william j.t. ellyson and stephen c. sillett - save the redwoods league

Epiphyte Communities on Sitka Spruce in an Old-Growth Redwood Forest

William J.T. Ellyson and Stephen C. Sillett

Recommended Citation: Ellyson, William J.T. and Stephen C. Sillett. 2003. Epiphyte communities on Sitka spruce in an old-growth redwood forest. Bryology and Lichenology 106 (2): 197-211.

This report was produced with funding from Save-the-Redwoods League. The views and findings contained within are those of the researchers and not necessarily those of Save-the-Redwoods League. Copyright is owned by Humboldt State University. Save-the-Redwoods League has been granted the right to reproduce, translate and otherwise use any and all of the report for non-commercial, educational, and research purposes. Through the license terms of this work you are free to copy, distribute and display the work under the following conditions: Attribution: you must attribute the work to William J.T. Ellyson and Stephen C. Sillett, Save-the-Redwoods League, and Humboldt State University. Noncommercial: You may not use this work for commercial purposes. For any reuse or distribution, you must include the above license terms of this work. © Humboldt State University. Permission required for uses with commercial, non-educational, or non-editorial purposes.

1

0007-2745/03/$1.65/0

THE BRYOLOGISTA JOURNAL OF BRYOLOGY AND LICHENOLOGY

VOLUME 106 NUMBER 2SUMMER 2003

The Bryologist 106(2), pp. 197 211Copyright q 2003 by the American Bryological and Lichenological Society, Inc.

Epiphyte Communities on Sitka Spruce in an Old-Growth Redwood Forest

WILLIAM J. T. ELLYSON AND STEPHEN C. SILLETT

Department of Biological Sciences, Humboldt State University, Arcata, California 95521, U.S.A.

Abstract. Using rope techniques for access, we surveyed epiphytes on five Sitka spruce treesup to 92 m tall in an old-growth redwood forest. We quantified epiphyte diversity by sampling5% of each tree’s surface area of axes (branches .5 cm diameter) and branchlets (branches ,5cm diameter, including foliage). Epiphyte communities included 57 macrolichen, 15 crustose li-chen, 17 bryophyte, and two fern species. The five most abundant species—Isothecium myosu-roides, Polypodium scouleri, Polypodium glycyrrhiza, Lobaria pulmonaria, and Frullania nisqual-lensis—contributed 42.1, 13.3, 8.4, 6.7, and 4.7% of the total epiphyte biomass, respectively. Therewas an average of 36.2 kg of bryophytes, 9.9 kg of lichens, 12.7 kg of ferns, and 131 kg ofassociated dead organic matter per tree. Axes supported 83% of the biomass and 98% of thedead organic matter. At the whole-tree level, bryophyte biomass was 11.3 times higher and lichenbiomass was 2.5 times lower on axes than branchlets. Ferns were restricted to axes. Ordinationanalysis revealed one dominant gradient in epiphyte composition that was positively correlatedwith height and lichen diversity, and negatively correlated with bryophyte diversity. Chlorolichensdominated the exposed portion of the gradient with equivalent amounts of cyanolichens and bryo-phytes. Mosses dominated the intermediate portion of the gradient with equivalent amounts ofliverworts, cyanolichens, and chlorolichens. There was very little lichen cover in the shelteredportion of the gradient, which was dominated by bryophytes. Extensive bryophyte mats with largequantities of dead organic matter promote biological diversity on Sitka spruce in redwood forestcanopies by storing water and serving as habitats for desiccation-sensitive organisms.

The west coast of North America is home tosome of the world’s tallest and most massive for-ests, including at least 20 conifer species exceeding60 m in height (Van Pelt 2001). The vast majorityof temperate rain forests that once extended fromAlaska to California have now been logged, and agreat deal of research has focused upon the fewremaining old-growth forests. These studies haveconsistently revealed a high biomass and diversityof canopy epiphytes, including lichens, bryophytes,and even vascular plants. Epiphyte communities inforests dominated by Douglas-fir [Pseudotsugamenziesii (Mirb.) Franco] have received a dispro-portionate amount of scientific attention, includingfloristic surveys of Douglas-fir (Clement & Shaw1999; McCune 1993; McCune et al. 2000; Pike etal. 1975, 1977; Sillett 1995), western hemlock

[Tsuga heterophylla (Raf.) Sarg.; Lyons et al. 2000;Sillett & Rambo 2000], and understory vine maple(Acer circinatum Pursh; Ruchty et al. 2001). Epi-phyte communities in temperate rain forests domi-nated by other conifers have scarcely been inves-tigated, except for studies of lower trunks (e.g., Gli-me & Hong 2002).

The tallest temperate rain forests contain coastredwood [Sequoia sempervirens (D. Don) Endl.],and so far only their vascular epiphyte communitieshave been surveyed (Sillett 1999). The high shadetolerance of redwood excludes most other treesfrom these forests, but a few species do co-exist.Sitka spruce [Picea sitchensis (Bong.) Carr.] trees,which grow along the alluvial terraces of some red-wood forests, appear from the ground to support agreater abundance of lichens and bryophytes than

198 [VOL. 106THE BRYOLOGIST

adjacent redwoods. Indeed, the diversity of Sitkaspruce epiphyte communities may rival those ofDouglas-fir forests in Oregon and Washington.

Our objective in this study was to quantify thebiomass and diversity of epiphytic lichens andbryophytes on Sitka spruce trees in an old-growthredwood forest. We used a sampling design thatfacilitated comparisons of our results with previousepiphyte studies in Douglas-fir forests. This studythus provides important baseline information onepiphyte communities in redwood rain forest can-opies.

STUDY AREA

Five tall Sitka spruce trees were selected for detailedstudy. The trees were distributed along the James IrvineTrail in Prairie Creek Redwoods State Park, HumboldtCounty, California. All trees were located along the allu-vial terraces of Godwood Creek between 2.5 and 4.5 kmfrom the Pacific Ocean. The basal area of this riparianforest exceeds 400 m2/ha, of which 76% is redwood, 20%is Sitka spruce, 3% is western hemlock, and 1% is PortOrford cedar [Chamaecyparis lawsoniana (A. Murr.)Parl.]. Sword fern [Polystichum munitum (Kaulf.) C.Presl.] dominates the ground cover and is interspersedwith huckleberry (Vaccinium ovatum Pursh and V. par-vifolium Smith), salal (Gaultheria shallon Pursh), cascara(Rhamnus purshiana DC.), and an assortment of herbs andbryophytes. Mean annual rainfall is 168 cm, and temper-atures range from 4 to 24 C in summer and 0 to 13 C inwinter (Redwood National and State Parks: http://parks.ca.gov/north/ncrd/pcrsp.htm).

METHODS

Tree access. We initially accessed tall tree crowns byshooting rubber-tipped arrows trailing Firelinet filamentover sturdy branches between 70 and 80 m high in thecrowns. Nylon cord was attached to the arrow end of thefilament, hauled over the branches, and then used to haula 10 mm static kernmantle rope into position. One end ofthe rope was anchored at ground level, leaving the otherend of the rope free for ascent via single rope technique.A pulley was installed near the top of the tree, and nyloncord was left in the tree between research days to maintainthe climbing path.

