VEGETATIONAL ANALYSIS OF THE
COASTAL PICEA SITCHENSIS FOREST ZONE
IN OLYMPIC NATIONAL PARK, WASHINGTON
A Thesis
Presented to
the Department of Biology
Western Washington State College
In Partial Fulfillment
of the Requirements for the Degree
Master of Science
Andrew Michael Kratz
June 1975 WASHINGTON NATURAL HERITAGE PROGRAM BLDG 17 AIRDUSTRIAL CENTER • OLYMPIA WA 98504 • 206.753.2449
September 20, 1978
Dr. J. F. Franklin USDA Forest Service 320 Jefferson Way Oregon State University
Dear Jerry,
How s it feel to be clean again? A little lime seems to have taken care of the moss on the north side of my nose.
I am enclosing a copy of my thesis, with apologies again for the three year delay. I will confess to a good deal of dissatisfaction with my thesis, which is one reason I have never entertained the idea of publishing the information. Soils data were poorly collected out of ignorance. Samples should have been air dried soon after collection, not with as much as a two week delay. Texture analysis was incomplete for "lack of time" and not knowing how to deal with heavy soils which once dried formed equivalents to adobe bricks - Calgon would not cause dispersion of the clays and grinding either reduced the minibricks to sand size or reduced the sand component to clay or silt size. In other words, there isn t much information in my soils data.
Tree growth rate data was hampered by a dull increment borer and my inexperience telling me that either I was doing something wrong or there was something wrong with the trees (not all the cores came out corkscrewed). I finally did get a sharp one.
I have distinct reservations on the use of prominence values beyond all other doubts of the implications of the other number games I used. Multiplying cover and frequency (acutually square root of frequency) renders indistinguish- able the very different situations of high cover with low frequency and the reverse.
For example:
SPECIES A - All 20 Daubenmire microplots with 5-25% coverclass (mean cover 15% and frequency = 100%) has a PV=150.
SPECIES B - 6 microplots in the 95-100% coverclass, remainder with none (mean cover = 29.25% and frequency = 30%) has a PV=160. September 20, 1978 Page 2
SPECIES C - 5 microplots in the 95-100% coverclass, others with none (mean cover = 24.38% and frequency = 25%) has a PV=122.
SPECIES D - 15 microplots in the 5-25% coverclass, 5 microplots in the 0-5% coverclass. This species has actual coverage varying from 4 to 10%. (Mean cover = 11.88% and frequency = 100%). PV=119.
Using PV s tends to obscure the similarity in distribution of species A and D and masks the differences between C and D. Species A and B would be weighted more similarly than would species B and C.
These are of course "theoretical" considerations and perhaps in the overall analysis of communities it doesn t make too great a difference. Nevertheless, I have little faith in number-manipulation that has such great potential for obscuring ecological "reality".
I want to thank you for having me along in the Hoh; it was good to be a part of the team. Though my field experience has been limited, that venture was the most outstanding - I wish it could have been longer. The group spirit was tremendous and the interdisciplinary approach made the campfire talks all the more interesting. I m looking forward to seeing the results.
Take care, and let me know if there s something I can do.
