49

Leaf area differences associated with old-growth forest communitiesin the western Oregon Cascades'

HENRY L. GHOLZ, FRANKLIN K. FITZ, AND R. H. WARING

School of Forestry, Oregon State University, Corvallis, Oregon 97331Received July 2,1975'

Accepted December 5,1975

GHOLZ, H. L., F. K. FITZ, and R. H. WARING. 1976. Leaf area differences associatedwith old-growth forest communities in the western Oregon Cascades. Can. J. For. Res.6: 49-57.

Total leaf area varied from 20 to 42 m'/m' in 250- to 450-year-old forest communitiesdeveloped under different temperature and moisture conditions. The largest values werein communities at midelevations where winter snowpack accumulated and growing-seasontemperatures were cool. Shrub and herb leaf area varied from 3% to 14% of the total.Equations for converting from foliage biomass to surface area are included for mostspecies encountered.

GHOLZ, H. L., F. K. FITZ et R. H. WARING. 1976. Leaf area differences associatedwith old-growth forest communities in the western Oregon Cascades. Can. J. For. Res.6: 49-57.

La surface foliaire totale variait de 20-42 m'/m' dans des groupements forestiers agesde 250-450 ans formes sous diverses conditions de temperature et d'humidite. Les valeursles plus fortes ont ete observees dans les peuplements situes a mi-elevation, oil Ia neiges'accumule en hiver et ou les temperatures de la saison de croissance sont fraiches. Lasurface foliaire des arbustes et des herbes variait de 3% a 4% de la surface foliaire totale.L'article fournit des equations de conversion de la biomasse de feuillage a la surfacefoliaire pour Ia plupart des espêces qui s'y rencontrent.

[Traduit par le journal]

Introduction

Leaf area and its spatial distribution is essen-tial for estimating photosynthesis, transpiration,respiration, canopy interception, and energytransmission to the ground. Thus, it is a basicstructural feature required in all terrestrial eco-system studies.

This study had two major objectives: (1) todevelop techniques enabling one to assessfoliage areas for a wide variety of species char-acteristic of forest communities in the CascadeMountains of western Oregon, and (2) toevaluate whether leaf area differences amongmajor vegetational units are associated withmeasured moisture and temperature gradients.

'The work reported in this paper was supported byNational Science Foundation Grant No. GB-36810Xto the Coniferous Forest Biome, U.S. Analysis ofEcosystems, International Biological Program. This isContribution No. 183 from the Coniferous ForestBiome and Paper 1004 of the Forest Research Labora-tory, School of Forestry, Oregon State University.

'Revised manuscript received October 22,1975.

Study AreasThe study was confined to the H. J. Andrews

Experimental Forest, a 6000-ha watershed 60km east of Eugene, Oregon (44° N, 122° W).Dyrness et al. (1974) recognized 23 matureforest communities there, and Zobel et al. (1974)described their ecological distributions in termsof measured gradients of plant moisture stress(Waring and Cleary 1967) and temperature(Cleary and Waring 1969) (Fig. 1). 'Referencestands' (RS) representing the major communi-ties were established for detailed analysis(Zobel et al. 1974). To assess possible differ-ences in leaf area, six reference stands repre-senting a broad spectrum of environments wereselected. Descriptions of these stands are givenin Table 1 and their locations in environmentalcoordinates are designated in Fig. 1. In allstands 250- to 450-year-old Douglas fir (Pseu-dotsuga menziesii (Mirb) Franco var. men-ziesii) was the major overstory tree, constitut-ing from 55 to 98% of the total tree foliagebiomass.

w0z

100

80

60

40

0 10 20 30

PLANT MOISTURE STRESS, BARSFIG. I. Distribution of plant communities in the

H. J. Andrews Experimental Forest (after Zobel et a!.1974). Moisture stress gradient represents predawnmeasurements when soil water reserves were minimalfor the year. The temperature gradient is a physiologi-cal assessment of daily air and soil temperaturesthroughout the growing season for Douglas fir. Num-bers refer to reference stands where detailed analysisof vegetation and environment were made.

Methods

Estimating Foliar Biomass (see Tables 2 and 3)Because most foliage data from coniferous forests

are reported as biomass, one must first use mensura-

50 CAN. J. FOR. RES. 6, 1976

TABLE 1. Descriptions of six stands (RS) studied

RS Community Elevation, m Description

1 Pseudotsuga menziesii / Holodiscus discolor 490 Hot, dry; Douglas-fir climax(PSME/HODI)

2 Tsuga heterophylla / Rhododendronmacrophyllum / Berberis nervosa

490 Modal; hemlock climax

(TSHE/RHMA/BENE)4 Abies amabilis / Tiarella unifolia 1310 Subalpine; silver-fir climax

(ABAM/TIUN)5 Tsuga heterophylla Abies

amabilis / Rhododendronmacrophyllum / Berberis nervosa

885 Midelevation; hemlock / silver-fir climax

(TSHE—ABAM/RHMA/BENE)6 Tsuga heterophylla / Castanopsis

chrysophylla (TSHE/CACH)610 Dry; hemlock climax

7 Tsuga heterophylla / Polystichummunitum / Oxalis oregano

490 Very wet; hemlock climax

(TSHE/POMU/OXOR)

TABLE 2. Regression equation constants for foliagebiomass of major tree speciesa

Species A B R2

P. menziesii —3.890 1.890 0.88T. heterophylla —4.195 2.119 0.96A. amabilis —5.480 2.380 0.97A. procera —4.990 2.190 0.99T. plicata 2.0997 1.915 0.99C. chrysophylla —3.0258 1.607 0.79

°All equations are of the form In(dry foliage wt.) = A+ Bln(DBH). Dry foliage wt. is in kilograms, DBH is in centimetres.

tional techniques to estimate biomass and then convertto leaf area.

Grier (unpublished) furnished foliage biomassequations for Douglas fir, golden chinkapin (Castemop-sis chrysophylla (Doug'.) A. DC.), and noble fir(Abies procera Rehd.). The equation for western hem-lock (Tsuga heterophylla (Raf.) Sarg.) was derivedfrom a composite of data from Krumlik (1974) andFujimori (1971). The equation for silver fir (Abiesamabilis (Doug). Forbes) was derived from a com-posite of data from Krumlik (1974) and Grier (unpub-lished). Foliage biomass for western red cedar (Thujaplicara Donn.) was calculated using a commonlogarithm equation for MO occidentalis L. (Reiners1972). Foliage biomass for mountain hemlock (Tsugarnertensiana (Bong.) Carr.), a minor overstory conifer,and Pacific yew (Taxus brevifolia Nutt.), a minorunderstory conifer, was estimated using the equationfor western hemlock. For sugar pine (Pinus lambertionaDoug'.), foliage biomass was estimated from theDouglas fir equation.

For the seven most abundant shrubs, methods weresought that would allow rapid, objective, and accurateestimates of foliage biomass. Easily measurable inde-

AcerBerbe.Rho&Gault'XeropPolyst

eAll&Fro

pendtwerewas cEachusingham24 hto thiapprtweigl

Se.on coforminvesfoliodiamlcngtvide;anal!thenisticsin atotalEstirMaxformAudgreatapp'priat(Vacscattplanrangtiveltion!estirencc

Tfforcent.

wasThiscovecasetribistor

GHOLZ ET AL.

TABLE 3. Understory foliage biomass equations

51

Species Sample size

Acer circinatumb 132Berberis nervosa 32Rhododendron macrophyllum 40Gaultheria shallon 32Xerophyllum tenax 22Polystichum munitwn 41Oxalis oregana 10

Equations

X = 9 .03(D2L„,a,)1X = 14.218 +1.984(%cover)In(X) = 0.067177 + 0.60981 ln(D2L.„)ln(X) = 1.5137 + 0.70263 In(c7ccover)X = 18.873 0.02798(D2LayidX = - 2 . 5695 0.06429(L„, X no. fronds)X = 0.4625(%cover)

R2

0.900.800.900.830.940.90

Sy = 2.2°All equations are significant at the 0.99 level. D, diameter at litter surface (cm): L. length (cm, except for Acer m).bFrom Russel (1973).

;e

R2

►.88.96.97.99.99.79

4- Btees.

iert

lassop-

fir:ro-vedand5ies)m-ub-.uja.lontersugafer,norion77-la

the

•ere'ate;de-

pendent variables, thought to reflect foliage biomass,were selected for each species. Destructive samplingwas conducted in the neighborhood of RS-5 and RS-7.Each plant was measured, or cover was estimated, byusing a 1-m 2 sampling frame, and the total foliage washarvested. Excised foliage was placed in ovens within24 h of harvest, dried for 24 h at 60 °C, and weighedto the nearest gram. Finally, data were plotted and anappropriate function was selected relating foliageweight to the given measurement(s).