Movement within the crown was facilitated by the useof an arborist-style, split-tail lanyard system. The lanyardutilized two adjustable loops of rope fastened to the climb-er’s harness that can be used simultaneously or individu-ally to allow the climber to ascend, descend, or traverseamong the branches. Each rope loop was adjustable bymeans of a friction hitch connected to a cambium saverwrapped around a sturdy overhead branch (Jepson 2000).The friction hitch allowed the climber to move up or downvia shortening or lengthening the rope loop, and the cam-bium saver minimized the impact of moving rope on del-icate epiphytes and other organisms.

Outer crown access. In the absence of a canopy crane(e.g., McCune et al. 2000), access to the outer and up-permost portions of tall tree crowns is usually quite lim-ited because thin branches cannot support a climber’sweight. We developed a safe and nondestructive ropetechnique to minimize these restrictions. Two 120 m ropeswere fixed near the top of the tree at their midpoints. Eachof the 60 m long ends was then threaded through sequen-

tial, vertically aligned Y-shaped branch forks spaced far-ther out and down in the crown from the central trunkuntil the rope was supported near the crown edge. Addi-tionally, each rope was spaced by azimuth so that therewas a rope on the outer crown in each cardinal direction.These four ropes provided a much-needed fulcrum to safe-ly access epiphytes far from the main trunk near the tipsof branches.

We snagged these ropes up to 8 m away with miniaturegrapnel hooks attached to fly-fishing line loaded on au-tomatic rewind reels mounted to our climbing harnesses.The spring-loaded reel minimized tangling and permittedquick retrieval and storage of the grapnel and line. Oncethe outer crown rope was pulled in to the climber, anascending device was attached to the rope from the climb-er’s harness. By lengthening a lanyard loop already an-chored to an overhead branch, the climber’s weight wastransferred to the outer crown rope, allowing safe non-destructive lateral movement. Despite these advancementsin access, some of the upper- and outermost twigs re-mained inaccessible unless destructive sampling was em-ployed.

Tree mapping. We mapped each tree to establish ran-dom plot locations and to quantify branch lengths andfoliar masses. A fiberglass tape was stretched from groundlevel to the treetop. Main trunk diameters were measuredat ground level, breast height (1.37 m), and 5 m heightintervals. We measured the following variables on eachbranch: height of origin, basal diameter, total horizontalextension from trunk, and cumulative length of branchsegments .5 cm diameter. Branches were individuallysketched, and all segments .5 cm diameter were num-bered sequentially. We then visually estimated lengths ofeach segment to the nearest 0.5 m. These data allowed usto estimate total lengths of branch segments .5 cm di-ameter on each tree.

In a mature tree, foliar units are discrete, repeating clus-ters of leaves that can be counted to quantify foliage onindividual branches. Each tree species has recognizablepatterns of branch growth and development (Halle et al.1978). The foliar units of Sitka spruce are clusters of twigsand leaves supported by branch segments of approximate-ly one cm basal diameter. Counting foliar units on eachbranch allowed us to estimate the foliar mass of an entiretree.

We randomly selected approximately 5% of each tree’stotal branch length and foliar units for epiphyte sampling.About one-fifth of these plots were randomly selected fordestructive sampling (Table 1). The tree mapping datasetensured that all branch surfaces on each tree had an equalprobability of being selected for sampling.

Epiphyte sampling. This study focused on branch-dwelling epiphytes. Trunks were not sampled. Epiphytesampling occurred on two substrate types: axes (i.e.,branches .5 cm diameter) and branchlets (i.e., branches,5 cm diameter, including foliage). We measured % coverof epiphytes on axes via the spiral transect method (Sillett1995). Axis plots occupied 0.5 m of branch length andconsisted of five wraps of a metric diameter-tape spaced10 cm apart. Percent cover of epiphytes was estimated byrecording the presence of species beneath tick marks onthe tape that are p-cm apart (Fig. 1). We used the diameterside of the tape for point intercept sampling to allow fora more reasonable number of tick marks than centimeterspacing would allow. Each time a species was found be-neath a tick mark, we recorded a hit for that species. Cov-er for each species in a plot was calculated as the numberof hits divided by the total number of possible hits in theplot (i.e., transect length in cm divided by p). We mea-

2003] 199ELLYSON & SILLETT: EPIPHYTE COMMUNITIES

TABLE 1. Summary of tree size and epiphyte sampling intensities on axes and branchlets for five Sitka spruce treesin an old-growth redwood forest. Sampling intensities are expressed as percent of tree totals. Numbers in parenthesesare sampling intensities for removed samples only. DBH is diameter of main trunk at breast height (1.37 m).

TreeHeight

(m)DBH(m)

Trunkvolume

(m3)

Axislength

(m)

Foliarmass(kg)

Sampling intensity

Axes Branchlets

12345

89.587.282.592.287.0

1.92.02.42.62.0

8075

101153

83

449337367652424

7998

101212147

5.8 (1.6)5.0 (0.9)6.0 (1.0)5.8 (1.0)5.8 (1.1)

6.5 (2.0)4.8 (0.9)4.8 (1.3)5.1 (1.1)6.0 (1.4)

FIGURE 1. Spiral transect sampling involves wrappinga diameter tape five times around a 0.5 m section ofbranch such that each wrap is 10 cm apart. Species ofepiphytes are recorded beneath tick marks that are p-cmapart.

sured the following additional variables in each axis plot:height above ground, distance from center of trunk to plot,and % sky. Percent sky was a visual estimate of canopyopenness above each plot.

We estimated % cover of epiphytes on branchlets within0.5 by 0.5 m quadrats. Approximately three foliar units fitin a single quadrat (mean 5 3.2, standard error 5 0.1).The cover of each epiphyte species within a quadrat wasvisually estimated to the nearest percent. We measured thefollowing additional variables for each quadrat: heightabove ground, distance from trunk to plot, and % sky.

All epiphytes were removed from approximately one-fifth of the plots and placed in paper bags. For axis plots,we used a pruning saw to carefully cut through thick bryo-phyte mats and minimize the impact of collecting on epi-phytes outside plot boundaries. The bark and cambiumwere not disturbed. For branchlet plots, we used pruningclippers and a saw to remove tree components and theirepiphytes. All destructive samples were transported to thelaboratory for processing.

In the laboratory, we sorted samples by epiphyte spe-cies. We also separated dead organic matter and tree fo-liage from epiphytes. Sorted samples were individuallyplaced in paper bags, oven dried for 24 hours at 60 C,and weighed to the nearest 0.001 g.

Whole-tree masses of epiphytes, dead organic matter,and tree foliage were calculated by dividing total massesof samples removed from a tree by the sampling intensity(i.e., approximately 0.01) for that tree. Masses were de-termined separately for axes and branchlets. Mass esti-mates for the fern Polypodium scouleri were obtainedfrom Sillett and Bailey (2003).

Identifying many chlorolichens in the field to specieswas challenging, because some species required chemicaltesting for confirmation, while others had subtle morpho-logical distinctions that were best viewed under a micro-scope. The genus Usnea was particularly challenging. Us-

nea rubicunda, longissima, and ceratina were easy toidentify, but other species were partitioned into the Usneacornuta group that has tufted thalli, and the U. filipendulagroup that has pendulous thalli with a blackening mainaxis. For similar reasons, Bryoria, Cladonia, and crustoselichens were not identified to species in the field, and cov-er estimates were recorded for genera only. In general,inconspicuous crustose lichens were not carefully exam-ined during field sampling. Removed plots, however, werethoroughly searched for crusts in the laboratory.