Yours truly,
Andrew Kratz
AK:pl
Enclosure iii
ACKNOWLEDGEMENT
I wish to thank my advisor, Dr. Richard W. Fonda, for his guiding influence and assistance during the full course of this study. I must extend a very special thanks to my youngest brother, Richard E. Kratz, for the very considerable help, companionship, and encouragement he gave me over many long miles during my summer in the field. Grateful acknowledgement is made to
Mi. Michael Bortel for his valuable help with computer analysis of data; to the administration and staff of Olympic National Park, and in particular
Mr. William Lester and Mr. Michael Kalahar, for their cooperation and provision of shelter when most needed; and to Drs. Ronald J. Taylor and -
Franklin C. Raney for their constructive critique of the manuscript. Finally,
I wish to thank the numerous unnamed people at home and on the Olympic
Peninsula who helped in so many ways to make this study possible. iv TABLE OF CONTENTS
ACKNOWLEDGEMENT iii
LIST OF FIGURES vi
LIST OF TABLES vii
INTRODUCTION 1
STUDY AREA 3
Location 3
History 3
Geology 3
Climate 7
METHODS 12
Vegetation 12
Field work 12
Data analysis 12
Tree growth rates 15
Soils 17
RESULTS 18
Vegetation 18
Picea sitchensis-Alnus rubra/Rubus spectabilis community 18
Picea sitchensis/Gaultheria shallon community 18
Picea sitchensis/Polystichum munitum community 18
Picea sitchensis/Carex obnupta community 23
Picea sitchensis/Maianthemum dilatatum community 23
Picea sitchensis/bryophytes community 24
Picea sitchensis-Tsuga heterophylla/Blechnum spicant community 24
Tsuga heterophylla-Picea sitchensis/Polystichum munitum community 25
Tree growth rates 26
Soils 26 V
DISCUSSION 33
Pattern 33
Disturbance 36
Phytogeography 37
SUMMARY 38
LITERATURE CITED 39
VITA 41 vi
LIST OF FIGURES
Figure
Study area and numbered sample plot locations 5
Estimated evapotranspiration for 15 cm water storage capacity soil on Tatoosh Island, calculated by monthly means for a 30 year period from 1931 to 1960 10
Hytherographs from the tropical rain forest in Guyana, the boreal forest in Manitoba, Canada, and the coastal Picea forest in Washington 11
Sampling plot 14
Phenogram showing the grouping of stands into communities 16
Transect taken near Norwegian Memorial extending 200 m into the forest from the edge of the beach 35 vii
LIST OF TABLES
Table
Percentage frequency of wind direction by qua4rant at Tatoosh Island and Moclips 9
Mean density, mean basal area, and standard errors of the mean of different tree species in the eight communities 19
Size class distribution for mean number of trees/ha in the eight communities 20
Understory species composition of the eight communities 21
Average growth rates and ages with standard errors of the mean for Picea sitchensis 37-57 cm dbh in seven of the eight communities 27
Chemical properties of soils under seven of the eight communities 28
7. Selected descriptions of soils under seven of the communities 30 1
INTRODUCTION
Picea sitchensis occurs naturally in a narrow band along the Pacific
Coast from Kodiak Island, Alaska, to Mendocino County, California. Along the west coast of Washington P. sitchensis is an element in a coastal forest band that is several kilometers wide. This forest band on the coastal plain of the western Olympic Peninsula is dominated by Tsuga heterophylla and Thuja plicata, but it is replaced by a narrow belt of Picea sitchensis-dominated forest adjacent to the beach.
There is little published information on these forests of Picea sitchensis. Few papers deal with the synecology of Picea communities, and even the autecology of the species is poorly understood. Fowells (1965) provides a good silvicultural summary. Picea sitchensis is generally considered to be a prolific species, but it produces less seed than its most common associate, Tsuga heterophylla. Flowering occurs between late March and April; seed fall lasts from about October through February. Seeds are scattered adequately for 0.5 km from their source, and they will germinate on almost any kind of seedbed if moisture is abundant. Silvicultural establishment is best on mineral soils with side shade and overhead light, but these conditions seldom occur in nature. Picea seedlings can also become established on rotten wood or moss, but on thick layers of moss they are subject to drought during dry spells. Along the Olympic Coast Rubus spectabilis, Sambucus racemosa, and Gaultheria shallon compete with, and easily outgrow, Picea seedlings.
