X 5 vegetation sam­pling matrix. Frol)l each sampled community 10 ID 30 Trifolium Ihompsonii individuals were randomly selected and measured for height, leaf number, shoot number, flowerhead number and root crown diameter.

At each sampled community the following sire fac. tors we.e recorded: elevation, slope, aspect, slope posi­tion, and a microterrain feature (flat, convex, concave). The soil depth was measured and a soil sample of the up­per 20cm was collected at each sample matrix and ana­lyzed for bulk density and% soil moisture. Toe soils dala we.e compared and suppleme.nied wilh SCS Soil survey laboraLOry information available on the Study area (Beiler 1969, 1978).

Anaiysis

Mean coverage and frequency values were calculat£d for all species in each sample unit and used IO consuuct a species list of cover and a table of plant communities with species cove. and coosiancy (Table I). Dendograms pro­duced by cluster analysis (Wilkinson 1992) we.e used IO identify relationships among Trifolium 1/rompsonii sam­pling clusters based on species composition and cover. Site fac!Or data on soil moisture, surface cover and ter-

rain fearures were tabulated for comparison by plant com­munity (Table 2). DataonclensityofTrifoliwnrhompsonii and measwemenrs of morphoiogical f~ we.e tabu­lated by plant community and analysis of variance was used LO 1CSt for differences among the community types..

Resulis

Trifolium thompsonii occurred in three broad yegeta­tive settings: forest patches, SLeppe-like parklands, and semiarid shrub Stq)pe. The forest patches consisted of a multilayered community of irees, shrubs, grasses and forbs dominated by Pseudotsuga menziesii occurring on mesic IOpographic· seuings such as gullies and northerly slopes. Steppe paddands were grass dominated landscapes with occasional mature Pi.nus ponderosa . individuals. The shrub steppe was the extensive drought LOlerant tall sbrub­bunchgrass (,Artemisia trkkn!alal Agropyron spiCIJlun:,) vegelative associations typical of the central Columbia basin.

Wilhin these lhree settings four plant communities were identified using cluster analysis of species compo­sition and cover values wilhin stands. These four com­munity types probably represent various successional stages of four plant associations as descn"bed by Franklin .and Dymess (1973) and Daubenmire (1970). They are Psew:w1suga menz~sii/Calmagrosris rubescens (PSMEI CARU) of the forest setting, Pillus ponduosa/Agropy­ro" spicatum (PIPO/AGSP) and Artemisia rriderual.tJ­vaseyana/Agropyron spicaium (ARV A/AGSP) of I.be steppe setting and Artemis/a 1ridenta1a-1riden1ataJA,gr<>­pyron spicarum (ARTR/ AGSP) of !he shrub Sleppe zone. Physiognomic differences between the community types we.e distinctive and easily recognizable. These four com­munity types are descn"bed below wilh refe.ence LO veg­etativecom_position and site fac!Or dala in Tables l and 2.

Planr Commwliry Descriptions

The PSME/CARU community occurred on noi:th fac­iog steep slopes (40-60 %) at mid LO upper elevations (> HMO m) in Chelan county. This was the weucst and coolest environment wilh estimated annual precipitation ranging 508 LO 1016 mm and estimated annual iempera­ture ofS-6 "C. Pseudo1suga menziesii was the overstory -dominant with Pinus ponderosa a codom.inant in some .sample units. Spiraea betulifolia and Pensremon jnuicosus were prominent shrub layer componentS. 1be grass, Calmagrosris rubescens was a significant CCHlomi­nant of this communjty type. Balsamorhiza sagi1ra1a, AcMllia millefo/ium and Trijoliwn 1/wmpsonii were the most prominent forb species. Other imponant hcrbacrous =ics were Festuca idalroensis, JJrornus tectorwn, Poa s~cundi, Collensia parvif/ora, Cryprhantha a/finis, Epi/obium paniculara. In Ibis community fire recently

• Trifolium tltompsclfii Stand Conditions FoUowing Wildfire • 247

Tablcl.Conmooymdmean~crofspecicoinfourTrifo/iJJmtliomp,,,niicommunityiypes.Tr-=%coverlessthanO.l. N=nnmbcr of sample unit&. Cot,,, """""""Y· = % canopy covcr.