Nomenclature of macrolichens, microlichens, mosses,and liverworts followed McCune and Geiser 1997; Brodoet al. 2001; Anderson et al. 1990; and Stotler and Cran-dall-Stotler 1977, respectively. Voucher specimens are de-posited in the Humboldt State University CryptogamicHerbarium (HSC) as well as in the herbarium of RedwoodNational and State Parks.

Data analyses. In order to reduce noise from infre-quent species, epiphytes occurring in fewer than 5% ofthe plots were removed from the data matrix prior to allmultivariate analyses. Bray-Curtis ordination was per-formed on the resulting primary matrix of 430 plots 3 31species using the variance–regression method of endpointselection and the Sørenson coefficient as the distance mea-sure (McCune & Mefford 1999). Presence/absence speciesdata were transformed by the Beals smoothing functionprior to ordination analysis. This transformation smoothedpatterns of joint occurrences of species and facilitated ex-traction of the dominant epiphyte compositional gradient(McCune 1994). Variance extracted by an ordination axiswas expressed by the coefficient of determination betweenEuclidean distances in ordination space and Sørenson dis-tances in the original species space. A secondary matrixof 430 plots 3 6 additional variables (i.e., height aboveground, % sky, number of lichen species, number of bryo-phyte species, lichen cover, and bryophyte cover) was alsoconstructed. Correlations among these variables and plotordination scores were used to aid interpretation of thedominant compositional gradient.

We performed cluster analysis on the transposed pri-mary matrix in order to delimit groups of epiphytes thatshared similar habitats or substrate preferences (McCune& Mefford 1999). Following the rationale of McCune etal. 2000, log-transformed cover values were relativized byspecies sums of squares prior to this analysis, which usedWard’s method of clustering and a Euclidean distance ma-trix, and the resulting dendrogram was scaled by Wishart’sobjective function converted to a percentage of informa-tion remaining.

We performed association analysis to test whether in-dividual epiphyte species pairs co-occurred more or lessoften than expected by chance. A 2 3 2 contingency tablewas generated for each pair of the 31 species used in the


TABLE 2. Epiphytes encountered on five Sitka spruce trees in an old-growth redwood forest. Cover of each taxonwas calculated separately for the sheltered, intermediate, and exposed portions of the epiphyte compositional gradient,which represented 98.5% of the information in an ordination analysis of 430 plots containing 31 species. Frequencieswere calculated separately for axes (branches .5 cm diameter) and branchlets (branches ,5 cm diameter, includingfoliage). Frequency values for all Bryoria and Usnea species were derived from removed samples only. Each species’proportion of the total estimated biomass of epiphytes on the five trees (i.e., 294.3 kg) is also indicated. An asterisk(*) indicates species groups that were identified in the field by genus only. ** Usnea cornuta cover values include U.chaetophora, U. cornuta, U. esperantiana, U. fragilescens var. mollis, U. scabrata, and U. wirthii. *** Usnea filipendulacover values include U. filipendula and U. madeirensis. Orthotrichum papillosum, 5 O. lyellii (Anderson et al. 1990).A dagger (†) indicates species found in study trees but outside plot boundaries.

Taxon Code

Cover (%)

ShelteredInter-

mediate Exposed

Frequency (%)

AxesBranch-

letsBiomass(% Total)

ALL EPIPHYTESCYANOLICHENSErioderma sorediatumLobaria oreganaLobaria pulmonariaLobaria scrobiculata

ERSOLOORLOPULOSC

77.10.2

0.1

48.55.8

0.33.90.3

32.07.0

0.54.20.6

25.2

3.418.9

8.7

71.00.9

11.257.632.6

,0.10.86.70.4

Nephroma bellumNephroma helveticumNephroma laevigatumPseudocyphellaria anomalaPseudocyphellaria anthraspisPseudocyphellaria crocataSticta limbataCHLOROLICHENSFRUTICOSE CHLOROLICHENSAlectoria imshaugiiAlectoria vancouverensisBryoria spp.*Bryoria capillarisBryoria fremontiiBryoria furcellataBryoria trichodes ssp. americanaCladonia spp.*Cladonia chlorophaea

NEBENEHENELAPSANPSAPPSCRSTLI

ALIMALVABRspp.BRCABRFRBRFUBRTRCLspp.CLCH

,0.1

,0.1,0.1,0.1

3.50.2

0.1,0.1

1.4

0.1

,0.10.20.60.2

,0.18.03.3

1.0,0.1

0.4

,0.1,0.1

0.90.20.40.1

11.86.4

2.7,0.1

,0.1

1.0

5.32.94.41.9

82.042.2

0.519.9

2.41.5

38.30.5

12.90.41.3

40.612.912.929.087.179.9

47.85.84.00.40.41.32.2

0.1,0.1,0.1

0.50.30.20.2

,0.12.0

,0.1,0.1,0.1,0.1

,0.1Cladonia furcataCladonia squamosaCladonia subulateCladonia sulphurinaCladonia transcendensRamalina farinaceaRamalina roesleriSphaerophorous globosusUsnea ceratinaUsnea chaetophoraUsnea cornuta**Usnea esperantianaUsnea filipendula***Usnea fragilescens var. mollisUsnea longissimaUsnea madeirensisUsnea rubicundaUsnea scabrata

CLFUCLSQCLSUCLSFCLTRRAFARAROSPGLUSCEUSCHUSCOUSESUSFIUSFRUSLOUSMAUSRUUSSC

0.1

,0.1

,0.1

,0.1

,0.10.4

1.3

0.3

0.1

,0.1

,0.1,0.1

0.1

2.3

0.6

0.2

,0.1

0.51.00.50.5

35.4

1.019.4

2.026.5

2.016.3

6.16.18.22.02.0

2.20.41.8

14.30.47.8

66.713.745.119.6

7.811.819.6

,0.1,0.1,0.1,0.1

0.2,0.1,0.1

0.1,0.1,0.1

0.90.10.70.10.30.1

,0.1,0.1

Usnea wirthiiFOLIOSE CHLOROLICHENSCavernularia hulteniiCavernularia lophyreaCetrelia cetrarioidesCetraria chlorophyllaHeterodermia leucomelosHypogymnia apinnataHypogymnia enteromorphaHypogymnia imshaugii

USWI

CAHUCALOCECECECHHELEHYAPHYENHYIM

,0.1

,0.1,0.1

3.6,0.1

,0.1,0.1

2.4

5.20.1

,0.1,0.1,0.1

0.13.3

6.134.0

1.5

1.92.9

25.70.50.5

45.179.023.2

0.90.43.1

40.259.8

0.1

,0.1,0.1,0.1,0.1

0.11.8

,0.1,0.1

Hypogymnia physodesMenegazzia terebrata

HYPHMETE ,0.1 ,0.1

0.50.5 1.3

,0.1,0.1


TABLE 2. Continued.