Several other species were thinly scattered or presenton one or two reference stands. Where similar growthforms and sizes were exhibited over the range of sitesinvestigated, an 'average-tree' sampling procedure wasfollowed (Baskerville 1965). This involved measuringdiameter at the litter surface (D) and average stemlength (L) and calculating a mean D 2L for all indi-viduals of a given species recorded in the vegetationalanalysis of a reference stand. Destructive analysis wasthen performed on selected plants that had character-istics of the mean. The sum of a species foliage biomassin a particular stand was estimated by multiplying thetotal number recorded by the mean foliage weight.Estimates for ocean spray (Holodiscus discolor (Pursh.)Maxim.), hazel (Corylus cornuta Marsh. var. call-fornica (D.C.) Sharp), and dogwood (Cornus nuttalliiAud. ex T. & G.) were made in this manner. Wheregreater variation in growth form and size occurred,application of the 'average-tree' method was inappro-priate. This was the case with red huckleberry(Vaccinium parvifolium Smith), which was foundscattered through two stands. For this species severalplants adjacent to each reference stand covering therange of size classes present in the stand were destruc-tively sampled. From this limited sample, D 2L rela-tionships with leaf weight were roughly established toestimate foliage weights for this species whereverencountered.

To reduce survey time, foliage biomass equationsfor herb species were not developed. Instead, herba-ceous cover was assumed to be directly related tofoliage surface area so that 10% cover on a 1-m 2 plotwas equivalent to a one-sided surface area of 0.1 m2.This assumption appeared invalid only at very highcover values, not commonly encountered. Only in thecase of Oxalis oregana Nutt. ex T. & G., which con-tributed a large amount of surface area to the under-story on RS-7 and similar wet sites, was a biomass

equation developed (Table 3). Equations relating per-centage cover to total aboveground biomass for almostall common herbs in the study area exist in Russel(1973).

Converting Biomass to Leaf AreaTo estimate leaf area from foliage biomass, a

coefficient for each species was required. To accomplishthis for needle-leaf species (Table 4), samples werecollected from midcrown, with a shotgun wherenecessary, and surface areas on fresh samples weredetermined using an optical planimeter (Miller et al.1956; Geppert 1968). Samples were dried for 24 h at60 °C, then weighed to the nearest 0.1 mg. Finally, allsurface areas were corrected for the three-dimensionalcharacter of the needles. This was done by projectinga microscope slide of the needle cross section ontopaper, then measuring the ratio between the circum-ference and projected diameter (Drew and Running1975). From 20 to 25 needles were used for eachcorrection and cross sections were cut from the mid-point of each needle. The red-cedar conversion wasadjusted for cross section and leaf-bearing twig weightby multiplying by I.S. In the case of mountain hemlock(Tsuga mertensiana), the Tsuga Izeterophylla coefficientwas applied.

Surface areas for broadleaves were obtained bydetermining similar coefficients (Table 5). Discs of aknown diameter were punched from leaves by usingvarious sizes of cork borers. Leaves were picked froma random selection of plants in each reference stand.For the long and narrow leaves of beargrass (Xerophyl-lum tenax (Pursh.) Nutt.) numerous 0.55-cm-diameterdiscs were punched for about three quarters of thelength of each sampled blade. One hundred discs foreach species represented on the stand were dried for24 h at 60 °C and weighed in groups of 10 to thenearest 0.1 rng. All broadleaf surface area figures formajor species were corrected for petiolar weight.

Estimating Total Leaf Area from VegetationalAnalyses

The vegetation on each reference stand was non-destructively surveyed in late June and July, 1974.This was accomplished for the shrub and herb layersby laying out twenty 4-rn 2 plots, composed of four 1-m2subplots on the inner 30 m x 30 m of the stand. A10-m-wide buffer strip along each side of the RS was

tax

52 CAN. J. FOR. RES. 6. 1976

ooCC1C.)

00 v:::)CT ■- ■-• 0

Ill VD r- C0

0D 0,Tr vn

enrn

r-cD

co)

a

cd

oror

8

O

0

z

2

le) C,J R.• .eV r- r- Ni

N oo en rlrn OT ro4 ot, C'

vD rt

.••••• .•••N Ni

rn r- •vDrn •-• • en vn

ev CT el --

••■•, ••■••• ■••

r, t"V 0 00 0r-•

to Ir._ r,„ n r Ni VD 0,,p■.1 .■■••• ■•••••

.•■■ ..... /... .•••■ ,....,zr ci• ---. eV ....-. 0, 0000

CA CD • rn en CA

e•si f'l ■0 r•I WI .... .....

,.... ......, ,.., .-., ,.... ,... ,...•

.... rn oo •-• en eet 1-- 0• • •eg 71- CD IT .0 Ni vo c)

CD cD 0, cD 00 r-- 00 •••-•

eV e'l •-• eV •-• •-• ■-• CV

•••••■ •••••• .•••••■

r- cr, rn rl V .n VO .o CVC 0 •

rl r- rA vD oo cr, CD Ni • • ken

N N N N N Nry r- vD

••••., •••■•• ••■••• ••••/ •••■■•

r- r- rn r- vD rn vn r- a CD

-T r- 7r vn rn rn rn ry

0, r- VD r- ken rn r, O N rn vnev cv CV --

-400

es40 0d6 (0

't) rn oo

N oo r"- ••••-•O 0 . . . .

0■■•

—

60

U

0)

7U,

.0 I... to ‘...C C 2 a 11'

, ., , - . CA ...-• S= Cn

1.)

;el :1 eel ti u. 3., v tio > . 0 > c..) ° . °> or >

U -?• 6 . , U! d < .'C.) •-• 6 ..—.. c 4.•—., v.,— —::..4 .••..-.•.-, .,..•• ..••••.

1--- 0 3 rA 2 7. v) v) ri) v) V)0 2 ..'"' v, 2to 0 0 ,, to 0 ,.., — — c4 ne c4 ce c40.) •-• el.) .•-• .... ct .L... ...." t j ..., U ,.. ,...., ,..., u , ...-.. -...... ...-,c) < ,= 0 < _14C 0 < < -.2 < « «..., ,..,

C —, cz • —, -, -Z=Ou = U Z==3

.-, -, — -, -,= = = Z X

sz

**>*,O

dC

0

,r

0eJ

GHOLZ ET AL. 53

TABLE 5. Surface area : weight conversions for common shrubs andherbs in the H. J. Andrews Experimental Forest

Species RS Cm2/ga %Petiole

Acer circinaturn 1 603.6 (53.8)5 737.8 (26.4)6 438.4 (50.6) 6.28 (1.05)7 688.2 (40.0)

Berberis nervosa 1 88.4 (4.4)5 176.2 (7.8)6 139.6 (2.8) 23.17 (1.91)7 171.2 (5.4)

Rhododendron macrophyllum 5 261.2 (25.2)6 166.8 (12.6) 5.40(1.08)

Gaultheria shallon 5 228.0 (14.6)6 166.4 (12.8) 2.58 (0.76)7 269.4 (18.8)

Xerophyllum tenax 4 88.4 (6.0)5 78.0 (4.4) 0.006 77.8 (2.0)

Polystichurn munitum 7 304.6 (6.6) 38.39 (3.90)Oxalis oregana 7 1193.6 (51.2) 41.77 (2.80)Castanopsis chrysophylla 6 140.4 (15.0) 2.73 (0.87)Corylus cornuta calif 424.2 (28.8) N.A.bHolodiscus discolor 420.6 (28.4) N.A.Cornus canadensis 4 245.8 (20.8) N.A.Vaccinium parvifolium 7 514.0 (15.2) N.A.Achyls triphylla 4 856.4 (38.4)

7 798.0 (78.2) N.A.Linnaea borealis 7 561.2 (9.4) N.A.Chimaphila umbellata 2 252.4 (9.8)

7 213.0 (8. 8) N.A.Cornus nuttallii 6 350.6 (33.4) 3.12 (0.35)Coptis laciniata 2 599.0 (19.6) N.A.

NorE: Standard deviations are shown in parentheses.°A!1 values are for two sidesbN.A., not available

not surveyed because of frequent disturbances there.Appropriate variables were measured for each speciesencountered. Cover for Oxalis, Oregon grape (Berberisnervosa Pursh.), salal (Gaultheria shallon Pursh.),and the herb layer was estimated on each subplot andan average was computed for the entire plot. Foliagebiomass for tree species was computed for the full50 m x 50 m area from diameter data collected byInternational Biological Program (IBP) personnel.'These data were used with the regressions of leaf dryweight on diameter at breast height (DBH) to com-pute species foliage biomass. Coefficients were applied

to foliage dry weights to determine leaf area by speciesin square metres per square metre. Values for individualspecies were summed to obtain totals for each refer-ence stand.

ResultsThe total foliage biomass, surface area, and

percentage surface area in the shrub and herblayers for each of the six stands are presentedin Fig. 2. The location of the values in Fig. 2corresponds to the environmental location ofeach reference stand as presented in Fig. I .