Frequently only the seedlings on rotten wood survive, because of low shrub cover on these logs. Picea sitchensis is a vigorous, fast-growing, shallow- rooted tree that readily overtops Tsuga heterophylla and Thuja plicata. It is classed as shade tolerant, but less so than these two competitors
(Franklin and Dyrness, 1973: 48). 2
Jones (1936) designated a climax forest of Tsuga heterophylla and Picea sitchensis along the Olympic coast, but he did not describe communities of this forest. Daubenmire (1969) mentioned that Picea sitchensis plays a climax role in a narrow strip along the ocean, but that it is seral to Tsuga heterophylla away from the beach. Franklin and Dyrness (1973) described the general composition of forests in the Picea sitchensis Zone along the coast of Oregon, but did not present compositional data. Cordes (1972), working on Vancouver Island, appears to have made the only synecological study of forests dominated by Picea sitchensis. He used the Braun-Blanquet approach.
To date there has been no study of Picea sitchensis-dominated forest communities on the Pacific Coast using the approaches of American ecologists.
This study was undertaken to answer the following questions: 1) what is the extent of the band of forest dominated by Picea sitchensis on the Olympic coast? 2) what kinds of plant communities make up this zone? 3) what environmental factors control the pattern of plant communities? I will show that the Picea sitchensis zone is restricted to within 200 m or less of the beach along the Olympic coast, and that it comprises at least eight plant community types. Community pattern, however; could not be related to any single limiting factor.
Botanical nomenclature follows Hitchcock and Cronquist (1973). 3
STUDY AREA
Location.- Olympic National Park is located on the Olympic Peninsula in the northwest corner of the State of Washington. The Park includes most of the Olympic Mountains and a separate 8,100 ha Pacific Ocean coastal strip stretching 80 km from just above the Ozette Indian Reservation on the north to the border of the Quinalt Indian Reservation on the south. The average width of this strip is about 2 km. Park property is interrupted along the
Pacific beach only by the Ozette, Quillayute, and Hoh Indian Reservations.
This coastal strip constitutes the study area (Fig. 1).
History.- Indians have lived along the coast for at least 2,000 years
(Kirk, 1974) and perhaps for 5-6,000 years (Kirk, pers. comm.) with little modification of the natural vegetation before the twentieth century
(Fagerlund, 1954). Europeans first entered the Olympic Peninsula in the late
18th century, and by 1840 fur trade and white settlement had greatly increased.
There has been no logging in the coastal strip, however, largely because there was no practical way to transport timber from the area. Only local land clearing and lumber needs have been met.
There was a brief riverine gold rush in 1894 along the coast; the most productive area was between the Quillayute River and Cape Flattery. The vegetation was hardly disturbed, however, because the gold was removed from riverbeds rather than from the land.
The coastal strip was added to the Olympic National Park in 1953, although it had been under National Park Service administration for some years prior to 1953.
Geology.- Little study has been made of the geology of this coast. The most recent and detailed work covers the area from Point Grenville north to the Hoh River (Rau, 1973). The following discussion is based primarily on the work of Rau (1973) and Danner (1955). 4
Fig. 1. Study area and numbered sample plot locations (•). Totoog.h Island INSET A Cape Flattery
Cape Alava °C46 tt Indian Reservation 35 Portage i Heicid 3738 -39 Point jof Arches
32 -31 34 :33 30 River Olympic ional Park River 0sr.• 0 Mount s Olympus
0
-n Whale Cre Ouinault Indian Reservation Point Grenville N • Moclips
20 19 18 Aberdeen Rialto Beach Grays Harbor
10 20 30 km
INSET B
N ° Norwegian Mlmorial • 14. 15 67 (-7
10 .11 .13 transect
28 41
r(:) 27 .3 Destruction Island 2
Ca mpground 23o Kalalach 24. 26
25 5 km
1 km The geology of the coast is complicated because of eustatic, isostatic, and other tectonic changes in land-sea level relations. In the late Pleisocene eustatic changes were small; a rise in sea level of about 2 m may have occurred during the Hypsithermal. Isostatic and other tectonic changes are usually large, measured in hundreds of meters, but are hard to distinguish between in areas that have been glaciated (Heusser, 1960: 189).
Rock outcrops along the coast are largely sandstones, shales, and conglomerates of marine origin, but some are volcanics composed of several kinds of dark, basic lava. The oldest rocks along the coast are possibly of lower Cretaceous age (Soleduck Formation). Sedimentary rocks of Oligocene,
Miocene, and Pliocene age are also found along the beach. Several offshore islands and reefs are formed of volcanic rock occurring with Miocene sediments.