Taxon Code

Cover (%)

ShelteredInter-

mediate Exposed

Frequency (%)

AxesBranch-

letsBiomass(% Total)

Parmelia hygrophilaParmelia saxatilis

PAHYPASA ,0.1 ,0.1 0.5

0.40.9

,0.1,0.1

Parmelia squarrosaParmelia sulcataParmotrema arnoldiiParmotrema chinense

PASQPASUPTARPTCH

,0.1 0.2,0.1,0.1,0.1

0.40.1

0.2

10.24.91.01.5

25.07.1

12.5

0.1,0.1,0.1

0.1Parmotrema crinitumPlatismatia glaucaPlatismatia herreiPlatismatia norvegicaCRUSTOSE CHLOROLICHENSCalicium virideCaloplaca sp.Graphis sp.Lecanora sp.Lecanora pulicarisLecanora subfuscaLepraria sp.Loxosporopsis coralliferaMycoblastus sanguinariusOchrolechia spp.*Ochrolechia juvenalisOchrolechia szatalaensisOchrolechia trochophora

PTCRPLGLPLHEPLNO

CAVICAsp.GRsp.LEsp.LEPULESULEsp.LOCOMYSAOCspp.OCJUOCSZOCTR

,0.1

1.90.1

,0.1

1.8,0.1

,0.1

0.8,0.1

0.1,0.1

0.7

0.2,0.1

0.10.4

1.0,0.1,0.1

0.2

,0.1

0.1

14.60.53.40.5

49.00.50.50.5

37.41.51.08.77.80.50.5

57.11.80.4

6.30.4

0.50.50.50.9

2.22.2

0.8,0.1,0.1,0.1

,0.1,0.1,0.1,0.1,0.1,0.1,0.1,0.1,0.1

,0.1,0.1,0.1

Pertusaria ophthalmizaThelotrema lepadinumXylographa sp.BRYOPHYTESMOSSESAntitrichia curtipendulaAulacomnium androgynum†Bryum capillareDicranum fuscescensEurhynchium oreganumHypnum circinaleIsothecium cristatum†Isothecium myosuroidesNeckera douglasiiOrthotrichum papillosumPlagiothecium undulatumLIVERWORTSCephaloziella rubella

PEOPTHLEXYsp.

ANCUAUANBMCADIFUEUORHYCIISCRISMYNEDOORPAPLUN

CERU

71.056.9

3.8

0.20.50.6

,0.1

42.89.00.1

,0.114.1

0.1

28.520.2

0.7

,0.1

16.03.30.3

8.3

0.1

6.22.4

,0.1

,0.1

1.60.60.2

3.8

2.9

0.599.094.216.5

1.514.6

0.50.5

92.264.6

9.20.5

96.10.5

5.40.4

100.080.810.3

0.9

63.862.917.4

98.2

,0.1,0.1,0.1

5.8,0.1,0.1

0.10.6

,0.1,0.141.2

5.60.1

,0.1

,0.1Douinia ovata†Frullania nisquallensisPorella navicularisRadula bolanderiScapania bolanderiFERNSPolypodium glycyrrhizaPolypodium scouleri

DOOVFRNIPONARABOSCBO

POGLPOSC

6.87.3

,0.1,0.1

2.21.90.3

6.41.70.2

0.30.20.1

3.70.1

93.265.0

3.41.0

43.739.8

4.4

97.827.2

0.4

3.12.70.4

,0.14.73.3

,0.1,0.1

8.413.3

multivariate analyses. The likelihood ratio (G, Sokal &Rohlf 1995) was used as a test statistic.

RESULTS

The Sitka spruce trees surveyed in this studywere between 82.5 to 92.2 m tall and 1.9 to 2.6 mdiameter at breast height. The largest Sitka spruce,tree four, had twice as much axis length and trunkvolume and almost three times as much foliar mass

as the other trees (Table 1). The proportion of axislength to foliar mass was similar among four of thetrees (i.e., 2.9 to 3.6 m/kg), but tree one had a muchhigher ratio (i.e., 5.7 m/kg), because it was im-pacted by a disturbance event that left a number ofbroken branches in the middle of its crown. Thisdeclining tree also had numerous conks of theheart-rot fungus, Phellinus pini [(Thore.:Fr.) A.Ames.].


TABLE 3. Epiphyte species diversity associated withvarious sample groupings on five Sitka spruce trees in anold-growth redwood forest. Species richness was ex-pressed as the mean number of species per plot 6 onestandard deviation. Beta diversity was calculated by di-viding the total number of species within a group by themean number of species per plot within a group. Samplessizes are indicated in parentheses.

Sampledelineation

Speciesrichness

Betadiversity

Totalnumber

of species

By epiphyte group (430)BryophytesLichensCyanolichensChlorolichensFerns

3.2 6 1.35.5 6 4.21.3 6 1.74.1 6 3.10.2 6 0.4

4.713.3

8.315.0

8.8

17721161

2

By substrateAxes (206)Branchlets (224)

7.0 6 2.69.2 6 3.9

8.55.7

6052

By ordination stratumSheltered (143)Intermediate (144)Exposed (143)

5.8 6 2.08.1 6 3.3

10.8 6 3.4

6.97.14.9

405753

TABLE 4. Estimated masses of epiphytes and associated dead organic matter (DOM) on axes (branches .5 cmdiameter) and branchlets (branches ,5 cm diameter) of five Sitka spruce trees in an old-growth redwood forest. Valuesare mean dry masses (kg) per tree 6 one standard error.

Bryophytes Cyanolichens Chlorolichens Ferns Biomass DOM

AxesBranchlets

33.3 6 8.32.9 6 0.8

1.6 6 0.53.9 6 1.5

1.1 6 0.43.4 6 1.2

12.7 6 2.10.0 6 0.0

48.7 6 8.010.2 6 1.5

128.1 6 19.52.8 6 0.6

The epiphyte flora of Sitka spruce in a redwoodforest. We encountered a total of 91 epiphyte spe-cies on the five trees, including 57 macrolichens,15 crustose lichens, 11 mosses, six liverworts, andtwo ferns (Table 2). Ten species contributed morethan 1% of total epiphyte biomass, including threemosses (Isothecium myosuroides, Antitrichia curti-pendula, and Neckera douglasii), two liverworts(Frullania nisquallensis and Porella navicularis),one cyanolichen (Lobaria pulmonaria), two chlo-rolichens (Alectoria vancouverensis and Hypogym-nia apinnata), and both Polypodium ferns (Table 2).Beta diversity was nearly three times higher for li-chens than for bryophytes, but plot-level speciesrichness was similar for lichens and bryophytes(Table 3). Branchlet plots supported more epiphytespecies and had a more consistent species compo-sition (i.e., lower beta diversity) than axis plotseven though we encountered more species on axesthan branchlets overall (Table 3).

Epiphyte biomass and proportions of epiphytegroups were very different on axes and branchlets(Table 4). Axes supported 83% of the epiphyte bio-mass and 98% of the mass of dead organic matter

on a tree’s branch systems. Overall, the mass ofdead organic matter was higher than epiphyte bio-mass (69 vs. 31% of total mass). Bryophytes ac-counted for 62% of the total epiphyte biomass andwere much more abundant on axes than branchlets(92 vs. 8% of bryophyte biomass). Lichens, on theother hand, were less abundant on axes thanbranchlets (27 vs. 73% of lichen biomass). Fernscontributed the remainder of epiphyte biomass(22%) and were nearly restricted to axes.