The coniferous needle surface area coeffi-

'Unpublished data on file at the Coniferous ForestBiome Data Bank, Oregon State University, ForestResearch Laboratory.

DOUGLAS-FtCLIMAX

MOISTHEMLOCK

/ 32.45 \ it 1428 1(18.2)

(15.31 13%5%

120.2820.78 II .7)

'(13.3) %a%

42.1(24.8)

8%

SILVER-FIRZONE

TRANSITIONZONE

33.13(19.9)

3%I I i I

RYH E M LOC K

ODAL

0 10 20 30PLANT MOISTURE STRESS, BARS

FIG. 2. Total foliage biomass (mt/ha, parentheticalvalues), total leaf area (m 2 /m 2, upper values), andpercentace of the total surface area in the shrubs andherbs (lower values) for the six reference standssuperimposed on Fig. 1. Projected leaf areas, one sidewith no cross-section corrections, are as follows: RS-1,8.7 m 2 /m 2 ; RS-2, 14.3; RS-4, 14.4; RS-5, 18.6; RS-6,9.1; RS-7, 12.0. Sugar pine (Pinus lambertiana Dougl.)would add about 0.4 mt/ha to RS-1 and 0.9 mt/ha toRS-6, but it is not included here because there is nosurface area conversion available.

cients (Table 4) show that surface area pergram of needle decreased substantially withincreasing age, although not all age class valuesare statistically different. In old-growth Douglasfir, where about 20% of the total foliage wascurrent, 15% 1 year old, 15% 2 years old, and50% 3 years or older, the surface area decreasedfrom about 204 cm2/g for the current foliageto 149 cm 2/g for the 3rd year and older foliage(statistically different at the 0.95 level). Thissuggests that substantial errors may result fromnot treating age classes separately. However,because of inadequate data, only a few speciesother than Douglas fir had leaf areas computedby age class.

Also, silver and noble fir growing in theunderstory had a greater surface area per gramthan those in the overstory (Table 4). Weexpect that this is also true for western hemlockand other species, which suggests in additionthat a variety of coefficients may be obtained

100

z

1- 800cr91

60I-cr

a.

40

54 CAN. J. FOR. RES. 6, 1976

from various areas of the crowns of the largeoverstory trees.

Broad- leaved species illustrate a greater rangeof coefficients than the conifers (Table 5). Forexample, beargrass had 77 cm 2/g„, whereasOxalis had 1194 cm2/g. Likewise, there is awide range of values within a single species. Forexample, vine maple (Acer circinatum Pursh.)ranged from 438 cm 2/g to 738 cm2/s, and salalfrom 166 cm 2 /g to 269 cm2/g.

Discussion

In observing forests with different leaf areasone may question whether these differences areassociated with environmental adaptations orare merely changes associated with stand devel-opment. There is general support for rapidincrease in leaf area from the time of establish-.ment until crown closure. This period may be asshort as 4 years in eastern deciduous forests(Marks 1974) or may extend beyond a decadeas reported in a white pine (Pinus strobus L.)plantation by Swank and Schreuder (1973).After closure the amounts of foliage may returnto a somewhat lower plateau (Kira and Shidei1967; Ovington 1957).

The reported peaks are, however, not alwaysconvincing and may be offset by unreportedincreases in understory foliage development ofolder forests. We can probably assume, barringmajor windstorms, fire, or other recent acutedisturbances, that old-growth forest communi-ties do exhibit equilibrium leaf areas. Whetherthis is greater or less than that at the time ofcrown closure awaits confirmation.

In this study the magnitude of the leaf areasestimated for midelevation stands was higherthan those associated with cooler or warmerenvironments with similar moisture regimes(Fig. 2). In fact, the value 42.1 m2/m2 (pro-jected leaf area of 18.6) for reference stand 5is near the maximum ever reported. Westmanand Whittaker (1975) recently reported asimilar value for a mature coast redwood forestin California. In a 26-year-old hemlock foreston the Oregon Coast, Fujimori (1971) esti-mated foliage at 21.1 mt/ha which converts to37 m2 /m 2 of total leaf area using the appro-priate coefficient from Table 4.

What permits such a large accumulation offoliage to develop in an area of summer drought?

GHOLZ ET AL. 55

In the humid deciduous forests of the easternUnited States, total leaf areas above 12 m2/m2are uncommon (Whittaker 1966). Even in atropical rain forest of Thailand an exceptionaltotal leaf area of 32 m 2/m2 was reported, withmost plots supporting less than 24 m2/m2(Ogawa et al. 1965 ).

From the moisture gradient defined directlyby the stress experienced by understory conifers(Fig. 2), it is clear that soil water is not verylimiting to the forest communities where thehighest foliage areas accumulate.

Apparently, temperature is the variable whichpermits large accumulations of foliage, butneither the warmest sites nor those with thelongest growing season support the largestamount of foliage. Rather, we hypothesize thatit is a special balance of temperatures that favorsmodest respiration rates but yet permits, evenduring the dormant season, considerable netphotosynthesis. In the redwood forests of Cali-fornia, the Coast Range forests of the PacificNorthwest, and the midelevation forests of thewestern Cascades, the winter temperatures areusually above freezing, at least during the day(Waring and Major 1964; Zobel et al. 1973 ).Where there is a snowpack, day temperaturesusually remain above —2 'C so the watercolumn in the stem is not often frozen (Zimmer-man 1964) and both transpiration and photo-synthesis can occur. During the summer,coastal fog, maritime air, or elevational-relatedcooling create mild growing-season tempera-tures around 10 c C (Waring and Major 1964;Zobel et al. 1973). For conifers these tempera-tures permit very respectable rates of photo-synthesis (Walker et al. 1972).

This proposed explanation could alsoaccount for the decrease in leaf area in wellwatered but warmer habitats where respirationincreases substantially (RS-2 and RS-7). Athigher elevations colder winter temperaturesand a shorter growing season notably limitphotosynthesis (Emmingham 1974) and to-gether with snow damage could explain thelower value found there.

There are some special problems associatedwith estimating leaf area which also meritdiscussion. First of all is an inconsistency inexpressing results. Usually data from broadleafspecies have been reported as representing onlya single surface. This is not a major problem if

the convention is clearly noted by the authors.We have freely converted many of the valuesreported by Whittaker (1966), Westman andWhittaker (1975 ), and Ogawa et al. (1965) tototal surface area. There is real difficulty,however, when data for needle-leaf species arereported inconsistently on both a projected andtotal surface area basis. The problem becomesacute when both broadleaf and needle-leafspecies occurring in the same stand are reporteddifferently. We believe there is need to stand-ardize, and because stomata and energy ex-change can occur on more than one leaf surface,we feel total surface area is the most desirableexpression. From recent studies comes furthersupport that interception, photosynthesis,respiration, and possibly even foliar nutrientconcentrations are biologically best interpretedin such context (Linder 1974; Sollins et al.1974; Gholz, unpublished).

A second problem in estimating surface areais converting from biomass data. For example,in Table 5, differences in species coefficientsrange from 77 cm2/g to 1194 cm2 /g, withfurther variation noted within a species, e.g.,Acer circinatum. Recent unpublished studies byGholz suggest much of the within-species varia-tion is a response to development under differentlight regimes. Light has similarly been attributedas the causal factor in explaining leaf area–weight differences in European beech (Kiraet al. 1969) and a variety of oak–hickory forestspecies (Monk et al. 1970). These follow earlierclassical experimental work (e.g., Blackman andRutter 1948).

Additionally, there are differences in leafarea–weight coefficients as leaves age. This pointwas made earlier in discussing Table 4. For bothproblems, development of separate regressionequations for individual stands is a possible but,as we found, laborious solution.

Where possible, we recommend a direct esti-mation of leaf area, bypassing the difficulties ofconverting from foliar biomass. This appearspossible from simple linear relationships, suchas recently discovered between cross-sectionalarea of functional conducting tissue in the sternand leaf weight or area (Grier and Waring 1974;Dixon 1971 ). More recently, Waring et al.(unpublished) found a similar linear coefficientfor shrub species Rhododendron macrophyllum.

These new coefficients may solve another

ge

age'orEtas

aF.-or;h.)alai

.reasare

s orevel--apid

be as► rests.cacier L.)1 73 ).' turnhidei

ways)rtednt ofrringtcuteiuni-Aherle of

areasigherrmerimespro-nd 5.man:d acrestDrestesti-ts toTro-

ll ofight?

56 CAN. I. FOR. RES. 6, 1976

problem common to estimating foliar area oftrees. Ogawa et al. (1965 ), Kira and Shidei(1967), and Shinozaki et al. (1964) reportedthat estimates of foliage weight from measure-ments of stem diameter were often overesti-mates, especially for larger trees. Some of thismay be a result of competition among trees(Satoo 1966). This could also occur by extrap-olating from small trees which contain onlyconducting tissue to large trees with non-conducting heartwood. In this study, regressionequations were developed from data includinglarge trees. Still, because of the log relationships,a small change in stem diameter suggests a fairlylarge change in leaf weight and area. Directmeasurement of linear coefficients could resultin a considerable gain in accuracy.