PSME P.ceo ARTR ARVA N=Z N=7 N"9 N=J Ca, Cov Ca, Cov Ca, Cov Con Cov ,... PWU p<>ltduMO 100 29.0 p p 0 0.0 0 0.0 Puw:louwga nvlfZiaii 100 32.0 0 0 0 • 0.0 0 0.0 Shrub Arte,ni,io tr;,J,AJaJa-tr 0 0.0 14 Tr 77 8.S 0 o.o Nt~t4 trid,usl4J4-\t2 0 0.0 0 0.0 0 0.0 67 1.2 ,1,..,,,., .. tripartiJa 0 0.0 0 0.0 11 1.0 67 14.0 ClllNldut: wlutima so 6.0 0 0.0 . 0 o.o 0 0.0 C~,a vueidij!ona 0 0.0 0 0.0 II 0.2 0 0.0 Erigero1t luu.ari, 0 0.0 14 0.7 0 0.0 0 0.0 ~-1,uad.e;,iJu 0 0.0 43 2.2 ss 2.6 100 8.S fjiogo,wm ,uy~""" 0 0.0 14 0.4 0 .o.o 0 0.0 Erit>ga,aun .ttric:bM 0 0.0 0 0.0 22 0.2 0 0.0 1/aplcpopp,u ,u,.cphyl/,u 0 0.0 14 Tr ss 1.2 0 0.0 P u,.sUlf'IOn /n,tlic.onu 100 ll.S 28 o.s 0 0.0 0 0.0 P/Joz io,,gifolio 100 1.8 57 1.7 100 13 34 0.8 Pwhhi.o lrkkn1a1a 0 0.0 0 0.0 66 4.9 0 0.0 Riba CU~wrt so 0.4 0 0.0 22 0.6 0 0 .0 Spi,,ua bd"1i[Q/;., 100 12.7 0 0.0 0 0.0 0 0.0 s~an- 50 1.6 14 0.4 0 0.0 0 0 .0 Vacc.ini.ian txtiducla.k 50 2.3 0 0.0 0 0.0 0 0.0 Crus Agropyr<M aisJ.aJJutt 0 0.0 0 0.0 11 Tr 0 o.o Agrof11'"",,,._.,. so 0.1 JOO 26.4 100 44.1 100- 26.7 s,,,,.., oc,;;J,"'4/it 0 0.0 0 0.0 II Tr 0 0.0 8rt:/IIPUI.J kt:IDrlll,t 100 0.4 57 2.0 88 l.4 67 ·- 4.4 C4JmogrMl'°1 rwbua111 100 37.0 0 0.0 0 0.0 0 0.0 Futwc.a~ 100 3.7 100 7.S 88 1-U 34 4.2 F utMCA rllMwa 0 0.0 28 Tr 0 0.0 0 0.0 Koe.luia crutalQ so 0.1 71 2.3 0 0.0 100 4.3 Mdicabuibosa 50 0.0 14 Tr 33 1.3 34 Tr 1'"'1 s.eetW:kJ JOO 1.2 85 4.1 100 7.1 100 s.o Fori> A,:hi/1,a ""11,fo/"- 100 7.1 100 7.3 n 2.0 JOO 2.4 Agouru irondif/oro 0 0.0 57 l.S 88 6.0 100 1.8 Alliwn ocwninat.wn. 0 o.o 0 0.0 n 10.7 0 o.o Mgdi.co argWIO 0 o.o 14 Tr 0 0.0 0 0.0 NtltNW'id 'l'IM«I 0 0.0 ss 11.6 0 0.0 34 2.5 A:rra.gtJlvs pw-slaiJ 0 0.0 43 1.3 () 0.0 0 o.o Balsam,,J,J,i.;a 30gittat.o. 100 9.9 100 11.7 JOO 17.6 67 3.8 Brodi.tua dolf8WU 0 0.0 0 0.0 II Tr 34 0.3 Casulkja tw~u,u so 0.1 28 o.s 22 0.l 0 0.0 ColJWi.o porviflora 100 o.s 100 3.l 66 1.0 100 0.9 Crypt.aNlta offini.s 100 0.9 8S 1.S 22 0.0 61 0.8 Epuoi,ium pa,,iCMlaJwm 100 1.6 100 3.2 44 ~ 100 1.2 GtWN fTijlOrlJ1'1'1 50 1.6 28 l .3 0 0.0 34 0.9 Hclian'1ua ptliol.ori, 0 0.0 28 2.4 0 0.0 0 0.0 Hiuaciwn albifwnmt so 0.8 57 1.S II Tr 34 0.9 LoctMCO suriola. 0 o.o 14 0.4 11 0.3 0 0.0 T.,;,,a,,Ji,,a linij/onu 0 0.0 14 Tr 0 0.0 0 0.0 l.illtc,p,,,...,.. n,d,roJ, 0 0.0 0 0.0 II l'r 0 0.0 lDmaliwn twdic.aJe so 0.4 57 1.6 n 3.S 0 0.0 l,,,pinia i<w;opkyll"' 0 0.0 0 0.0 44 1.3 0 0.0 u,pvu,s :,ricews 0 o.o 71 1.4 66 2.8 34 Tr Mic:rostuis graciJ.is 100 0.4 .57 0.2 n l.6 100 l.O Polll1VJIU1,1m, micraNlwm 0 0.0 57 0.2 100 0.6 34 0.2 . Sisy,r.brum alJisurtJUn so 0.4 0 0.0 22 0.0 61 1.2 T rifoliwn 1Mmps.on.ii 100 6.S 100 1/J,1 100 21.2 100 ll.6