Bryophytes clearly dominated epiphyte commu-nities on Sitka spruce, forming thick mats that con-sistently covered branch surfaces in the lower two-thirds of the crowns. Relatively high cover of bothmosses and liverworts occurred throughout treecrowns on axes, but bryophyte cover was quite lowon branchlets (Fig. 2). The moss Isothecium myo-suroides was the most abundant species, contrib-uting over 41% of total epiphyte biomass. The liv-erwort Frullania nisquallensis occupied the broad-est range of epiphytic habitats, occurring in 96% ofall plots sampled (Table 2).

Polypodium ferns were most abundant between40 and 50 m above the ground, although smallamounts of ferns occurred nearly to the treetops(Fig. 2). Despite contributing over 13% of total epi-phyte biomass, P. scouleri occurred in less than 5%of the plots and was not among the 31 species con-sidered in multivariate analyses. Biomass estimatesfor P. scouleri were derived from Sillett and Bailey(2003), who sampled every fern mat on the fivetrees (n 5 89). According to these estimates, over60% of fern biomass was attributable to P. scouleri.We probably underestimated biomass of P. glyc-yrrhiza, because some of our sampling occurredduring the early fall when this deciduous specieshas no leaves.

Cyanolichens contributed less to overall epiphytebiomass than bryophytes and ferns, but they wereslightly more abundant than chlorolichens (Table4). Lobaria pulmonaria was by far the most abun-dant cyanolichen with nearly three times as muchbiomass as all other cyanolichens combined. Cy-anolichens were scarce in the lower crowns, butthey were less restricted in their vertical range thanmost chlorolichens (Fig. 2). This broader distribu-tion is also reflected by the lower beta diversity forcyanolichens than chlorolichens (Table 3).


FIGURE 2. Epiphyte cover in 10 m height strata for axes and branchlets on Sitka spruce in an old-growth redwoodforest. Values are means 6 one standard error (n 5 5 trees).

Two-thirds of all epiphyte species encounteredon the five trees were chlorolichens. These specieswere divided into four morphological groups thatvaried in abundance and distribution. Foliose andfruticose chlorolichens were mostly restricted to theupper crowns and represented the bulk of chloro-lichen biomass. Together, these groups coverednearly 40% of the tree surfaces in the top 10 m ofthe crowns (Fig. 2). Crustose lichens and Cladoniawere more widely distributed, reaching peak abun-dances on axes between 50 and 70 m above theground (Fig. 2).

Associations among epiphytes. There weremany positive and negative associations among the31 epiphyte species investigated in the multivariateanalyses (Fig. 3). It is precisely because of thesenumerous associations that multivariate data reduc-tions via ordination and cluster analyses were sosuccessful (see below). Most bryophyte species co-occurred in mats except for Orthotrichum papillos-um, which often intermingled with lichens, espe-cially Parmotrema crinitum, on branchlets. Asso-

ciations among lichens were both positive and neg-ative and did not clearly relate either to type ofphotobiont or to thallus morphology. The only li-chens positively associated with bryophytes wereCladonia and Lepraria. These lichens were, in turn,negatively associated with many other lichen spe-cies. Lepraria was typically found on the under-sides of large branches supporting thick overlyingbryophyte mats. The fern P. glycyrrhiza always oc-curred amidst bryophyte mats. Multivariate analy-ses provided further insights on species-habitat re-lationships.

Dominant compositional gradient. The Bray-Curtis ordination performed on Beals-smoothedpresence-absence data extracted one axis explain-ing 98.5% of the total variation in the species data.Height (r 5 0.63), % sky (r 5 0.66), lichen speciesrichness (r 5 0.76), and lichen cover (r 5 0.35)were positively correlated, while bryophyte speciesrichness (r 5 20.60) and bryophyte cover (r 520.72) were negatively correlated with ordinationscores along this gradient. Ranking all 430 plots by


FIGURE 3. Dendrogram (left) and associations analyses (right) for epiphytes on Sitka spruce trees in an old-growthredwood forest. Each square in the associations grid represents the result of a 2 by 2 contingency table for a speciespair. Only species occurring in 5% or more of the sampled plots are included. Shaded squares indicate statisticallysignificant positive or negative associations. An X indicates that the species pair did not co-occur in any sampled plot.An O indicates 100% co-occurrence for the less frequent species of the pair. Order of species was determined by thedendrogram. Ten species groups are indicated at the level of 50% information remaining on the dendrogram. Corre-lations of species along the dominant compositional gradient are also listed. Refer to Table 2 for full names of speciescodes.

ordination scores and dividing the ranked list intothirds assisted with the interpretation of the gradi-ent. The resulting three sets of plots were classifiedas exposed, intermediate, and sheltered (see Dis-cussion). Exposed plots occurred predominately onbranchlets in outer and upper crowns, intermediateplots were on axes in middle to upper crowns andon branchlets in middle to lower crowns, and shel-tered plots were mostly restricted to axes in middleto lower crowns (Fig. 4).

Epiphyte composition among the three groups ofplots varied dramatically (Fig. 5). Exposed plotswere dominated by chlorolichens with roughlyequivalent amounts of co-occurring cyanolichensand bryophytes. Liverworts dominated the bryo-phyte component in exposed plots. Sheltered plotswere overwhelmingly dominated by bryophyteswith very little lichen cover. In sheltered plots,crustose lichens (93% Lepraria) and Cladoniadominated the lichen component, and mosses dom-inated the bryophyte component. Intermediate plotswere dominated by mosses with roughly equivalentamounts of co-occurring liverworts, cyanolichens,and chlorolichens. Exposed plots had the lowestbeta diversity (Table 3), indicating the relatively ho-mogenous distribution of macrolichens in this por-tion of the gradient.

Species groups. Cluster analysis revealed dis-

tinct groups of species that had similar distributionpatterns within the crowns. The dendrogram wastrimmed at the level of 50% information remaining,and the resulting ten groups were interpreted bygraphically combining the dendrogram with resultsfrom ordination and association analyses (Fig. 3)and by examining scatterplot distributions of indi-vidual species (Fig. 6). Species within a group tend-ed to be positively associated with each other, havesimilar associations with other species, and havesimilar distributions within the crowns.

Vertical distribution patterns as well as affinitiesfor axes versus branchlets varied considerablyamong the ten groups of species. Some groups con-tained species that occurred throughout the crowns,while others contained species with more limiteddistributions. Each group of species is described be-low.

1. High exposure, high frequency lichens: Alec-toria vancouverensis, Hypogymnia apinnata, Usneacornuta, Heterodermia leucomelos, Lobaria pul-monaria, Parmotrema crinitum, Parmelia squar-rosa, and Cavernularia hultenii. These specieswere most frequent and abundant in upper crowns,but they also occurred in middle and sometimeslower crowns. Heterodermia and Cavernulariawere both nearly restricted to branchlets (Fig. 6[1A–H]).


FIGURE 4. Spatial distribution of 430 sample plots onaxes and branchlets of five Sitka spruce trees in an old-growth redwood forest. Classifications of the plots as ex-posed, intermediate, and sheltered were based upon ordi-nation analysis of sample plots in species space.