Increasingly, accurate estimates of leaf areaare necessary to test ecological hypotheses con-cerning the functional recovery of ecosystems.How long recovery takes is often a function ofreestablishment of leaf area, sometimes by shrubrather than tree species. As the importance ofunderstory species in nutrient cycling has be-come better appreciated, accurate estimates oftheir contribution to uptake, storage, and litterfall are needed. With the development ofaccurate methods, we hope quantitative esti-mates of leaf area will become more commonlyreported in vegetational analyses.

In summary, when the total leaf areas of sixold-growth forest communities in westernOregon were estimated, differences were ob-served which were attributed mainly to varia-tions in the physical environment. The highestvalues may represent maximum accumulationand reflect particular combinations of environ-ment that permit adequate photosynthesis withmodest respiration in maritime-influenced cooltemperate forests of the Pacific Northwest.Methods developed in this paper will enableecologists to explore further the foliage rela-tions of both disturbed and undisturbed plantcommunities.

Acknowledgements

We thank Dr. W. H. Emmingham, Dr. C. C.Grier, and S. W. Running, of the ConiferousForest Biome, Oregon State University, for helpin the preparation and review of the manuscript.

B ASKERVILLE, G. L. 1965. Estimation of dry weight oftree components and total standing crop in coniferstands. Ecology, 46(6): 867-869.

B LACKMAN, G. E., and A. J. RurrER. 194.8. Physiologi-cal and ecological studies in the analysis of plantenvironments. III. The interaction between lightintensity and mineral nutrient supply in leaf devel-opment and in the net assimilation rate of theBluebell (Scilla non-scripta). Ann. Bot (London),N.S. 11(45): 1-26.

CLEARY, B. D., and R. H. WARING. 1969. Temperature:collection of data and its analysis for the interpreta-tion of plant growth and distribution. Can. J. Bot.47: 167-173.

DIXON, A. F. G. 1971. The role of aphids in woodformation. J. Appl. Ecol. 8: 165-179.

DREW, A. P., and S. W. RUNNING. 1975. A comparisonof two techniques for measuring surface area ofconifer needles. For. Sci. 21(3): 231-233.

D YRNESS, C. T., J. F. FRANKLIN, and W. H. Mont. 1974.A preliminary classification of forest communitiesin the central portion of the Western Cascades inOregon. Bull. No. 4, Conif. For. Biome, USIBP,Univ. Wash., Seattle, Washington.

E MMINGHAM, W. H. 1974. Physiological responses offour Douglas-fir populations in three contrastingfield environments. Ph.D. Thesis, Oregon State Uni-versity, Corvallis, Oregon.

FUJIMORI, T. 1971. Primary productivity of a youngTsuga heterophylla stand and some speculationsabout biomass of forest communities on the Oregoncoast. U.S. Dep. Agric. For. Ser. Res. Pap. PNW-123.

G EPPERT, R. R. 1968. The effect of age on transpirationrates of Douglas-fir seedlings. M.Sc. Thesis, OregonState Univ., Corvallis, Oregon.

G RIER, C. C., and R. H. WARING. 1974. Conifer foliagemass related to sapwood area. For. Sci. 20(3):205-206.

K IRA, T., and T. SHIDEI. 1967. Primary production andturnover of organic matter in different forest eco-systems of the Western Pacific. Jap. J. Ecol. 17(2):70-87.

K IRA, T., K. SHINOZAKI, and K. Hozumi. 1969. Struc-ture of forest canopies as related to their primaryproductivity. Plant Cell Physiol. 10: 129-142.

KRUEGER, K. W., and R. H. Rum!. 1969. Comparativephotosynthesis of red alder, Douglas fir, Sitka spruceand western hemlock seedlings. Can. J. Bot. 47:519-527.

KRUMLIK, J. G. 1974. Biomass and nutrient distribu-tion in two old-growth forest ecosystems in S.Coastal British Columbia. M.S. Thesis, Univ. B.C.,Vancouver, British Columbia.

L INDER, S. 1974. A proposal for the use of standardizedmethods for chlorophyll determinations in ecologicaland eco-physiological investigations. Physiol. Plant.32: 154-156.

M ARKS, P. W. 1974. The role of pin cherry (Prunuspensylvanica L.) in the maintenance of stability innorthern hardwood ecosystems. Ecol. Monogr.44(1): 73-88.

GHOLZ ET AL. 57

MILLER., E. E., C. A. SHADBOLT, and L. HOLM. 1956.Use of an optical planimeter for measuring leafarea. Plant Physiol. 31(6) : 484-487.

MONK, C. D., G. I. CHILDS, and S. A. N ICHOLSON. 1970.Biomass, litter and leaf surface area estimates of anoak-hickory forest. Oikos, 21: 138-141.

OGAWA, H., K. YODA, K. OGINO, and T. KIRA. 1965.Comparative ecological studies on three main typesof forest vegetation in Thailand. II. Plant biomass.In Nature and life in S.E. Asia. Vol. IV. Edited byT. Kira and K. Iwata. Japan Soc. Prom. Sci., Tokyo.pp. 49-80.

OVINGTON, J. D. 1957. Dry-matter production by Pinussylvestris L. Ann. Bot. (London), N.S. 21(82):287-314.

REINERS, W. A. 1972. Structure and energetics of threeMinnesota forests. Ecol. Monogr. 42(1) : 71-94.

RUSSELL, D. W. 1973. The life history of vine mapleon the H. J. Andrews Experimental Forest. M.S.Thesis, Oregon State University, Corvallis, Oregon.

SATOO, T. 1966. Production and distribution of drymatter in forest ecosystems. Misc. Info. Tokyo Univ.For. 16: 1-15.

SHINOZAKI, K., K. YODA, K. Hozumt, and T. KmA.1964. A quantitative analysis of plant form-thepipe model theory. II. Further evidence of the theoryand its application in forest ecology. Jap. J. Ecol.14(4): 133-139.

SOLLINS, P.. R. H. WARING, and D. W. COLE. 1974. Asystematic framework for modeling and studyingthe physiology of a coniferous forest ecosystem. InIntegrated research in the coniferous biome. Editedby R. H. Waring and R. L. Edmonds. Bull. No. 5,Conif. For. Biome, USIBP, Univ. Wash., Seattle,Washington.

SWANK, W. T., and H. T. S CHREUDER. 1973. Temporalchanges in biomass, surface area, and net productionfor a Pinus strobus L. forest. In IUFRO biomassstudies. Working Party on the Mensuration of the

Forest Biomass. S4.01. Mensuration, Growth, andYield. Coll. Life Sci. Agric. Univ. of Maine, Orono,Maine.

WALKER, R. B., D. R. M. Scorr, D. J. SALO, and K. L.REED. 1972. Terrestrial process studies in conifers:a review. In Proceedings-research on coniferousforest ecosystems-a symposium. Edited by J. F.Franklin, L. J. Dempster, and R. H. Waring. Pac.NW For. Range Exp. Stn. U.S. Dep. Agric. For.Ser., Portland, Oregon.

WARING, R. H.. and B. D. CLEARY. 1967. Plant moisturestress: evaluation by pressure bomb. Science, 155:1248-1254.

WARING, R. H., and J. MAJOR. 1964. Some vegetationof the California coastal redwood region in relationto gradients of moisture, nutrients, light, and tem-perature. Ecol. Monogr. 34: 167-215.

WESTMAN, W. E., and R. H. WHITTAKER. 1975. Thepygmy forest region of Northern California: studieson biomass and primary productivity. J. Ecol. 63(2):493-520.

WrirrrAKER, R. H. 1966. Forest dimensions and produc-tion in the Great Smoky Mountains. Ecology. 47(1):103-121.

Z IMMERMANN, M. H. 1964. Effect of low temperatureon the ascent of sap in trees. Plant Physiol. 39: 568-572.

ZOBEL, D. B., W. A. McKEE, G. M. HAWK, and C. T.DYRNESS. 1973. Variation in air and soil tempera-tures in forest communities in the H. J. AndrewsExperimental Forest, 1970-1972. Int. Rep. 127,Conif. For. Biome, Ecosystem Analysis Studies,USIBP, Univ. Wash.. Seattle, Washington. 1974. Correlation of forest communities with

environment and phonology on the H. 1. AndrewsExperimental Forest, Oregon. In Integrated researchin the coniferous biome. Edited by R. H. Waringand R. L. Edmonds. Bull. No. 5, Conif. For. Biome.USIBP, Univ. Wash., Seattle, Washington.

so•-••••••••••••,,,«. t-m• . •

ofer

gi-Lnt

.el-then),

ire:eta-Bot.

rood

isona of

1974.nities/es inSIBP,

ses of.acting

Uni-

youngationstregonPNW-

rationregon

oliage1)(3):

andCCO-

7(2):

jtruc-ary

-ativepruCe. 47:

tribu-n S.B.C.,

dized^gical)1ant.