During the Eocene the present coast was under the sea. Lava flows of the Metchosin volcanics spread out on the sea floor, but since uplift of the
Olympic Mountains only remnants exist along the coast, mostly at Point of
Arches, Portage Head, and probably Point Grenville (Fig. 1). After lava deposition the area underwent several erosion cycles.
Most of the coastal plain is mantled with unconsolidated marine or freshwater sediments including sand, gravel, clay, and peat which were deposited at several different times during the Pleistocene. Some of the deposits in the area between Point Grenville and the Hoh River may have come from alpine glaciers that reached the present coastline. After the deposition of the younger sand and gravel deposits, and probably before much vegetation had developed, windblown sand and silt covered the coastal area to a depth of
1 m or more.
In the area between Whale Creek and Kalaloch (Fig. 1) late Pleistocene sand and gravel deposits are revealed in wave-cut cliffs. These materials once formed a broad surface that extended from the foothills of the Olympics to sea level at some distance from the present coast. Destruction Island,
6 km offshore, was once part of this surface, perhaps only 6,000 years ago.
Eastward erosion has produced cliffs ca. 25 m high along the present coastline, but the cliffs at Kalaloch are lower because of downwarping.
Crustal movements during the Pleistocene in the Soleduck Valley area tilted and folded some of the older Pleistocene sediments. Southward, the crust was downwarped throughout the Pleistocene epoch, with the central axis in the Kalaloch area.
Thousands of years of erosion formed the present coastline, and these forces are likely to continue. Land surveys in this century show that there is an average erosion rate of 1.1 m per year in some places south of the
Hoh River. From 1902 to 1962 two adjacent government lots lost 51 and 60 percent of their total surfaces. This high rate of erosion is partly because of the susceptability to erosion of sand, gravel, and clay deposits. Part of the Hoh rock assemblage, another deposit which is common along the beach, is highly erodable. Technically this portion is a tectonic melange, consisting of chaotically mixed resistant rocks set in a fine-grained matrix of softer materials. These deposits are structurally weak because of the clay minerals which expand when wetted by waves and precipitation. Consequently, the Hoh melange slumps periodically where it is exposed along the coast.
Climate.- The climate of the Picea sitchensis Zone is the wettest and mildest of any northwestern vegetation zones (Franklin and Dyrness, 1973).
The Cascade Mountains to the east shield the Olympic Peninsula from arctic air flows in the winter and continental air flows in the summer. The west coast of the peninsula is, however, exposed to gale force winds and heavy rains during winter. On the coast, winds blow most frequently from the east during fall and winter, but from the south or west during spring and summer (Table 1). The strongest winds are associated with winter storms moving eastward across the Pacific Ocean (Phillips and Donaldson, 1972).
Precipitation averages 1800-2300 mm per year. It is distributed unevenly through the year with a maximum in winter; December is the wettest month, July is the driest. Assuming the soil has a 15 cm waterholding capacity in the root zone, the plants will experience a water deficit of less than 1 cm during the summer months as indicated by the difference between potential and actual evapotranspiration (Fig. 2). Actual moisture stress experienced by the plants in any given habitat is likely to differ from this estimation because of soil type, water table depth, drainage patterns, rooting characteristics of the individual plants, etc. Figure 2 represents an average for a 30 year period and does not reflect the extremes which can have important effects on the vegetation. Rainfall intensity is light to moderate.
Fog is most frequent during the summer, with fog drip resulting from condensation in the canopy.
Mean monthly temperatures range from 6 C in winter to 14 C in summer.
During the warmest summer months the afternoon temperatures are about 15-20 C and the lows are around 7-12 C. In winter the average maximums are around
10 C and the average minimums are near 2 C. Maximums reach 0 C or lower on an average of about 7-12 days during the year (Phillips and Donaldson, 1972).