2. Middle to upper crown cyanolichen: Lobariaoregana. Although similar in crown distribution toLobaria pulmonaria, L. oregana was much less fre-quent and abundant. Both Lobaria species contrast-ed with other cyanolichens by occupying axes near-ly as often as branchlets. Lobaria oregana was theonly cyanolichen not encountered below 65 m (Fig.6 [2]).

3. Broad-ranging, mat-forming moss: Antitrichiacurtipendula. This relatively infrequent moss wasabundant when found on axes within the lower two-thirds of the crowns, but it also occurred on branch-lets in middle to lower crowns and occasionallyeven on axes in upper crowns (Fig. 6 [3]).

4. Middle to upper crown moss: Orthotrichumpapillosum. This moss was the only bryophytemore abundant in exposed and intermediate plots

than in sheltered plots. It was more frequent onbranchlets than axes (Fig. 6 [4]).

5. Intermediate exposure lichens: Ochrolechiaand Sphaerophorus globosus. These taxa had theirgreatest abundance on intermediate exposure plots(Table 2), but they were negatively associated witheach other. Ochrolechia typically covered bare barkon axes, whereas Sphaerophorus often occurredamidst bryophyte mats or on branchlets (Fig. 6[5A,B]).

6. Middle to upper crown chlorolichens: Par-melia sulcata, Parmotrema chinense, Usnea filipen-dula, and Usnea longissima. These chlorolichensdid not extend into uppermost crowns and were lessfrequent than chlorolichens of group 1. Parmeliasulcata and Usnea filipendula were nearly as fre-quent on axes as branchlets, while the other twospecies were much more frequent on branchlets(Fig. 6 [6A–D]).

7. High exposure cyanolichens: Lobaria scrobi-culata, Pseudocyphellaria anomola, Sticta limbata,and Pseudocyphellaria crocata. These cyanolich-ens occurred nearly to the treetops, and they weremuch more frequent on branchlets than axes, es-pecially Sticta (Fig. 6 [7A–D]).

8. Intermediate exposure cyanolichens: Nephro-ma bellum and Pseudocyphellaria anthraspis. Un-like all other cyanolichens, these species were mostabundant on intermediate exposure plots (Fig. 6[8A,B]).

9. Bryophyte-covered axis epiphytes: Cladonia,Lepraria, Polypodium glycyrrhiza, and Dicranumfuscescens. These species were found almost exclu-sively on axes covered with bryophyte mats. Cla-donia, Lepraria, and P. glycyrrhiza exhibited broadvertical distributions, whereas Dicranum was near-ly absent from upper crowns (Fig. 6 [9A–D]).

10. Crown-wide, mat-forming bryophytes: Frul-lania nisquallensis, Isothecium myosuroides, Neck-era douglasii, and Porella navicularis. These bryo-phytes were distributed throughout the crowns onboth axes and branchlets, but their cover was con-sistently higher on axes. Frullania was the onlymember of the group that was more frequent onbranchlets than axes (Table 2), thus explaining itspositive correlation along the dominant composi-tional gradient (Fig. 6 [10A–D]).

DISCUSSION

Epiphytes on Sitka spruce vs. Douglas-fir.Quantitative comparisons between epiphyte com-munities on Sitka spruce and Douglas-fir are pos-sible because of the extensive research on Douglas-fir epiphytes (e.g., Clement & Shaw 1999; McCune1993; McCune et al. 2000; Pike et al. 1975; Sillett1995). Douglas-fir and Sitka spruce are the second


FIGURE 5. Relative proportions of tree total epiphyte cover for exposed, intermediate, and sheltered portions of thedominant compositional gradient on five Sitka spruce trees in an old-growth redwood forest. Classifications as exposed,intermediate, and sheltered were based upon ordination analysis of sample plots in species space. Ferns in this diagraminclude only P. glycyrrhiza.

and third tallest tree species on Earth, respectively;only coast redwood is taller. Both species, whichare members of Pinaceae, also have a high affinityfor epiphytes, and their geographic ranges overlap.Differences include shade tolerance (Sitka spruce.Douglas-fir) and habitat preference (i.e., Sitkaspruce is restricted to moist, low elevation coastalforests, while Douglas-fir extends into higher ele-vations in drier, more continental climates; Burns& Honkala 1990). Sitka spruce also differs fromDouglas-fir in having persistent leaf bases on itstwigs.

There is substantial overlap in epiphyte speciescomposition between Sitka spruce and Douglas-fir,but there are some important floristic differences.

The high diversity of Usnea (11 spp.) on Sitkaspruce in this study contrasts with the low diversity(3 spp.) observed on Douglas-fir by McCune et al.(2000). Hypogymnia diversity is surprisingly lowon Sitka spruce with all but three of 185 occur-rences being H. apinnata in this study. Eight andseven species of this genus were reported on Doug-las-fir by McCune et al. (2000) and Sillett (1995),respectively. Several cyanolichens reported onDouglas-fir did not occur on the Sitka spruce westudied, including Nephroma occultum, Pseudocy-phellaria rainierensis, and Sticta weigelii. Theseold-growth associated species (see Rosentreter1995; Sillett & Neitlich 1996) apparently reach thesouthern limits of their geographic ranges in


Oregon. Finally, the chlorolichens Heterodermialeucomelos, Parmelia squarrosa, Parmotrema crin-itum, Ramalina roesleri, Usnea rubicunda, and U.wirthii and the cyanolichen Erioderma sorediatumhave not yet been reported on Douglas-fir.

Several bryophytes reported on both Douglas-firand Sitka spruce differed markedly in their abun-dance on the two tree species. Antitrichia curtipen-dula was the most abundant moss in a 700-year-oldDouglas-fir forest (Sillett 1995), while Isotheciummyosuroides was seven times more abundant thanA. curtipendula on Sitka spruce in this study. Doui-nia ovata can be abundant on Douglas-fir, coveringthe lower surfaces of large, moss-laden branches(Sillett 1995). We found D. ovata only twice onSitka spruce, and both times it was outside plotboundaries. Frullania nisquallensis was the mostabundant liverwort in this study, but it was muchmore sparsely distributed in the 450-year-old and700-year-old Douglas-fir forests studied by Pike etal. (1975) and Sillett (1995), respectively. This te-nacious liverwort occurred on nearly every avail-able substrate throughout Sitka spruce crowns, in-cluding bark, foliage, and even other epiphytes.

Cyanolichen abundance also differed betweenthis study and previous studies of Douglas-fir. Lo-baria oregana is consistently reported as the dom-inant epiphytic lichen in old-growth Douglas-firforests, but L. pulmonaria was far more abundantthan L. oregana on the five Sitka spruce trees sur-veyed in this study. However, we did observe anincredibly high quantity of L. oregana on a 97-m-tall Sitka spruce tree in Prairie Creek RedwoodsState Park (Sillett & Ellyson, unpubl.). Such tree-to-tree variation in L. oregana abundance may bea consequence of this species’ very limited dis-persal capabilities (Sillett et al. 2000a,b). The pau-city of epiphytic bryophytes and cyanolichens onredwoods suggests that this tree species may notprovide suitable habitats for many species, includ-ing L. oregana. The overwhelming dominance ofredwoods in these forests may present a problemfor slowly dispersing epiphytes; suitable trees maybe spatially distant in a forest composed mostly ofinhospitable redwoods. Dispersal limitations mayalso be invoked to explain the otherwise puzzlingabsence of Peltigera britannica on Sitka spruce.This cyanolichen, which does occur in northernCalifornia, is associated with bryophyte mats inold-growth Douglas-fir forest canopies (Sillett1995). Suitable substrates for P. britannica are thusplentiful on Sitka spruce.