-anusty innogr.

49

Leaf area differences associated with old-growth forest communitiesin the western Oregon Cascades1

HENRY L. GHOLZ, FRANKLIN K. FITZ, AND R. H. WARING

School of Forestry, Oregon State University, Corvallis, Oregon 97331Received July 2, 1975'

Accepted December 5,1975

GHOLZ, H. L., F. K. Frrz, and R. H. WARING. 1976. Leaf area differences associatedwith old-growth forest communities in the western Oregon Cascades. Can. J. For. Res.6: 49-57.

Total leaf area varied from 20 to 42 m'/m' in 250- to 450-year-old forest communitiesdeveloped under different temperature and moisture conditions. The largest values werein communities at midelevations where winter snowpack accumulated and growing-seasontemperatures were cool. Shrub and herb leaf area varied from 3% to 14% of the total.Equations for converting from foliage biomass to surface area are included for mostspecies encountered.

GHOLZ, H. L., F. K. Frrz et R. H. WARING. 1976. Leaf area differences associatedwith old-growth forest communities in the western Oregon Cascades. Can. J. For. Res.6: 49-57.

La surface foliaire totale variait de 20-42 m'/m' dans des groupements forestiers agesde 250-450 ans formes sous diverses conditions de temperature et d'humidite. Les valeursles plus fortes ont ete observees dans les peuplements situes a mi-elevation, ou la neiges'accumule en hiver et ou les temperatures de la saison de croissance sont fraiches. Lasurface foliaire des arbustes et des herbes variait de 3% a 4% de la surface foliaire totale.L'article fournit des equations de conversion de la biomasse de feuillage a la surfacefoliaire pour la plupart des especes qui s'y rencontrent.

[Traduit par le journal]

Introduction

Leaf area and its spatial distribution is essen-tial for estimating photosynthesis, transpiration,respiration, canopy interception, and energytransmission to the ground. Thus, it is a basicstructural feature required in all terrestrial eco-system studies.

This study had two major objectives: (1) todevelop techniques enabling one to assessfoliage areas for a wide variety of species char-acteristic of forest communities in the CascadeMountains of western Oregon, and (2) toevaluate whether leaf area differences amongmajor vegetational units are associated withmeasured moisture and temperature gradients.

`The work reported in this paper was supported byNational Science Foundation Grant No. GB-36810Xto the Coniferous Forest Biome, U.S. Analysis ofEcosystems, International Biological Program. This isContribution No. 183 from the Coniferous ForestBiome and Paper 1004 of the Forest Research Labora-tory, School of Forestry, Oregon State University.

'Revised manuscript received October 22,1975.

Study AreasThe study was confined to the H. J. Andrews

Experimental Forest, a 6000-ha watershed 60km east of Eugene, Oregon (44° N, 122° W).Dyrness et al. (1974) recognized 23 matureforest communities there, and Zobel et al. (1974)described their ecological distributions in termsof measured gradients of plant moisture stress(Waring and Cleary 1967) and temperature(Cleary and Waring 1969) (Fig. 1). 'Referencestands' (RS) representing the major communi-ties were established for detailed analysis(Zobel et al. 1974). To assess possible differ-ences in leaf area, six reference stands repre-senting a broad spectrum of environments wereselected. Descriptions of these stands are givenin Table 1 and their locations in environmentalcoordinates are designated in Fig. 1. In allstands 250- to 450-year-old Douglas fir (Pseu-dotsuga menziesii (Mirb) Franco var. men-ziesii) was the major overstory tree, constitut-ing from 55 to 98% of the total tree foliagebiomass.

DOUGLAS-FIRCLIMAX100

z80

60

40

50 CAN. J. FOR. RES. 6, 1976

TABLE 1. Descriptions of six stands (RS) studied

RS Community Elevation, m Description

1 Pseudotsuga menziesii / Holodiscus discolor 490 Hot, dry; Douglas-fir climax(PSME/HODI)

2 Tsuga heterophylla / Rhododendronmacrophyllum / Berberis nervosa

490 Modal; hemlock climax

(TSHE/RHMA/BENE)4 Abies amabilis / Tiarella unifolia 1310 Subalpine; silver-fir climax

(ABAM/TIUN)5 Tsuga heterophylla — Abies

amabilis / Rhododendronmacrophyllum / Berberis nervosa

885 Midelevation; hemlock / silver-fir climax

(TSHE—ABAM/RHMA/BENE)6 Tsuga heterophylla / Castanopsis

chrysophylla (TSHE/CACH)610 Dry; hemlock climax

7 Tsuga heterophylla / Polystichummuniturn / Oxalis oregano

490 Very wet; hemlock climax

(TSHE/POMU/OXOR)

0

10

20 30

PLANT MOISTURE STRESS, BARSFro. 1. Distribution of plant communities in the

H. J. Andrews Experimental Forest (after Zobel et al.1974). Moisture stress gradient represents predawnmeasurements when soil water reserves were minimalfor the year. The temperature gradient is a physiologi-cal assessment of daily air and soil temperaturesthroughout the growing season for Douglas fir. Num-bers refer to reference stands where detailed analysisof vegetation and environment were made.

Methods

Estimating Foliar Biomass (see Tables 2 and 3)Because most foliage data from coniferous forests

are reported as biomass, one must first use mensura-

TABLE 2. Regression equation constants for foliagebiomass of major tree speciesa

Species A B R2

P. menziesii —3.890 1.890 0.88T. heterophylla —4.195 2.119 0.96A. amabilis —5.480 2.380 0.97A. procera —4.990 2.190 0.99T. plicata 2 .0997 1.915 0.99C. chrysophylla —3.0258 1.607 0.79

aAll equations are of the form ln(dry foliage wt.) == A+ Bin(DBH). Dry foliage wt. is in kilograms, DBH is in centimetres.

tional techniques to estimate biomass and then convertto leaf area.

Grier (unpublished) furnished foliage biomassequations for Douglas fir, golden chinkapin (Castanop-sis chrysophylla (Dougl.) A. DC.), and noble fir

procera Rehd.). The equation for western hem-lock (Tsuga heterophylla (Raf.) Sarg.) was derivedfrom a composite of data from Krumlik (1974) andFujimori (1971). The equation for silver fir (Abiesamabilis (Dougl. Forbes) was derived from a com-posite of data from Krumlik (1974) and Grier (unpub-lished). Foliage biomass for western red cedar (Thujaplicata Donn.) was calculated using a commonlogarithm equation for Thuja occidentalis L. (Reiners1972). Foliage biomass for mountain hemlock (Tsugamertensiana (Bong.) Carr.), a minor overstory conifer,and Pacific yew (Taxus brevifolia Nutt.), a minorunderstory conifer, was estimated using the equationfor western hemlock. For sugar pine (Pines lamberrianaDougl.), foliage biomass was estimated from theDouglas fir equation.

For the seven most abundant shrubs, methods weresought that would allow rapid, objective, and accurateestimates of foliage biomass. Easily measurable inde-

AcerBertnRhodGauliXeroiPolysOxali

aAtibFr,

pendwerewasEactusinihary24 hto ifapprweig

Scon cforainvefolicdianlengvideanalthenisticin atotaEstiMa)foreAucgrez.apppria(Vascatplatrangytivetioresti,enc

1forceofoliwa5.Thicoycas,tritstor

GHOLZ ET AL. 51

TABLE 3. Understory foliage biomass equations

Species Sample size Equationa R2

Acer circinatum4 132 X = 9.03(D2Lmax)i 0.90Berberis nervosa 32 X = 14.218+1.984(%cover) 0.80Rhododendron macrophyllurn 40 In(X) = 0.067177 + 0.60981 In(D2L.,) 0.90 Gaultheria shallon 32 ln( X) = 1.5137 0.70263 ln(7c,cover) 0.83Xerophyllum tenax 22 X = 18.873 + 0.02798(D2Layd 0.94Polystichum muniturn 41 X = -2.5695 + 0.06429(Lavg X no. fronds) 0.90Oxalis oregana 10 X = 0.4625(%cover) Syr = 2.2

aAll equations are significant at the 0.99 level. D, diameter at litter surface (cm); L, length (cm, except for Acer m).5From Russel (1973).

pendent variables, thought to reflect foliage biomass,were selected for each species. Destructive samplingwas conducted in the neighborhood of RS-5 and RS-7.Each plant was measured, or cover was estimated, byusing a 1-m 2 sampling frame, and the total foliage washarvested. Excised foliage was placed in ovens within24 h of harvest, dried for 24 h at 60 °C, and weighedto the nearest gram. Finally, data were plotted and anappropriate function was selected relating foliageweight to the given measurement(s).