Relative humidity ranges between 80-95% throughout the day year round.
Annual sunshine along the north Pacific coast is less than 40% (Heusser, 1960:
30).
A comparison of hytherographs from Tatoosh Island, Washington, Fort Good
Hope, Manitoba, and Georgetown, Guyana reveals that the spruce zone in the
Olympics has the general range of precipitation and the overall uniformity of temperature of the tropical rainforest, but that mean monthly temperatures are more similar to those of the warmest months in the boreal forest (Fig. 3). TABLE 1. Percentage frequency of wind direction by quadrant at Tatoosh Island and Moclips. Calm periods are not included. Data based on an 11 year period from 1948 to 1958 (Phillips and Donaldson, 1972).
Tatoosh Island Moclips N E S w N E S W J 4.8 54.2 28.7 11.9 5.1 66.8 11.9 6.9 F 6.0 42.2 32.2 18.8 5.1 58.0 12.9 13.9 M 7.8 38.9 29.6 22.4 10.8 37.9 16.0 24.2
A 9.1 26.6 34.2 28.4 8.7 25.6 19.8 33.7 M 8.0 15.6 40.1 34.2 9.1 16.9 17.2 45.2 J 6.2 14.2 47.9 30.0 8.3 9.5 10.9 56.3
J 4.9 8.7 62.8 22.2 11.0 7.9 14.4 45.4 A 6.2 14.1 63.9 13.9 8.0 11.4 13.0 43.4 S 12.9 27.7 42.6 14.3 10.2 19.3 13.1 31.1
0 9.7 /1/1.8 30.5 13.5 9.9 40.1 22.0 13.4 N 5.7 53.5 27.4 12.6 11.0 55.2 14.5 11.4 D 3.0 /1/1.0 33.3 18.8 9.8 50.6 19.4 10.2 10
30 -
25 - AVERAGE MONTHLY PRECIPITATION
AVERAGE POTENTIAL MONTHLY EVAPOTRANSPIRATION
20 - AVERAGE ACTUAL MONTHLY EVAPOTRANSPIRATION
10 - WATER DEFICIT
5 -
1 I 1 i I I I I I JAN FEB MAR1 APR MAY JUN JUL AUG SEP OCT NOV DEC
Fig. 2. Estimated evapotranspiration for 15 cm water storage capacity soil on Tatoosh Island, calculated by monthly means for a 30 year period from 1931 to 1960. Average actual evapotranspiration and potential evapotranspiration are different only from June through August. Data from Phillips and Donaldson (1972). 11
30 —
10 7 c9 2 1 12 6 GEORGETOWN, GUYANA
20--
8
10 10 — 11
12 3 2 TATOOSH ISLAND, WASHINGTON
10
-10 -. FORT GOOD HOPE, MANITOBA
-20 — 11
1 -30 — 1 1 I I 5 10 15 20 25 30 35 MEAN MONTHLY PRECIPITATION (cm)
Fig. 3. Hytherographs from the tropical rain forest in Guyana, the boreal forest in Manitoba, Canada, and the coastal Picea forest in Washington. Months are numbered consecutively (1 = January). Data from Phillips and Donaldson (1972) for Washington, Fonda (pers. comm.) for Guyana and Manitoba. Data for Washington based on the 30 year period from 1931 to 1960 (Phillips and Donaldson, 1972). Data for Guyana and Manitoba based on unknown period (Fonda, pers. comm.). 12
METHODS
Vegetation
Field work.- A preliminary survey of the Picea forests in the study area showed that this coastal zone included several different community types which
could be largely differentiated on the basis of understory vegetation. A general classification of the vegetation by the major understory components was then devised and a total of 38 sample plots were studied during summer 1973, with between three and fifteen plots per tentative community type.