Tree-to-tree variation. Of the 91 species of epi-phytes encountered in this study, 32 of these werefound on a single tree, while only 30 species werefound on all five trees. High tree-to-tree variationin epiphyte communities is also indicated by the

reduction in beta diversity when calculated for in-dividual trees (mean 5 6.3) compared to all fivetrees together (10.1). Dispersal limitations of epi-phytes probably contributed to this heterogeneity,but variation in canopy structure may also be im-portant.

While the five study trees varied somewhat insize, site-to-site variation in canopy structure ap-pears to be more important to epiphytes than treesize itself. Only tree three had a non-emergent up-per crown and was closely surrounded by tall,neighboring trees. On average, plots in this treewere less exposed to sunlight than plots in the othertrees (38% vs. 49% sky, one-way ANOVA, F 53.8, p , 0.005). Ten species were found only onthis tree, including two normally terrestrial mosses(i.e., Eurhynchium oreganum and Plagiotheciumundulatum). Aboveground adventitious roots werealso found amidst bryophyte mats on several of itsbranches. These roots had well-developed ectomy-corrhizal associations with Leucopaxillus amarus[(Alb. & Schw.:Fr.) Kuhn.], whose mushroomswere found on the branches. Dense shading, terres-trial mosses, and the presence of canopy roots sug-gest that epiphytes on tree three were subject to amore humid, sheltered environment than epiphyteson the other four trees.

Whole-tree epiphyte mass estimates. Excludinglarge redwoods, which can support hundreds of kgof ferns alone (Sillett & Bailey 2003), our massestimates indicate that Sitka spruce in redwood for-ests can support more epiphytic material (i.e., 59kg biomass and 131 kg dead organic matter pertree) than any conifer previously studied. For com-parison, Pike et al. (1977) found 17.8 kg of epi-phytes per tree on Douglas-fir in Oregon, Clementand Shaw (1999) found 27.1 kg of epiphytes pertree on Douglas-fir in Washington, and Rhoades(1981) found 12.8 kg of epiphytes per tree on sub-alpine fir [Abies lasiocarpa (Hook.) Nutt.].

Biomass ratios of bryophytes, cyanolichens, andchlorolichens vary greatly among the coniferousforests for which whole-tree estimates of epiphytebiomass are available (i.e., 8:1:1 in this study, 1:5:1 in Pike et al. 1977, 1:1:11 in Clement & Shaw1999, and 1:3:2 in Rhoades 1981). Sitka spruce isthe only tall conifer surveyed to date with a higherbryophyte than lichen biomass. The relatively smallproportions of chlorolichen and cyanolichen bio-mass on Sitka spruce are likely due to the verticaldisplacement of lichens by bryophytes in very wetforests (Sillett & Goward 1998; Sillett & Neitlich1996). In merely moist forests, bryophytes are re-stricted to the lower and middle canopy, cyanolich-ens dominate the middle canopy, and chlorolichensare most abundant in the upper canopy (McCune1993; Sillett & Rambo 2000). In relatively dry for-


FIGURE 6. Distribution of 31 epiphyte species on five Sitka spruce trees in an old-growth redwood forest. Graphsdepict crown locations of each plot containing a given species among the 430 plots sampled on all five trees. Symbolsindicate cover of epiphytes. A plus sign (1) indicates less than 1% cover, an open circle indicates 1 to 5% cover, anda black triangle indicates greater than 5% cover. Axis plots are displayed on the right, and branchlet plots are displayedon the left side of each graph. Epiphyte species are presented in ten groups derived from cluster analysis (see dendro-gram in Fig. 3).

ests, bryophytes and cyanolichens are scarce, andchlorolichens dominate the canopy (Clement &Shaw 1999).

Epiphyte gradients in tall forests. The fact thatepiphyte composition changes with height has been

well documented in many forests. There are pro-nounced changes in microclimatic conditions fromthe top to the bottom of forests that are particularlyrelevant to poikilohydric epiphytes. Humidity de-creases and exposure to desiccating wind increases


FIGURE 6. Continued.

with height in the canopy (Campbell & Coxson2001; Parker 1995). Since individual species varyin desiccation tolerance, epiphyte communities arestrongly stratified by height, especially in tall, moistforests (McCune 1993; McCune et al. 2000).

We interpreted the dominant compositional gra-dient in Sitka spruce epiphyte communities as anexposure gradient because of positive correlationsbetween ordination scores and both height and %

sky as well as the predictable, associated changesin epiphyte communities along such a gradient (i.e.,increasing lichen and decreasing bryophyte diver-sity). And while the vertical component of the ex-posure gradient is particularly pronounced in anold-growth redwood forest, a horizontal componentto this gradient also exists. Exposure to desiccationprobably increases along branch systems from axesin the inner crown to branchlets in the outer crown.


In this study, exposed plots extended much fartherdown in the crowns on branchlets than on axes. Adownward extension of exposed plots was also ob-served along the clearcut edge of an old-growthDouglas-fir forest (Sillett 1995). However, com-munity-level differences between axes and branch-lets probably represent an interplay between des-iccation tolerance and succession. Desiccation sen-sitive epiphytes (e.g., many bryophytes) may be un-able to survive in the xeric environment of exposedsites in upper and outer crowns, while desiccationtolerant epiphytes (e.g., many lichens) may be un-able to compete with aggressive, mat-forming bryo-phytes in the mesic environment of sheltered sitesin lower and inner crowns. Manipulative experi-ments (e.g., Stone 1989) are needed to evaluate therelative importance of these factors for epiphyteson Sitka spruce.

Ecological importance of Sitka spruce epi-phytes. Lichens, bryophytes, and ferns co-exist onSitka spruce branches, but the overwhelming dom-inance of bryophytes is indicative of a losing suc-cessional battle for most lichens. Relentless bryo-phytes cover entire axes, often doubling the appar-ent basal diameters of branches. Only foliatedbranchlets remain free of these mats. The immensesurface areas of sponge-like bryophyte mats pro-mote nutrient capture from leachates and through-fall (Nadkarni 1986), and their high water holdingcapacity increases moisture availability for associ-ated organisms (Sillett & McCune 1998; Veneklaaset al. 1990).

In addition to Polypodium ferns and a few li-chens, a wide variety of other organisms flourishamidst bryophytes on Sitka spruce. We regularlyobserve fungi, mollusks, annelids, and arthropodson thick bryophyte mats during the rainy season.Rodents, marbled murrelets (Brachyramphus mar-moratus), and wandering salamanders (Aneides va-grans) also utilize these mats. These epiphytes maytherefore serve a critical role in maintaining thebiodiversity of redwood forest canopies.