Several other species were thinly scattered or presenton one or two reference stands. Where similar growthforms and sizes were exhibited over the range of sitesinvestigated, an 'average-tree' sampling procedure wasfollowed (Baskerville 1965). This involved measuringdiameter at the litter surface (D) and average stemlength (L) and calculating a mean D = L for all indi-viduals of a given species recorded in the vegetationalanalysis of a reference stand. Destructive analysis wasthen performed on selected plants that had character-istics of the mean. The sum of a species foliage biomassin a particular stand was estimated by multiplying thetotal number recorded by the mean foliage weight.Estimates for ocean spray (Holodiscus discolor (Pursh.)Maxim.), hazel (Corylus cornuta Marsh. var. cali-fornica (D.C.) Sharp), and dogwood (Cornus nuttalliiAud. ex T. & G.) were made in this manner. Wheregreater variation in growth form and size occurred,application of the 'average-tree' method was inappro-priate. This was the case with red huckleberry(Vaccinium parvifolium Smith), which was foundscattered through two stands. For this species severalplanes adjacent to each reference stand covering therange of size classes present in the stand were destruc-tively sampled. From this limited sample, D 21, rela-tionships with leaf weight were roughly established toestimate foliage weights for this species whereverencountered.

To reduce survey time, foliage biomass equationsfor herb species were not developed. Instead, herba-ceous cover was assumed to be directly related tofoliage surface area so that 10% cover on a 1-m = plotwas equivalent to a one-sided surface area of 0.1 m2.This assumption appeared invalid only at very highcover values, not commonly encountered. Only in thecase of Oxalis oregana Nutt. ex T. & G., which con-tributed a large amount of surface area to the under-story on RS-7 and similar wet sites, was a biomass

equation developed (Table 3). Equations relating per-centage cover to total aboveground biomass for almostall common herbs in the study area exist in Russel(1973).

Converting Biomass to Leaf AreaTo estimate leaf area from foliage biomass, a

coefficient for each species was required. To accomplishthis for needle-leaf species (Table 4), samples werecollected from midcrown, with a shotgun wherenecessary, and surface areas on fresh samples weredetermined using an optical planimeter (Miller et a!.1956; Geppert 1968). Samples were dried for 2413 at60 °C, then weighed to the nearest 0.1 mg. Finally, allsurface areas were corrected for the three-dimensionalcharacter of the needles. This was done by projectinga microscope slide of the needle cross section ontopaper, then measuring the ratio between the circum-ference and projected diameter (Drew and Running1975). From 20 to 25 needles were used for eachcorrection and cross sections were cut from the mid-point of each needle. The red-cedar conversion wasadjusted for cross section and leaf-bearing twig weightby multiplying by 1.5. In the case of mountain hemlock(Tsuga mertensiana), the Tsuga heterophylla coefficientwas applied.

Surface areas for broadleaves were obtained bydetermining similar coefficients (Table 5). Discs of aknown diameter were punched from leaves by usingvarious sizes of cork borers. Leaves were picked froma random selection of plants in each reference stand.For the long and narrow leaves of beargrass (Xerophyl-lum tenax (Pursh.) Nutt.) numerous 0.55-cm-diameterdiscs were punched for about three quarters of thelength of each sampled blade. One hundred discs foreach species represented on the stand were dried for24 h at 60 °C and weighed in groups of 10 to thenearest 0.1 mg. All broadleaf surface area figures formajor species were corrected for petiolar weight.

Estimating Total Leaf Area from VegetationalA nalyses

The vegetation on each reference stand was non-destructively surveyed in late June and July, 1974.This was accomplished for the shrub and herb layersby laying out twenty 4-rn ? plots, composed of four 1-m2subplots on the inner 30 m X 30 m of the stand. A10-m-wide buffer strip along each side of the RS was

0o.

L.

•

0

°•••.O

0.

z

oo0r--en r-rn

C S to 8. . 0 u 0 .,.., C

V) ..., V, 4,-+el el ce cc: ,...i3 '0 2

> . 0 > • .

•

>uU -t2 d . u (..) -52,..) ..... •-• ...;

cl, 0o..) -. -: d

a 3 i-1-a •"' trig o 0 c:4 2a ...? 2 V3

"

r**7 't -;

,5 —2! '1",.,- *6 f; :2 ci'',L) __., ....... 8 t., u.1:C) <=f) < o CD «2:2" <<—, c) ct - ct • --. —

X=, - —

Z ,:-.)C) (.) Z.?: - —.

T =

en WI r-

▪

r- •

cy • v::)re, .-.

..•-•• C7i4444,

r, r4 O cr r e oo CDrn ce Tr .0 Oroo r- r- r- vl CT

■. ■-•

....... 44., ,••••■ .44., ....... ...... 44, ,..... .....,r- 0r Tr en r4 vn 1,-. vD IN ......• - • CD

N N e•4 LC) •••-, 00 0, 01 0 C-I •CA rsi CV t".4 rs1 t^-4 •-• ,•-■ .--• ts4 •••• T--.■•-•4

cr r- r r, r■.o r, V1 r1 0,.0. 'I' r •CI• •-■

„ „. r4 rq

..•••■CT 0000

C7i 0 • el TnCV en it, es41 h■■

■-••• ■44.4

-. en co e- en rn r CD. • . .

eq 7t CD a i47 r4 ■ID CDCD CD cr cx) r`•• CO ••-•N N --N ...

CAN. J. FOR. RES. 6. 1976

22 r"--O

-

0

Tr rlCD 0.6 CD

52

cnerCbV

4,1

2(-4

3Vz

01-48

V0.

PI 00r-. .̂

...., •-• .....N. .••••■

W

C

L-a'5.23

9

1*.

.0

0

3

2

CD •WI`.0 •-••

0•-•

COoe C

Tr 0 s O C

0,eVv.0

eq C-tocre.3

CC

ECa

2r4 ,zr

-- 4o

O8

g

2

. .

to ,,,sa ,_.

^ ^g

uU

o coo

C

4

r-

co"Li '''', >VI 0

CCC

..-. .--. ....,r-

)v'' c' 'C.. 5 -o

rS

-ct< <--, --, •—,= =

3 .E-o

0.c)

q vu . C

11 ill .11. Of00 C o Z r.

; :07

c71

r.

GHOLZ ET AL. 53

TABLE 5. Surface area : weight conversions for common shrubs andherbs in the H. J. Andrews Experimental Forest

Species RS Cm2/ga %Petiole

Acer circinatum 1 603.6 (53.8)5 737.8 (26.4)6 438.4 (50.6) 6.28 (1.05)7 688.2 (40.0)

Berberis nervosa 1 88.4 (4.4)5 176.2 (7.8)6 139.6 (2.8) 23.17 (1.91)7 171.2 (5.4)

Rhododendron macrophyllum 5 261.2 (25.2)6 166.8 (12.6) 5.40(1.08)

Gaultheria shallon 5 228.0 (14.6)6 166.4 (12.8) 2.58 (0.76)7 269.4 (18.8)

Xerophyllum tenax 4 88.4 (6.0)5 78.0 (4.4) 0.006 77.8 (2.0)

Polystichum munitum 7 304.6 (6.6) 38.39 (3.90)Oxalis oregana 7 1193.6 (51.2) 41.77 (2.80)Castanopsis chrysophylla 6 140.4 (15.0) 2.73 (0.87)Corpus cornuta calif 1 424.2 (28.8)Flolodiscus discolor 1 420.6 (28.4) N.A.Cornus canadensis 4 245.8 (20.8) N.A.Vaccinium parvifolium 7 514.0 (15.2) N.A.Achyls triphylla 4 856.4 (38.4)

7 798.0 (78.2) N.A.Linnaea borealis 7 561.2 (9.4) N.A.Chimaphila umbellata 2 252.4 (9.8)

7 213.0 (8.8) N.A.Cornus nuttallii 6 350.6 (33.4) 3.12 (0.35)Coptis laciniata 2 599.0 (19.6) N.A.

Non: Standard deviations are shown in parentheses.aAll values are for two sidesbN.A., not available

not surveyed because of frequent disturbances there.Appropriate variables were measured for each speciesencountered. Cover for Oxalis, Oregon grape (Berberisnervosa Pursh.), salal (Gaultheria shallon Pursh.),and the herb layer was estimated on each subplot andan average was computed for the entire plot. Foliagebiomass for tree species was computed for the full50 m x 50 m area from diameter data collected byInternational Biological Program (IBP) personnel.'These data were used with the regressions of leaf dryweight on diameter at breast height (DBH) to com-pute species foliage biomass. Coefficients were applied

'Unpublished data on file at the Coniferous ForestBiome Data Bank, Oregon State University, ForestResearch Laboratory.

to foliage dry weights to determine leaf area by speciesin square metres per square metre. Values for individualspecies were summed to obtain totals for each refer-ence stand.