Scherer, G., Evereu, R. and Zamora, B.

Table l. Rm,g-,of values fouiie f:iaors in .Usamplc units of fc1lr , ........... :iy i:ypes ,woci,w,d with oCTrifoii,,m tmmpsonu.

PSME PIPO ARTR Ml.VA

Site V.vibl,. EIO'llllioo (m} 1115-1060 810-399 890-1096 816-11171 Slq,e('il\lv 40-50 30-60 15-,l'O 30-40 A,pea( 3(,() 360-235 045-270 3(,().235 P=;ptt (mm) 508-1016 ~ ~ 31)4-406 Temp(C) 5-6 8-10 8 • 10 8· 10

Soil Surface liuu(%) 44-63 40-60 19-69 37-$ Ba,c Grouod (%) II 8-34 13-SS 8-40 Rocle(%) I 0 0.26 0 Wood(%) 4-11 I 0.10 0 Baul Arca (%) 18 11-29 7.z, 71.71 Ctyplogaau ('ii\) 1-4 1-4. 5 0.1.S

Solllfllllt

~ 35-48 25.55 24-75 48-56 . (31«)

Available W~ 1.22-1.40 1.11).1.30 1.05-1.30 1.21-1.29

(an-an) 0.3~.43 0.43--0.50 0.15--0Sl 0.30--0.35

~anic Manet 6.5 6.4 7.1 6.4

%) 2.5. 4.5 2.8-6.4 1.5-4.5 2.6-3.0 CEC(meq) 10-20 9.15 8·18 8.4 -12.0

Cea= 10.40 10.20 5-45 10.20 Fr>gmcnu (%) S,md($) 7Ul3 65-'9 33,,59 67- 68 Silt ('ii\) 15-21 19-31 36-53 25-32 C.y('l\) 1-3 1-4 Z. 13 1- 6

destroyed lhe crown foliage of nearly all Pseud01suga menziesii and Pinus ponderosa.