ACKNOWLEDGMENTS

Funding for this research was provided by grants fromGlobal Forest Society (GF-18-1999-54) and Save-the-Redwoods League. We thank Humboldt State Universitystudents for assistance with fieldwork (Leah Larsen, J.Brett Lovelace, and Susan Stearns) and curation of her-barium specimens (Sunny Bennett). We also thank RogerRosentreter and Fred Rhoades for comments on the man-uscript.

LITERATURE CITED

ANDERSON, L. E., H. A. CRUM & W. R. BUCK. 1990. Listof the mosses of North America north of Mexico. THE

BRYOLOGIST 93: 448–499.BURNS, R. M. & B. H. HONKALA. 1990. Silvics of North

America, Vol. 1, Conifers. Agricultural Handbook654. U.S.D.A. Forest Service, Washington, D.C.

BRODO, I. M., S. D. SHARNOFF & S. SHARNOFF. 2001. Li-chens of North America. Yale University Press, NewHaven, CT.

CAMPBELL, J. & D. S. COXSON. 2001. Canopy microcli-mate and arboreal lichen loading in subalpine spruce-fir forest. Canadian Journal of Botany 79: 537–555.

CLEMENT, J. P. & D. C. SHAW. 1999. Crown structure andthe distribution of epiphyte functional groups in Pseu-dotsuga menziesii trees. Ecoscience 6: 243–254.

GLIME, J. M. & W. S. HONG. 2002. Bole epiphytes on threeconifer species from Queen Charlotte Islands, Canada.THE BRYOLOGIST 105: 451–464.

HALLE, F., R. A. A. OLDEMAN & P. B. TOMLINSON. 1978.Tropical Trees and Forests: An Architectural Analysis.Springer-Verlag, NY.

JEPSON, J. 2000. The Tree Climber’s Companion. BeaverTree Publishing, Longville, MN.

LYONS, B., N. M. NADKARNI & M. P. NORTH. 2000. Spatialdistribution and succession of epiphytes on Tsuga het-erophylla in an old-growth Douglas-fir forest. Cana-dian Journal of Botany 78: 957–968.

MCCUNE, B. 1993. Gradients in epiphyte biomass in threePseudotsuga-Tsuga forests of different ages in westernOregon and Washington. THE BRYOLOGIST 96: 405–411.

———. 1994. Improving community analysis with theBeals smoothing function. Ecoscience 1: 82–86.

——— & L. GEISER. 1997. Macrolichens of the PacificNorthwest. Oregon State University Press, Corvallis,OR.

——— & M. J. MEFFORD. 1999. Multivariate analysis onthe PC-ORD system. Version 4. MJM Software, Gle-neden Beach, OR.

———, R. ROSENTRETER, J. M. PONZETTI & D. C. SHAW.2000. Epiphyte habitats in an old conifer forest inwestern Washington. THE BRYOLOGIST 103: 417–426.

NADKARNI, N. M. 1986. The nutritional effects of epi-phytes on host trees with special reference to alterationof precipitation chemistry. Selbyana 9: 44–51.

PARKER, G. G. 1995. Structure and microclimate of forestcanopies, pp. 73–106. In M. D. Lowman & N. M.Nadkarni (eds.), Forest Canopies. Academic Press,NY.

PIKE, L. H., W. C. DENISON, D. M. TRACY, M. A. SHER-WOOD & F. M. RHOADES. 1975. Floristic survey of epi-phytic lichens and bryophytes growing on old-growthconifers in western Oregon. THE BRYOLOGIST 78: 389–402.

———, R. A. RYDELL & W. C. DENISON. 1977. A 400-year-old Douglas-fir tree and its epiphytes: biomass,surface area, and their distributions. Canadian Journalof Forest Research 7: 680–699.

RHOADES, F. M. 1981. Biomass of epiphytic lichens andbryophytes on Abies lasiocarpa on Mt. Baker lavaflow, Washington. THE BRYOLOGIST 84: 39–47.

ROSENTRETER, R. 1995. Lichen diversity in managed for-ests of the Pacific Northwest, USA, pp. 103–124. InC. Scheidegger, P. A. Wolseley & G. Thor (eds.), Con-servation Biology of Lichenized Fungi. Herausgeber,Birmensdorf, Germany.

RUCHTY, A., A. L. ROSSO & B. MCCUNE. 2001. Changesin epiphyte communities as the shrub, Acer circina-tum, develops and ages. THE BRYOLOGIST 104: 274–281.

SILLETT, S. C. 1995. Branch epiphyte assemblages in theforest interior and on the clearcut edge of a 700-year


old forest canopy in Western Oregon. THE BRYOLOGIST

98: 301–312.———. 1999. Tree crown structure and vascular epiphyte

distribution in Sequoia sempervirens rain forest can-opies. Selbyana 20: 76–97.

——— & M. G. BAILEY. 2003. Effects of tree crownstructure on biomass of the epiphytic fern Polypodiumscouleri in redwood forests. American Journal of Bot-any 90: 255–261.

——— & T. GOWARD. 1998. Ecology and conservationof Pseudocyphellaria rainierensis, a Pacific Northwestendemic lichen, pp. 377–388. In M. G. Glenn, R. C.Harris, R. Dirig & M. S. Cole (eds.), LichenographiaThomsoniana: North American Lichenology in Honorof John W. Thomson. Myxcotaxon Ltd., Ithaca, NY.

——— & B. MCCUNE. 1998. Survival and growth of li-chen transplants in Douglas-fir forest canopies. THE

BRYOLOGIST 101: 20–31.——— & P. N. NEITLICH. 1996. Emerging themes in epi-

phyte research in westside forests with special refer-ence to cyanolichens. Northwest Science 70: 54–60.

——— & T. R. RAMBO. 2000. Vertical distribution ofdominant epiphytes in Douglas-fir forests of the cen-tral Oregon Cascades. Northwest Science 74: 44–49.

———, B. MCCUNE, J. E. PECK & T. R. RAMBO. 2000a.

Four years of epiphyte colonization in Douglas-fir for-est canopies. THE BRYOLOGIST 103: 661–669.

———, B. MCCUNE, J. E. PECK, T. R. RAMBO & A. RUCH-TY. 2000b. Dispersal limitations of epiphytic lichensresult in species dependent on old-growth forests. Eco-logical Applications 10: 789–799.

SOKAL, R. R. & F. J. ROHLF. 1995. Biometry, 3rd ed. W.H. Freeman and Company, NY.

STONE, D. F. 1989. Epiphyte succession on Quercus gar-ryana branches in the Willamette Valley of westernOregon. THE BRYOLOGIST 92: 81–94.

STOTLER, R. & B. CRANDALL-STOTLER. 1977. A checklistof the liverworts and hornworts of North America.THE BRYOLOGIST 80: 405–428.

VAN PELT, R. 2001. Forest Giants of the Pacific Coast.Global Forest Society and University of WashingtonPress, Seattle, WA.

VENEKLAAS, E. J., R. J. ZAGT, A. VAN LEERDAM, R. VAN

EK, A. J. BROEKHOVEN & M. VAN GENDEREN. 1990.Hydrological properties of the epiphytic mass of amontane tropical rain forest, Columbia. Vegetatio 89:183–192.

ms. received March 19, 2002; accepted Jan. 8, 2003.

william j.t. ellyson and stephen c. sillett - save the redwoods league

Documents