of epiphytic bryophyte
Ordination
British Columbia*
South-Coastal
communities
in a wet-temperate
coniferous
N. C. Kenkel1 & G. E. Bradfield2**
1
Department
of Plant Sciences, University of Western Ontario, London, Ontario N6A
2
Vancouver, British Columbia
Department
of Botany, University of British Columbia,
Bark, British Columbia,
Keywords:
Bryophyte,
Coniferous
forest, Epiphyte,
Gradients,
forest,
3K7, Canada
V6T2B1,
Canada
Ordination
Abstract
coniferous tree species (Pseudotsuga menziesii,
at Capilano
and Thujaplicata)
Tsuga heterophylla,
were
for
were
British
studied
their
communities.
data
obtained
Columbia,
Canyon,
Quantitative
epiphytic
for fourteen bryophyte
species by sampling at heights of 0.5 m, 1m and 2 m above ground level. Scapania
bolanderi
is an abundant
and widely
distributed
Bazzania
den?dala,
species; Dicranum
fuscescens,
Three
and
reptans
Lepidozia
Hypnum
are
circinale
also
common.
An
environmentally
uniform
study
area
was
the effects of factors other than bark and elevation on the distribution
of epiphytes. In
are
tree
terms
in
similar
of
their
but
ordination
methods
general,
species
epiphytic assemblages,
using
data exposed compositional
variation that may be explained by differences
inmicroclimate
and
quantitative
In this respect, the results point the way to further studies to examine epiphyte associations
in
bark-type.
to specific factors in the microenvironment.
closer detail, and to relate these associations
to minimize
chosen
Introduction
have long recognized that
Ecologists
and lichens are sensitive indicators of
environmental
variation, and thus may
of vegetational
delineation
units to a
with
possible
1973; Sj?gren,
Studies
1973).
bryophytes
small-scale
assist in the
degree not
vascular
1958,
(Barkman,
plants
& van der Maarel,
1961; Westhoff
have
shown
that
nonvascular
cryp
are useful
of site conditions
indicators
1953;
1969; Szczawinski,
(Hoffman & Kazmierski,
Jonescu,
1970), and may play an important role in
the cycling of nutrients through certain ecosystems
(Pike, 1971; Pike et al, 1977; Forman,
1975). In
togams
* Nomenclature
Cronquist
(1976)
hepatics.
** We
thank
&
Hitchcock
vascular
follows
plants
Schofield
follows
nomenclature
bryophyte
and Stotler & Crandall-Stotler
(1977) for
for
(1973);
for mosses
Dr.
W.
B.
Schofield
for aiding
in bryophyte
identification.
Vegetatio
? Dr. W.
$1.60.
45, 147-154 (1981). 0042-3106/81/0453-0147
Junk Publishers,
The Hague.
in The Netherlands.
Printed
are a significant
some areas, epiphytic bryophytes
of a site's vegetation
1960;
(Iwatsuki,
component
Scott, 1970, 1971; Pike et al, 1975).
A tree presents a variety of microhabitats
that are
open to colonization
by corticolous
species (Bark
man,
1958). Environmental
gradients
relating to
can be
changing
height, aspect and inclination
in epiphytic
changes
along the tree bole have
(Hale, 1952; Kershaw,
1964; Harris,
1971; Yarranton,
1972), and for
1975; Ras
1976; Rasmussen,
(Slack,
bryophytes
mussen & Hertig,
1977; Stringer & Stringer,
1974;
& Odani
Iwatsuki,
1960). Hosokawa
(1957) and
et al ( 1964) reported that light intensity
Hosokawa
increases and relative humidity
decreases
with
recognized.
Compositional
communities with elevation
been reported for lichens
increasing elevation, noting that epiphytes adapted
occur at higher levels. It
to more xeric conditions
has been suggested
that the bole may be divided
into height zones representing
unique combina
tions of habitat
factors and associated
vegetational
148
components
1958; Iwatsuki,
(Barkman,
1960);
the delineation
of such zones may be
however,
somewhat arbitrary (Pike et al, 1975).
also vary circumferential
Epiphytic communities
on
a
at
constant
level
the
bole. This variation
is
ly
usually correlated with aspect and inclination. The
effect of aspect tends to be greater in open stands
where differential
to wind and light is
exposure
more
pronounced
(Barkman,
1958); Yarranton
(1972) and Gough
(1975) reported compositional
changes with aspect for epiphytic lichens. In closed
inclination may be more
forests,
important. An
bark
surface
will
receive more
upward
facing
moisture
and stemflow than
through precipitation
the underside which may remain dry even after a
rainstorm (Sj?gren,
1961; Pike et al, 1975). Ras
mussen (1975) found differences
in epiphyte species
to
and
be
correlated with
composition
diversity
inclination.
between substrate (bark) and epi
Relationships
have
also
been examined. Barkman
phytes
(1958)
lists
relief,
hardness,
scaling,
water-holding
capac
ity, nutrient availability,
acidity, and chemistry as
factors of potential
influence. Stringer & Stringer
that bark texture was important
(1974) concluded
was
for lichen establishment,
but that microclimate
more important for bryophytes. Jesberger & Sheard
and
(1973) identified substrate, water availability,
stand age as important factors.
methods
Ordination
structure
in epiphytic
using polar ordination,
a
represents
mainly
have been used to examine
Beals
vegetation.