All stands but one were sampled using a single 15 x 25 m macroplot
randomly located in a homogeneous portion of the stand (Fig. 4). A 25 m line was run down the center of the macroplot to establish its length, then a 7.5 m
line was repeatedly run perpendicularly to the centerline to establish the width and locate the understory plots within the macroplot. The herbaceous understory was sampled by systematically placing a 20 x 50 cm frame at 20 points
inside the macroplot, and shrubs were sampled in eight 1 m radius circular plots regularly spaced within the macroplot. Shrub and herb coverage was estimated by the cover classed of Daubenmire (1959), and his mid-points were used in analyzing the data. All trees larger than 2.5 cm dbh were measured on the
15 x 25 m macroplot. Tree reproduction was counted in the 1 m radius circular plots. Bryophytes and lichens were treated together as a single component of
the understory.
Data analysis.- After all data had been collected and summarized by stand, coefficients of similarity for stands were calculated based on prominence values (PV) of all species. The values for understory species were calculated as PV = mean % cover x frequency l (Beals, 1960). An alternate formula was used for tree species (PV = basal area x density 2 ), because there were no frequency values for trees. Both methods of calculating prominence values 13
Fig. 4. Sampling plot. Trees were tallied on 15 x 25 m plot, shrubs on 1 m radius plots, and herbs on 20 x 50 cm frames. • ..! .1,-4,•=s-A.
Im RADIUS f-20x50cm FRAME
I5m 10 15 20 25
25m 15 yielded numbers between 0.28 and 945, and growth forms (tree, shrub, herb) were equally weighted; the highest prominence value in a given stand could be held by a tree, shrub, or herb.
Similarity coefficients were calculated by comparing all stands two at 00)b2 w a time using the formula C = (1 where C is the coefficient of similarity, a + a and b are the sums of the prominence values calculated for stands A and B, and w is the sum of the lowest common value of the two stands being compared.
A phenogram was then constructed from the similarity matrix following the methods of Mountford as described in Southwood (1966: 342-344). The phenogram (Fig. 5) shows a grouping of the stands based on their similarities.
Division of the stands into communities was made by the selection of a minimum percentage similarity for members of the communities. In Fig. 5 the 55% similarity level shows the grouping of stands into community types described under Results. The phenogram required modification of the original community classification, after which mean values of cover and frequency for understory species, and basal area and density for tree species of each community were calculated. These values appear in Tables 2, 3, and 4. The number of plots in each finalized community type is shown in Table 2.
Tree growth rates
To assess differences in growth conditions increment borings from Picea in seven of eight communities were examined for growth rates. Only trees of
37-57 cm dbh were drilled since annular ring width decreases as a function of tree diameter even under otherwise constant conditions. Because annular rings are usually larger on the downhill side of a tree, all trees were drilled from the south side, except in the Picea-Alnus/Rubus stand in the
Rialto Beach area where they were drilled from the east side, since the coast runs east-west there. 16 0 9 PICEA/CAREX
22-
16 34 PICEA/GAULTHERIA 27 11 7 17 13 39 TSUGA-PICEA/POLYSTICHUM 15 12 37- PICEA-TSUGA/BLECHNUM 36 30- 8 25 PICEA/BRYOPHYTES 32 4 26- 28 PICEA/MAIANTHEMUM 18 2_ 40 41 PICEA-ALNUS/RUBUS 38 21 19 5 20 35 PICEA/POLYSTICHUM 31 14 29 33 1_
I I I I I r I 10 20 30 40 50 60 70 80 90 100 PERCENT SIMILARITY
Fig. 5. Phenogram showing the grouping of stands into communities. 17
Data on Picea growth rates were subjected to analysis of variance using a completely randomized design and the level of significance chosen at 5% before testing.
Soils
I sampled soils of each community shown in Table 2, except for the
Picea-Tsuga/Blechnum community which was revealed after the phenogram was constructed. Soil pits were dug in or adjacent to 17 of the macroplots, and the profiles were described. A composite 1 kg sample of soil was collected from each horizon, and air dried within 14 days. Samples were passed through a 2 mm sieve and subsequent analyses were carried out on the 2 mm fraction.