Thedominantvegciation in PIPO/ AGSPwasanAgro­[7Jron spica1um- F es1uca idahoensis grass understory (Table 1). Pim,s ponderosa was present as solitary ma­ture individuals in mesic topographic settings wiill cover rarely exceeding 2 % on ille entire study area of the Re­~h NaruraJ. Arca. Only a few low shrubs, Eriogonum heracleoides,Phlox longifolia, Hap/opappus stenophyflus, were present wiih less than 5% overall cover. No tall shrubs were recorded. The most prominent species in the forb component were A1Ue1111tJTia rosea, Balsomorhiza sagittata, Trifolium thompsonii and A chi/lea millefolium. The annuals Coiiensui parviflora, Epilobium paniculaLa, Crypthanlha affinis were nearly ubiquitous in the samples but with small cover values. Fire disturbance was gener­ally in tbe grass shrub understory with no tree crown dam­age.

The AAVA/ AGSP was located on tbe Thompson's Clove:r Research Nalllial Area (Figure !) and was adja­cent to PIPO/ AGSP. Individuals of Arlemisia 1rid<!illata­vasayana and Artemisia triparti1a represented ille over­Story cover of this community cypc. The grass species Agropyron spicatum, Poa sec;uuta, Festuca idahoensis rangjngin mean cover from4-27 % were vegetativedomi­nant.s. Trifolium 1hompsonii, Balsanwrhiza sagittata,

Anwuuzrui ro~a, Acheilia millefoliJlln were tbe major forb components. Annoals were net prominet1t in this community, except for a Bromus ~ctorum coverage of 4 %, tbe highest of this study. Charred buochgrass clumps and shrub skelerons indicated recent (1988) fire disiur­bance.

In tbe ARTR/ AGSP community cype !he overstory is dominated by the tall shrubs Anemisia tridentata­lrideillata and Purshia lridemata, wiih mean coverage valoes of 9% and 5%. The herbaceous dominants were represenied by Agropyron spicatwn, F estuca idalwensis, and Poa secunda with op to 44% mean cover. 11lis com­munity was rich in forbs with nearly 20 recorded The most important were Trifolium thompsonii,Balsamorhirza sagiltala, AJJium acuminallun, Agoseris grandiflora, Loma1ium nudicale, Achellui millefolium.. and Lupi.mu sericeus. There was no evidence of recent fire in this com­municy. However, some forbs and bunch grasses appeared to have been grazed.

Community Structure

Table 3 shows the SlruCIW'C of the four communities by life form. · Grasses and forbs were prominent in ille communities of PIPO, ARTR, and ARV A with forbs nearly equal in presence but more variable in cover and :species composition among ille groups. Tree cover was clearly distinctive in th1e PSME/ CARU community but almost of no importance in !he ihreeoihers. Noteworthy, was a low value of shrub cover but high value of forb cover in PTPO/ AGSP, suggesting recent disturbance in lhose layers; lilcely by fires, since rree boles were charred but crown foliage was most! y unharmed. The structure

data shows ·a decline in shrub and tree components and increase in grass and forb components from the mesie forests to the arid shrub s teppe.

Plant density- Trifolium thompsonii

Density of Trifolium thompsonii ranged from O to 16 plants-m-' on 258 plotS from within ille four community types. There were significant differences in the mean density of Trifolium thompsonii among all ille commu­nity types. Generally. Trifolium thompsonii densities were greater in the steppe-like settings than those in !he PSME setting and specifically they were highest in ille ARTR shrub- Steppe of Douglas County where individual plot frame samples contained as many as 16 plants m•. By contras!, plot samples in the forested (PSME) had high vegetative coverage in trees. shallow soil depths, and modest available water (Table 2 and Figure 2). These forest like features suggest quite different site conditions for Trifo/ium thompsonii occupancy. which may be lhe cause of its low density value of 0.6 plants meter.