Results

The total foliage biomass, surface area, andpercentage surface area in the shrub and herblayers for each of the six stands are presentedin Fig. 2. The location of the values in Fig. 2corresponds to the environmental location ofeach reference stand as presented in Fig. 1.

The coniferous needle surface area coeffi-

MOIST

DOUGLASCLIMAX

-Fl20.21111.7 )20.711‘i

NEMLOC %(13.3)

32.45 11228 1 (18.2) 1 14Y

(15.3) 1 3% HEMLOCK5% tt ODAL

TRANSITIONZONE

42.1(24.8)

9%•■• •••• ■•■■ ••■••

SILVER-FIRZONE

33.13(19.9)

3%

54 CAN. I. FOR. RES. 6, 1976

100

0z80

310cc

ar 60

a.

40

0 10 20

30

PLANT MOISTURE STRESS, BARS

FIG. 2. Total foliage biomass (mt/ha, parentheticalvalues), total leaf area (m'/m 2, upper values), andpercentage of the total surface area in the shrubs andherbs (lower values) for the six reference standssuperimposed on Fig. 1. Projected leaf areas, one sidewith no cross-section corrections, are as follows: RS-1,8.7 m 2 /m'; RS-2, 14.3; RS-4, 14.4; RS-5, 18.6; RS-6,9.1; RS-7, 12.0. Sugar pine (Pinus lambertiana Dougl.)would add about 0.4 mt/ha to RS-1 and 0.9 mt/ha toRS-6, but it is not included here because there is nosurface area conversion available.

cients (Table 4) show that surface area pergram of needle decreased substantially withincreasing age, although not all age class valuesare statistically different. In old-growth Douglasfir, where about 20% of the total foliage wascurrent, 15% 1 year old, 15% 2 years old, and50% 3 years or older, the surface area decreasedfrom about 204 cm 2/g for the current foliageto 149 cm 2/g for the 3rd year and older foliage(statistically different at the 0.95 level). Thissuggests that substantial errors may result fromnot treating age classes separately. However,because of inadequate data, only a few speciesother than Douglas fir had leaf areas computedby age class.

Also, silver and noble fir growing in theunderstory had a greater surface area per gramthan those in the overstory (Table 4). Weexpect that this is also true for western hemlockand other species, which suggests in additionthat a variety of coefficients may be obtained

from various areas of the crowns of the largeoverstory trees.

Broad-leaved species illustrate a greater rangeof coefficients than the conifers (Table 5). Forexample, beargrass had 77 cm 2/g, whereasOxalis had 1194 cm2/g. Likewise, there is awide range of values within a single species. Forexample, vine maple (Acer circinatum Pursh.)ranged from 438 cm 2/g to 738 cm 2/g, and salalfrom 166 cm 2/g to 269 cm2/g.

DiscussionIn observing forests with different leaf areas

one may question whether these differences areassociated with environmental adaptations orare merely changes associated with stand devel-opment. There is general support for rapidincrease in leaf area from the time of establish-ment until crown closure. This period may be asshort as 4 years in eastern deciduous forests(Marks 1974) or may extend beyond a decadeas reported in a white pine (Pinus strobus L.)plantation by Swank and Schreuder (1973).After closure the amounts of foliage may returnto a somewhat lower plateau (Kira and Shidei1967; Ovington 1957).

The reported peaks are, however, not alwaysconvincing and may be offset by unreportedincreases in understory foliage development ofolder forests. We can probably assume, barringmajor windstorms, fire, or other recent acutedisturbances, that old-growth forest communi-ties do exhibit equilibrium leaf areas. Whetherthis is greater or less than that at the time ofcrown closure awaits confirmation.

In this study the magnitude of the leaf areasestimated for midelevation stands was higherthan those associated with cooler or warmerenvironments with similar moisture regimes(Fig. 2). In fact, the value 42.1 m 2/m2 (pro-jected leaf area of 18.6) for reference stand 5is near the maximum ever reported. Westmanand Whittaker (1975) recently reported asimilar value for a mature coast redwood forestin California. In a 26-year-old hemlock foreston the Oregon Coast, Fujimori (1971) esti-mated foliage at 21.1 mt/ha which converts to37 m2 /m2 of total leaf area using the appro-priate coefficient from Table 4.

What permits such a large accumulation offoliage to develop in an area of summer drought?

GHOLZ ET AL. 55

In the humid deciduous forests of the easternUnited States, total leaf areas above 12 m2/m2are uncommon (Whittaker 1966). Even in atropical rain forest of Thailand an exceptionaltotal leaf area of 32 m 2/m2 was reported, withmost plots supporting less than 24 m2/m2(Ogawa et al. 1965 ).

From the moisture gradient defined directlyby the stress experienced by understory conifers(Fig. 2 ), it is clear that soil water is not verylimiting to the forest communities where thehighest foliage areas accumulate.

Apparently, temperature is the variable whichpermits large accumulations of foliage, butneither the warmest sites nor those with thelongest growing season support the largestamount of foliage. Rather, we hypothesize thatit is a special balance of temperatures that favorsmodest respiration rates but yet permits, evenduring the dormant season, considerable netphotosynthesis. In the redwood forests of Cali-fornia, the Coast Range forests of the PacificNorthwest, and the midelevation forests of thewestern Cascades. the winter temperatures areusually above freezing, at least during the day(Waring and Major 1964; Zobel et al. 1973).Where there is a snowpack, day temperaturesusually remain above —2 °C so the watercolumn in the stem is not often frozen (Zimmer-man 1964) and both transpiration and photo-synthesis can occur. During the summer,coastal fog, maritime air, or elevational-relatedcooling create mild growing-season tempera-tures around 10 °C (Waring and Major 1964;Zobel et al. 1973). For conifers these tempera-tures permit very respectable rates of photo-synthesis (Walker et al. 1972 ).

This proposed explanation could alsoaccount for the decrease in leaf area in wellwatered but warmer habitats where respirationincreases substantially (RS-2 and RS-7). Athigher elevations colder winter temperaturesand a shorter growing season notably limitphotosynthesis (Emmingham 1974) and to-gether with snow damage could explain thelower value found there.

There are some special problems associatedwith estimating leaf area which also meritdiscussion. First of all is an inconsistency inexpressing results. Usually data from broadleafspecies have been reported as representing onlya single surface. This is not a major problem if

the convention is clearly noted by the authors.We have freely converted many of the valuesreported by Whittaker (1966), Westman andWhittaker (1975 ), and Ogawa et al. (1965) tototal surface area. There is real difficulty,however, when data for needle-leaf species arereported inconsistently on both a projected andtotal surface area basis. The problem becomesacute when both broadleaf and needle-leafspecies occurring in the same stand are reporteddifferently. We believe there is need to stand-ardize, and because stomata and energy ex-change can occur on more than one leaf surface,we feel total surface area is the most desirableexpression. From recent studies comes furthersupport that interception, photosynthesis,respiration, and possibly even foliar nutrientconcentrations are biologically best interpretedin such context (Linder 1974; Sollins et al_1974; Gholz, unpublished).

A second problem in estimating surface areais converting from biomass data. For example,in Table 5, differences in species coefficientsrange from 77 cm 2/g to 1194 cm 2/g, withfurther variation noted within a species, e.g.,Acer circinatum. Recent unpublished studies byGholz suggest much of the within-species varia-tion is a response to development under differentlight regimes. Light has similarly been attributedas the causal factor in explaining leaf area–weight differences in European beech (Kiraet a!. 1969) and a variety of oak–hickory forestspecies (Monk et al. 1970). These follow earlierclassical experimental work (e.g., Blackman andRutter 1948 ).

Additionally, there are differences in leafarea–weight coefficients as leaves age. This pointwas made earlier in discussing Table 4. For bothproblems, development of separate regressionequations for individual stands is a possible but,as we found, laborious solution.

Where possible, we recommend a direct esti-mation of leaf area, bypassing the difficulties ofconverting from foliar biomass. This appearspossible from simple linear relationships, suchas recently discovered between cross-sectionalarea of functional conducting tissue in the stemand leaf weight or area (Grier and Waring 1974;Dixon 1971 ). More recently, Waring, et al.(unpublished) found a similar linear coefficientfor shrub species Rhododendron macrophylhun.

These new coefficients may solve another

arge

IngeForreasis aFor•sh.).alal

reasareor

vei-pidish-e. asestsadeL.)3).urndei

IysedofIg:e

r

56 CAN. J. FOR. RES. 6, 1976

problem common to estimating foliar area oftrees. Ogawa et al. (1965 ), Kira and Shidei(1967), and Shinozaki et al. (1964) reportedthat estimates of foliage weight from measure-ments of stem diameter were often overesti-mates, especially for larger trees. Some of thismay be a result of competition among trees(Satoo 1966). This could also occur by extrap-olating from small trees which contain onlyconducting tissue to large trees with non-conducting heartwood. In this study, regressionequations were developed from data includinglarge trees. Still, because of the log relationships,a small change in stem diameter suggests a fairlylarge change in leaf weight and area. Directmeasurement of linear coefficients could resultin a considerable gain in accuracy.