(1965),
that substrate
determined
continuous
environmental
& Sheard
(1973) reached a
in a study of eight tree species
In another
components
using principal
analysis.
results
from
study,
analysis
principal components
stress,
suggested that site factors including moisture
nutrient
im
status, and light regime were more
variable.
Jesberger
similar conclusion
the performance
portant than aspect in determining
of epiphytes (Sheard & Jonescu,
1974).
were
In the present study, ordination methods
to the
used to answer specific questions pertaining
structure
the epiphytic
of a wet
vegetation
in
south-coastal
coniferous
forest
British
temperate
tree
Columbia.
three
The distinguishability
of
to
with
their epiphytic communities,
respect
species
and compositional
changes with height above
ground, were examined with principal components
of
analysis and canonical vari?tes analysis. The extent
to which patterns of association
among epiphytes
are
similar
on
with principal
Study
tree
different
coordinates
was
species
examined
analysis.
area
The study area is located adjacent to the Capi
lano River gorge in North Vancouver
at an eleva
tion of about 150 m above sea level. Geologically,
the area is characterised
laid
by alluvial deposits
over a complex of
down during the last glaciation
rock (Armstrong,
plutonic and metamorphic
1968).
data for Cloverly
Meteorological
study
area),
suggest
tion of about
an
average
an
average
2000 mm, most
annual
(ca. 2 km from the
annual
precipita
of which
of
temperature
is rain, and
9.8
?C.
The
of the forest understorey
microclimate
is charac
terised by consistently
maintained
high humidity
partly by the closed forest structure, and partly by
the canyon
The area lies within
topography.
Coastal
Western
Hemlock
bio
Krajina's
(1959)
is similar to
zone, and the vegetation
geoclimatic
that of nearby Lynn Canyon, described by Krause
& Schofield
trees are Pseu
(1977). The dominant
menziesii
var. menziesii,
Franco
dotsuga
(Mirbel)
and Tsugaheterophylla(Rai.)
ThujaplicataDonn.,
Sarg.,
form
which
a
closed
relatively
ex
canopy
tending from about 20 m (where branching begins)
to about 50 m. Scattered understorey
trees include
Taxus brevifolia Nutt. and Acer circinatum Pursh.
A
discontinuous
theria
low
shallon
Pursh,
shrub
layer
Rubus
contains
Gaul
Pursh,
spectabilis
as well as depauper
Smith,
Vacciniumparvifolium
ate saplings of Tsuga heterophylla.
The ferns,
munitum
Polystichum
(Kaulf.) Presl, Blechnum
spicant (L.) Roth, and Athyrium filix-femina
(L.)
Roth,
are
bryophytes
Warnst.,
Dominant
forest-floor
loreus(Hedw.)
Plagiotheciumundulatum(Hcdw.)B.S.G.,
Hylocomium
number
conspicuous.
include Rhytidiadelphus
splendens
(Hedw.)
B.S.G.,
and
a
of Mniaceae.
There is evidence of logging at higher elevations in
the canyon, although the study area apparently was
not extensively
logged and for many years has been a
park.
Methods
A preliminary
survey of the epiphytic vegetation
indicated that inclination of the tree bole from the
149
Trees
grow.
in determining
is important
vertical
grow
rarely
exactly
where epiphytes
and
vertically
to the upper
are normally
confined
bryophytes
trunk.
Lichens
the
of
side
(mostly
(wetter)
are often
indeterminable
species of Cladonia)
the effect of
present on the lower side. To minimize
was
to
the
confined
inclination, sampling
upper side
of trees where bryophytes are most abundant.
The similarity
on
separate
for fourteen
estimates
bryophyte
species
in 216
quadrats.