Chelan County

";;~ec River ~

t Nort.h

- Trifolium thompsonii Stand Conditions Following Wildf:u:e -

0

0

Columbia ltive;r

Douglas County

Douglas County Sites , -I 2 3

2 3 4 mJ

• Pct$Cnt sund.G illCQfsm !~ccB

~ Fi.ra during 19S0'.2.

El Fires during 195()1s

III] F'il'e.$ during t960"s

[D Fitts during 1970's

~ Fires durinc 19SO's

249

Fl2ure l. Present l:nowns!Jlnd.s.o(TrifoUum thompsonii in Ch.clan and Douglas countics_W ashingron wuhknownareas of fire disturbance in 'l"rijolium thompsonii range since 1930 (ad.aptecl !rom Canfield 1977 and WcruuchccNational Forest 1990).

250 Scherer, G., Eve."'ett, R. and Zunora, B.

Table 3. Proportion cf a,va by lifcfomtin follfcommwiity types 3SS<lCiated wilh Trifolium thompsonii.

PSME PIPO ARTR ARVA

fo<b(%) 17.7 61.4 l8.3 40.8 Grus(%) 2S.J 34.4 46.0 43.S Sluub ( %) 2.1.3 42 ZS.5 14.5 Tree(~) 35.9 I 0 0 Tow 100 JOO 99.8 98.8

No<cs: Canmunitytype~cman,:PSME,,1',"""""1,gamt.,ai,uiiJ Ca.lmogrO/ffl.r ni>ucur.1":PIPOa POnduwa pint/ Agropyron spicahtltr; AR~ Art~ia tr~ ;tgropyrt>ft rpicatllf'lf; ARVA: ArtbrWii:: rridelt14.J.a-vo.ur;a,rd Agropyr01t spicm""'-

Morphological characteristics- Trifo/;um thompsonii

Approxin,ately 350 •p,cimens ofTrijolium thompsollii were measured for ihe five morphological features of height. leaf number, shoot number, tlowcrhead number, and root crown diameter. Because of plant denshy dit· ferences there were varying sample sizes from each of the community cypes. Based on mean values, a "typical" Trifolium thompsonii specimen was 27.6 cm in height. a three shoot plant with.five leaflet per leaf, two tlowerheads and root diameter of four mm. Data presented in Figare 2 show relationships of Trifolium thompsonii height, TOOt diameter and flow<:l"head numbers in each community aligned in order by high fire iniensity ro low fite inlCD­sity. This information suggests Iha! T rifolium thompsonii achieves significant plant height and flowering potential on sites affected by wildfire, such as PSME and PIPO. Flowerhead numbers on PSME were significantly greater

(p < 0 .05) than those of the other communities. However, as the confiden~ interval bars suggest. there were no sig­nificam differences in the height or root diameter data among the other communities. We note the trend com­pariS()n of 1rifolium tlrompsonii plant density rising 10 ARTR-no fire side and the height/ flowerhead counts me:isures declining in the same direction, but presently offer no explanation for this pauem.

Discussion

Steep rocky slopes with deep valleys and open pine­shrub-g,ass communities are the most distinctive land­scape features of the semi-arid foothills on each side of the Columbia River north of Wenatchee, WA. It is an ecotonal area where dry climate spe,:ies of lhe PiluJ.s pon­derosa /Agropyron spicatum association loses landscape influence to even more drought tolerant Artemisia trident<Ua /Agropyron spica1um communities (Franklin and Dyrness 1973). Topogr,y,hic intluCllCes here create intergrade variations in the vegetative mosaic such as Pinus ponderosa communitie., giving way ro Pseudotsuga menziesii associations in north facing ravines and at eJ-

evatiO/JS of 800M er more (Daubenmire and Daubenmire 1968). This landscape is pan of a larger eastside area Iha! is subjected regularly 10 ligh1nmg ignited wildfire disturbances (Agee 1992). The results of this were ob­served in lhe physiognomy of the -vegetative mosaic as sc.cin:hed trunks of Pinus ponderosa, fire-lcilled pole size Pseudotsuga menziuii, burned skeletons of Artemisia triden101a (or large areas of none at all), and charred clumps of bunch grasses of Agropyron spica/um and F estuca idahoensis. Amidst this regular!y disturbed land­scape condition, Trijolium t/rompsomi appears as a domi­nam forb but in discontinuous patches of up t0 70 ha in si2e over some 80 km2 of Ibis area. The patchiness ofhabi­talS along lhis ecotOoe is particularly characteristic of the part of its range west of the Columbia River in Chelan county.