Increasingly, accurate estimates of leaf areaare necessary to test ecological hypotheses con-cerning the functional recovery of ecosystems.How long recovery takes is often a function ofreestablishment of leaf area, sometimes by shrubrather than tree species. As the importance ofunderstory species in nutrient cycling has be-come better appreciated, accurate estimates oftheir contribution to uptake, storage, and litterfall are needed. With the development ofaccurate methods, we hope quantitative esti-mates of leaf area will become more commonlyreported in vegetational analyses.

In summary, when the total leaf areas of sixold-growth forest communities in westernOregon were estimated, differences were ob-served which were attributed mainly to varia-tions in the physical environment. The highestvalues may represent maximum accumulationand reflect particular combinations of environ-ment that permit adequate photosynthesis withmodest respiration in maritime-influenced cooltemperate forests of the Pacific Northwest.Methods developed in this paper will enableecologists to explore further the foliage rela-tions of both disturbed and undisturbed plantcommunities.

AcknowledgementsWe thank Dr. W. H. Emmingham, Dr. C. C.

Grier, and S. W. Running, of the ConiferousForest Biome, Oregon State University, for helpin the preparation and review of the manuscript.

BASKERVILLE, G. L. 1965. Estimation of dry weight oftree components and total standing crop in coniferstands. Ecology, 46(6): 867-869.

BLACKMAN, G. E., and A. J. RUTTER. 1948. Physiologi-cal and ecological studies in the analysis of plantenvironments. HI. The interaction between lightintensity and mineral nutrient supply in leaf devel-opment and in the net assimilation rate of theBluebell (Scilla non-scripta). Ann. Bot. (London),N.S. 11(45): 1-26.

CLEARY, B. D., and R. H. WARING. 1969. Temperature:collection of data and its analysis for the interpreta-tion of plant growth and distribution. Can. J. Rot.47: 167-173.

DIXON, A. F. G. 1971. The role of aphids in woodformation. J. Appl. Ecol. 8: 165-179.

DREW, A. P., and S. W. RUNNING. 1975. A comparisonof two techniques for measuring surface area ofconifer needles. For. Sci. 21(3): 231-233.

DYRNESS, C. T., J. F. FRANKLIN, and W. H. Mom. 1974.A preliminary classification of forest communitiesin the central portion of the Western Cascades inOregon. Bull. No. 4, Conif. For. Biome, USIBP,Univ. Wash., Seattle, Washington.

EMMINGHAM, W. H. 1974. Physiological responses offour Douglas-fir populations in three contrastingfield environments. Ph.D. Thesis, Oregon State Uni-versity, Corvallis, Oregon.

FUJIMORI, T. 1971. Primary productivity of a youngTsuga heterophylla stand and some speculationsabout biomass of forest communities on the Oregoncoast. U.S. Dep. Agric. For. Ser. Res. Pap. PNW-123.

GEPPERT, R. R. 1968. The effect of age on transpirationrates of Douglas-fir seedlings. M.Sc. Thesis, OregonState Univ., Corvallis, Oregon.

GRIER, C. C., and R. H. WARING. 1974. Conifer foliagemass related to sapwood area. For. Sci. 20(3):205-206.

KIRA, T., and T. SHIDEI. 1967. Primary production andturnover of organic matter in different forest eco-systems of the Western Pacific. Jap. J. Ecol. 17(2):70-87.

KIRA, T., K. SHINOZAKI, and K. HozuMi. 1969. Struc-ture of forest canopies as related to their primaryproductivity. Plant Cell Physiol. 10: 129-142.

KRUEGER, K. W., and R. H. Rum. 1969. Comparativephotosynthesis of red alder, Douglas fir, Sitka spruceand western hemlock seedlings. Can. J. Bot. 47:519-527.

KRUMLIK, J. G. 1974. Biomass and nutrient distribu-tion in two old-growth forest ecosystems in S.Coastal British Columbia. M.S. Thesis, Univ. B.C.,Vancouver, British Columbia.

LINDER, S. 1974. A proposal for the use of standardizedmethods for chlorophyll determinations in ecologicaland eco-physiological investigations. Physiol. Plant.32: 154-156.

MARKS, P. W. 1974. The role of pin cherry (Prunuspensylvanica L.) in the maintenance of stability innorthern hardwood ecosystems. Ecol. Monogr.44( / ): 73-88.

GHOLZ ET AL. 57

ight of:onifer

iologi-plantlight

devel-if theidon),

attire:preta-

BOt.

wood

,rison:a of

1974.nitieses in^IBP,

es ofstingUni-

Dungiotts:gon

ionon

ge):

d

MILLER, E. E., C. A. SHADBOLT, and L. HOLM. 1956.Use of an optical planimeter for measuring leafarea. Plant Physiol. 31(6) : 484-487.

MONK, C. D., G. I. CHILDS, and S. A. N ICHOLSON. 1970.Biomass, litter and leaf surface area estimates of anoak-hickory forest. Oikos, 21: 138-141.

OGAWA, H., K. YODA, K. OGINO, and T. KntA. 1965.Comparative ecological studies on three main typesof forest vegetation in Thailand. II. Plant biomass.In Nature and life in S.E. Asia. Vol. IV. Edited byT. Kira and K. Iwata. Japan Soc. Prom. Sci., Tokyo.pp. 49-80.

OVINGTON, J. D. 1957. Dry-matter production by Pinussylvestris L. Ann. Bot. (London), N.S. 21(82):287-314.

REINERS, W. A. 1972. Structure and energetics of threeMinnesota forests. Ecol. Monogr. 42(1): 71-94.

RUSSELL, D. W. 1973. The life history of vine mapleon the H. J. Andrews Experimental Forest. M.S.Thesis, Oregon State University, Corvallis, Oregon.

SATOO, T. 1966. Production and distribution of drymatter in forest ecosystems. Misc. Info. Tokyo Univ.For. 16: 1-15.

S IIDIOZAKI, K., K. YODA, K. Hozumt, and T. KIRA.1964. A quantitative analysis of plant form-thepipe model theory. II. Further evidence of the theoryand its application in forest ecology. Jap. J. Ecol.14(4): 133-139.

S OLLLNS. P., R. H. WARING, and D. W. COLE. 1974. Asystematic framework for modeling and studyingthe physiology of a coniferous forest ecosystem. InIntegrated research in the coniferous biome. Editedby R. H. Waring and R. L. Edmonds. Bull. No. 5,Conif. For. Biome, USIBP, Univ. Wash., Seattle,Washington.

SWANK, W. T., and H. T. S CHREUDER. 1973. Temporalchanges in biomass, surface area, and net productionfor a Pinus strobus L. forest. In IUFRO biomassstudies. Working Party on the Mensuration of the

Forest Biomass. S4.01. Mensuration, Growth, andYield. Coll. Life Sci. Agric. Univ. of Maine, Orono,Maine.

WALKER, R. B., D. R. M. Scorr, D. J. S .11.3, and K. L.REED. 1972. Terrestrial process studies in conifers:a review. In Proceedings-research on coniferousforest ecosystems-a symposium. Edited by J. F.Franklin, L. J. Dempster, and R. H. Waring. Pac.NW For. Range Exp. Stn. U.S. Dep. Agric. For.Ser., Portland, Oregon.

WARING, R. H., and B. D. C LEARY. 1967. Plant moisturestress: evaluation by pressure bomb. Science, 155:1248-1254.

WARING, R. H., and J. MAJOR. 1964. Some vegetationof the California coastal redwood region in relationto gradients of moisture, nutrients, light, and tem-perature. Ecol. Monogr. 34: 167-215.

WESTMAN, W. E., and R. H. WarrrAKER. 1975. Thepygmy forest region of Northern California: studieson biomass and primary productivity. J. Ecol. 63(2):493-520.

WHITTAKER, R. H. 1966. Forest dimensions and produc-tion in the Great Smoky Mountains. Ecology. 47(1):103-121.

Z IMMERMANN, M. H. 1964. Effect of low temperatureon the ascent of sap in trees. Plant Physiol. 39: 568-572.

ZOBEL, D. B., W. A. McKEE, G. M. HAWK, and C. T.DYR:VESS. 1973. Variation in air and soil tempera-tures in forest communities in the H. J. AndrewsExperimental Forest, 1970-1972. Int. Rep. 127,Conif. For. Biome, Ecosystem Analysis Studies,USIBP, Univ. Wash.. Seattle, Washington.

1974. Correlation of forest communities withenvironment and phonology on the H. J. AndrewsExperimental Forest, Oregon. In Integrated researchin the coniferous biome. Edited by R. H. Waringand R. L. Edmonds. Bull. No. 5, Conif. For. Biome.USIBP, Univ. Wash., Seattle, Washington.

-ow, •vrap. • • ....nr.,••••■•••••••••.• •