Twenty-six of the samples from 5 and 30 cm depth in 14 soil pits were analysed for pH by the glass electrode method, NO 3-N by calcium oxide extraction with color developed by phenoldisulfonic acid and read on a colorimeter, Mg and Na by ammonium acetate extraction for atomic absorption and flame spectrophotometer, respecitively. These were standard analyses performed by technicians at Washington State University, Pullman. Two of the pits were less than 30 cm deep, however, so that only 5 cm deep samples were analysed for them.
Data on chemical properties of the soil were subjected to analysis of variance using a completely randomized design and the level of significance chosen at 5% before testing. 18 RESULTS
Vegetation
Picea sitchensis-Alnus rubra/Rubus spectabilis community.- This was one
of the most common communities facing the coast. Picea sitchensis and Alnus
rubra dominated the community (Tables 2 and 3). The two species had similar
density, but Picea had a larger total basal area. Tsuga heterophylla was minor in this community. One stand, south of La Push, had Picea seedlings,
but the other two stands had no tree reproduction.
The shrub layer was about 2 m tall and had a mean cover of 83%; herbaceous
cover was 57%. The dominant understory species were Rubus spectabilis,
Polystichum munitum, Sambucus racemosa, Gaultheria shallon, and Athyrium
filix-femina (Table 4).
Picea sitchensis/Gaultheria shallon community.- This community was commonly encountered on level to steeply sloping surfaces where it sometimes covered areas of a few hectares. It was generally exposed to storm winds off the ocean. Picea sitchensis dominated the canopy (Tables 2 and 3). Tree reproduction was sparse (Table 3).
The shrub layer ranged from 1.5-3 m in height. Shrub cover averaged 85%, herbaceous cover 15%. The dominant shrub species, Gaultheria shallon, was sometimes accompanied by Rubus spectabilis (Table 4). Polystichum munitum was the most important herb species and was present in all stands of this community.
Picea sitchensis/Polystichum munitum community.- This common community occurred on level to steeply sloping surfaces. It occasionally was found next to the beach but was usually protected from ocean winds by a line of young trees or Gaultheria shallon. Picea sitchensis clearly dominated this community.
Tsuga heterophylla and Alnus rubra were also present but were always minor 1 q
TABLE 2. Mean density, mean basal area, and standard errors of the mean of the different tree species in the eight communities.
Mean Mean Density Bas1 Area Community Tree (1i/ha) (m /ha) Plots
Picea-Alnus/Rubus Picea 151 + 49 63 + 20 Tsuga 27 + 15 <1 + <1 3 Alnus 133 + 80 20 + 11 Pyrus 9 + 9 _<1 T <1 Total 320 33 Picea/Gaultheria Picea 373 + 44 74 + 9 Tsuga 9+ 9 <1 + <1 6 Alnus 13 + 9 2 + 1 Total 396 76 - Picea/Polystichum Picea 368 + 59 79 + 11 Tsuga 53 + 22 7 + 4 10 Alnus 72 + 29 8 + 3 Total 493 94 Picea/Carex Picea 222 + 105 27 + 2 Tsuga 18 + 18 4 + 4 3 Alnus 53 + 27 6 + 2 Populus 18 + 18 5 + 5 Total 311 42 Picea/Maianthemum Picea 860 + 171 67 + 7 Tsuga 33 T 33 1 + 1 4 Thuja 40 + 40 2 + 2 Total 933 70 Picea/bryophytes Picea 869 + 139 69 + 5 Tsuga 229 + 101 6 + 3 Thuja 139 + 97 3 + 2 5 Alnus 5 + 5 ,1 + <1 Pyrus 16 + 16 (1 + <1 Total 1259 79 Picea-Tsuga/Blechnum Picea 213 + 80 146 + 21 Tsuga 360 + 173 36 + 5 2 Total 573 182 Tsuga-Picea/Polystichum Picea 147 + 52 39 + 10 Tsuga 693 + 154 67 + 8 4 Total 840 105 20 TABLE 3. Size class distribution for mean number of trees/ha in the eight communities.