Trifolium thompsonii easrof Columbia River in Dou­glas county occurred in a more unifonn physiognomy of Artemisia trideJtJala/ Agropyron spic<Uum zonal associa­tion with an area of Pinus ponderosa forest t0 the north on the summit ridge of Badger mountain (Daubenmire 1970). On a Jong tenn basis, this area has ~rienced similar landscape disturbances as the west side but wilh less regularity in recent years, probably due to greater agriculwral actiyity such as fruit on:hards, wheat farm. ing-and caule ranching (E. Gutzwiller 1993, per.; comm). Shrub-grass vegetation was typical cover in all but thinly soiled scab patches on knQlls, ridges, basalt cliffs and ta­lus screes. Trijolium 1hompsonii appears in this land­scape in patches up ro about 3 ha on modest slopes and terraces of all aspects but not necessarily as a dominant forb evenly dispersed among the Artemis/a 1ridentata I Agropyron spicatum communities.

The presemsuccessional condition of lhe Pinus pon­derosa /Agropyron spicatum association on the Thompson's Clover Research Natural AlC3 in Chelan rounty can be described as seral because of clear evidence of a recent (1988) wild.fire which involved much of this landscape (WNFl990). The species composition of PIPO of th.is study (Table I), however, suggests !hat this com­munity may still be near its most successionally advanced :slage. The COmJ)Osition nearly ma!Ches Ille established PinJJS ponderosa /Agropyron spicatum habitat type docu­mented by Daubenmire and Daubenmire (1968). There may be some argument on the authenticity of the PIPO community of th.is study because oi the presence of F estuca idar.oensis, a grass spedes not normally found in the Pinus ponderosa /Agropyron spit;a1um habitat type, Higher elevations and slightly more mesic conditions on 11ortherly aspects of PIPO probably explains this. Veg­etative evidence of fire disturbance is revealed by the per­sistent presence of Antennaria rosea and Epilobium p<Uliculatum and significant cover values ofsuch annuals as Crypthantha afflllis, He/ionthu.s petiolaris,Microsteris graci//is,andBromus 1ec1orum (Wright 19~). Pr=ce of these species in the understory are clues of an early

- Trijolium thompsollii Stand Conditions Following Wildfire - 251

Density

-·· PSME ARVA PIPO ARTR Hi Med Lo No I Fire Fite. fire fire

'---- --·- --------- ___ _j

' -I ~ Root Diameter E § " • T

"' E

&l I •

• i. ~ <

PSME ARVA PIPO ARTR HJ Med Lo No Fire Fire ffre n,.

.------- - -·-- -

I ' -·

PSME Hi Fire

ARVA Med Fire

Height

·--·--- - ··-·- -·-·. ~----4.5 1

' • 1 Flowerheads

c,, j i 3

ARTR No r.ro

" 2$

_D .. DJ1 ! ?

" : 1.6 :,:

' o,,

0

PSME ARVA PIPO ARTR HI Med Lo N<> Fire! Fire flre fire

C-ommunity type abbreviations are: PSME=- PSeGJdotsuga menziesii/ Calma-grostis rubesoens PIPO= Por.derosa pine/Agropyron spic-aium; ARTR=Artemisla bidentata/ Agropyron splcat>, m ARVA= Attemisia tridentata/ Agropyroc, spicatum

~re 2. Trifoliwn thompsonii plant density and morphological fearures in four community types. Vertical b.,.. = 95% confidence znterVals.

sc.raJ stage in this landscape. If this community is seral, then the vigor of Trifoliwn thompsonli individuals sampled here may support the proposition that this species is a se­ral opportunist in response to disturbance.

It was in the PSME community, however, that Trifo­liwn thompsollii appeared to be most opportunistic. lndi­vidaals here were tallest and had the greatest floral-~ potential. This PSME community resembled the Pseudolsuga menziesii I Symphoricarpus a/bus habitat type as descnoed by Daubenmire and Daubenmire (1968), largely because of the presence of the shrubs Spirea. /Jeru/ifolia, Ceanorhus velUD/lus and Vaccinium'spp. This comparison, though. is confounded by the presence of Caimagrostis rubesce.ns tha! was sampled on slightly drier pans (ridges) of this community. The shrubs were located in slightly wetter gully locations. It is important to note that the Pseudotsuga menziesii was burned in the tree

crowns by the 1988 wildfire. This lacl::oftreecrown per­mitted considerable light penetration 10 the understory and surface. With .these openings and reduced competition, light tolerant heliophytes such as Trifolium thompsonii may have opportunity to develop particularly vigorous indh'iduals. Of course, this strategy depends on a taxoo being present in lhe community, either in the soil seed reserves or in a reduced vegetative form prior to the event of disturbance. The low density value of T~ifo/ium thompsonii recorded in the PSME community suggestS !hat its presence was probably limited in the predisrurbance phase.

The ARTR community of this study appeared to have !he least disturbance, certainly from fires. A nearby rancher grazes cattle in the area_ Grazing was observed on the bunch grasses and on Trifoliwn rhompsonii. There was no indication of receatfires, indeed with such exten-

252 Schete.r, G., Everett, R. and Zamora, B.

sive coverage of fire in10lcrant Artemisia 1ridenta1a­triden1ata, fire probably has not been a disturbance fac­tor for decades (Table l). This community is thoroughly described in Daubenmire ( 1970) as a major climax asso­ciation of the Columbia Basin area. Trifolwn thompS()nii is present here in substantial densities and coverages (Table l). The individuals, however, were significantly shorter than those of lhe other community types of this study. Because lhis well established ARTR community was so successionally advanced, we argue that TrifoliJUn thompsonii is also an impo~ climax community forb. Since lhis study did not aceuracely identify the severity of previous grazing on this landscape, Trifolium thompsonii probably should be considered a seral forb in this community until more information is available on its ecological life bisiory at this new location east of the Co­lumbia River, particularly wilh the broad influence of agricultural disturbances.

ln summary, several communities have been desoibed where Trifolium thompsonii occurred. They are well known plant associations in cenain stages of lheir suc­cessional develcpmenL Trifolium thompsonii showed varying population densities and plant vigor expressions in each of these associations. These variations in expres­sion can be explained, in pan, by the different site poten­tials and degree of resource capture by Trifolium thorr.psonii in each of lhese communities. The various site potentials are partly explained by the type, severity and recency of disturbances these communities have ex­perienced: from !he inlCnsely burned trees of !he PSME community, where this species is tall and vigorous but very low in density IO the only slightly disturbed commu­nity of ARTR where it is significantly shorter bot greater in population density.

References

Agee, I.K. 1993. Fire Ecology of <he Pacific Northwest Island Press, Covelo, CA. 493 pages.

Beiler, V.E. 1969. Soil Survey of Chelm Are3, Washingion. USDA Soil Conservation Sc,vice, Washington D.C.

Beiler, V.E. 1978. Soil Suivey of Douglas County. Washington. USDA Soil Consavation Service. Washington O.C.

Bmuing, S. C. 1984. Fite in the sagebrush-grass ecosystem: successional changes. IN: Sanders, IC. and J. Ouxham 1985. Range!Jmd Fire Effects. a sympooiwn. Nov 1984. USOI. Buzeao of Land Managcmcn~ Bo~ ID.

Canfield, I.E. 1977. The ecological life hisuny of Trifou"'11 thompsonii Monon. wiJh reference IO its resuicted distribu­tion. M.S. Thesis, University ofWashingion. 153 pages.

Daubcmrure. R. 1959. Canopy coverage method of vegetation analysis. Northwest Science. 33:43-64.

Daubenmire, R. and J. Daubenmire. 1968. Fore.st Vegetation of Eastern Washington ·and NorthcmJdaho. Washington Agric. Experiment Station. WSU. Eim:ruion Bulletin 60.

Daubenmire, R. 1970. Steppe Vegeu,tion of Washington. WSU Extension Belletin No. 70, Pullman. WA.

Franklin.. I.P. and C.T. [)ymcso. 1973. Naruru VegeLUion of Oregon· and Washington. OSU Press, Cor,allis, OR. 452 pages.

Gamon. J. andN. Sprague.1988. Babitatmanag<menrguidelincs for Trifoluun t/u:mpsoniiMo-roon on die Wena1ehee National Fo=t. Washington Namn! Hctiu,ge Program. Olympia. WA. Unpublished report ro Wenatchee National Forest Su­p,;visor

Gutzwiler. E. 1993. P=nal COIIlllllllUc.tion with Ed Gutzwikr. 3531 Badger ML Rd.. E. Weruuehcc, WA. 98802. Fa:,m.s privao: land on Trifolium thcmpsonii sites.

HilChcoc:k. C.l.. and A. Cronquist. 1973 . ..flora of the Pacific Northwest University of Washing10n Press. Seartle, WA. 73-0 pages.

Kenru.on, I. and R. I. Taylor. 1979. Status Report on Trifolium . lhompsonii. File document. USFWS, Pon!and, OR.

National WeaJher Service (NWS). <nimatc dau, on ftle at W enaichce Office, Weruuchee.

Tiedcmann;A.R,J.Oj<rtson,andP.McColky.Im.Toompson's c!O"Ver Rcscan:bNIIUralArea. Supplemen1No. S. IN: Fr";'1klin, I.F. F.Hal.l, C.T.Dymess ond C. Maser. 1972. Feda-al Re­Wllcll N3lliral Areas in Oregon and Wasbinglotl. A guide­bool: for Scientists .,,d Educ.aJors. USDA Fores, Serv~­PNW. 498 pages.

Wen=bceNationaJ Fores, (WNF). 1990. Rcco«!s of fire evenu cas""'1 Entiat mountlhu: On fil~ at Entiat Ra.,g..- District. Entw. WA.

Wilkinson, L and M. Hill. 1992. Sysw. Slatistics, Ver. 5.2. Evmsion, Ill. 724 pages.

WNHP. 1994. Thr03lened. End.1.igettd and Sensitive Vascul2r Plants of Washington. Washington Natural Heri13&c Pro­gram.

' ;

Proceedings courtesy of:

& Washington Foundation for the Environment

Organized by:

TIONAL ASSOCUQ'ION OF

lANDRRE

Proceedings: First Conference on

Fire Effects on Rare and Endangered Species and Habitats

Conference organizer:

Co-sponsors:

Coeur d'Alene, Idaho November, 1995

Dr. Jason M. Greenlee Editor

.-.,;;;;;;,...TIIINAI. ASSOCllmON OF

LA.NORRI!

International Association of Wildland Fire Wildlife Forever

The Nature Conservancy U.S. Fish and Wildlife Service, NIFC Abundant Life Seed Foundation Montana Prescribed Fire Services, Inc. Environmental & Botanical Consultants USDA Forest Service (Reg-6)

Washington Foundation for the Environment Dr. Jon Keeley

· Colorado Natural Heritage Program

Proceedings courtesy of:

-- · -:._~~-,. ...:. ... ---~. . .. ";:-.,;.;.-

Wildlife • Forever

& Washington Foundation for the Environment -Dedicated to preserving and enhancing our state's environmental heritage by supporting educational and Innovative projects