Three
components
including principal
(PCA), principal coordinates
1978)
(P-Co- A, Gower, 1966;Orl?ci,
vari?tes analysis (CV?, Seal, 1964)
analysis
(axes) analysis
and canonical
were
used
methods
ordination
to
techniques,
the
analyse
these
data.
assume
methods
As
eigenvector
linear
relation
axes. This
variables and ordination
un
be
sometimes
may
ecologically
assumption
realistic thereby limiting the successful exposure of
vegetation
gradients that are linked to underlying
ships between
environmental
gradients; however, for summarizing
in a
information and displaying overall relationships
are appropriate
set of data, the methods
(Nichols,
1977).
Four PC A's were run, one on the data from each
of the three sampling heights,
and one on the
combined data for trees formed by averaging species
coverage
values
across
three
quadrats.
This
approach made possible the inspection of patterns of
variation within and between tree species at separate
and combined levels. Each analysis was based on an
unstandardized
covariance matrix for 14 species.
CVA was run after the data were arranged into
groups composed of quadrats from similar heights
on
each
tree
species.
Nine
quadrat-groups
were
to the nine height-species
formed, corresponding
axes calculated by
The ordination
combinations.
CVA
coincide
with
the lines of 'best' linear
discrimination
centroids;
among the quadrat-group
the
results
thus,
display an optimum configuration
of the data from which various aspects
can be inferred.
interrelationships
of group
relations among bryophyte
species
were
examined
with
species
P-Co
were
the analyses
run, bryophytes
in
less
than
five
occurring
quadrats on any tree
were
from
the data. Similarities
removed
species
between bryophytes were defined by the cos theta
(angular separation) function expressed as:
A.
Before
=
25 each of Thuja
A total of 72 trees comprising
and 22 Tsuga
and
menziesii,
Pseudotsuga
plicata
were
for
the
selected
study in a
heterophylla
trees
devoid of
fashion,
excluding
pseudorandom
bryophytes and those showing signs of disturbance
near trails. Trees were sampled at three height levels
(0.5 m, 1.0m, and 2.0 m from the base) using a 15 cm
cover
to obtain percentage
X
15 cm quadrat
tree
? XijXhj
j
J
and Xn- denote the cover recorded for
X?j
/and
h in quadrat/, and n is the total number
species
the expression relates to
of quadrats. Geometrically,
vectors in
the angle subtending
species positional
the
?-dimensional
thus,
lengths of
(quadrat) space;
the vectors, which are determined by the abundances
of species, do not influence the similarity measure.
programs for PCA and P-Co-A were
Computer
developed
by one of us (Bradfield); CVA was run
where
a program
from the MIDAS
statistical
All
&
Guire,
(Fox
1976).
analyses were
package
on
an
AMD
AHL
6-II
470
computer at
V/
performed
of British Columbia
the University
Computing
using
Centre.
Results
The mean
cover data for the fourteen
observed
epiphytic bryophytes
during sampling are
common species,
most
in
Table
1.
the
five
Of
given
is
the
bolanderi
Scapania
clearly
leading dominant
at all heights
Bazzania
examined.
den?dala,
and
bolanderi,
reptans show a
Scapania
Lepidozia
percent
tendency to decline in cover with increasing height;
trend. The
circinale shows the opposite
Hypnum
cover of Dicranum fuscescens
is highest on Tsuga
and lowest on Thuja plicata.
heterophylla
The remaining nine species occur with generally
low frequency and cover making
of
interpretation
their distributional
patterns more difficult. Species
such as Rhizomnium
glabrescens,
Plagiothecium
and Rhyti
undulatum,
Hylocomium
splendens,
are
on
common
the forest
loreus
diadelphus
fairly
on tree
but occasionally
grow epiphytically
more
a
common
Isothecium
is
bases;
stoloniferum
epiphyte at higher levels.
floor,
150
Table
percent cover data
and Tsuga heterophylla.
for epiphytic
1. Mean
menziesii
1m and 2 m on Thuja plicata,
at 0.5 m,
bryophytes
Thuja
Pseudotsuga
menziesii
=
(n
25)
plicata
=
25)
(n
Scapania
Dicranum
Bazzania
Aust.
bolanderi
.5m
lm
44.2
28.2
15.6
40.8
0.6
0.8
0.2
4.8
Turn.
fuscescens
den?dala
Hypnum
reptans
circinale
(L.) Dum.
Hook.
Plagiochila
porelloides
Nees) Linderb.
(Torrey
Ptilidium
(Aust.)
Rhizomnium
glabrescens
undulatum
Hylocomium
Cephalozia
splendens
lunulifolia
Lophocolea
heterophylla
loreus
Rhytidiadelphus
Underw.
.5 m
lm
33.4
21.2
40.2
40.0
32.7
7.0
21.4
28.2
25.0
11.8
2m
3.0
2.0
6.8
3.2
19.1
5.7
3.2
5.0
3.6
0.6
12.6
4.0
3.4
3.6
2.3
0.7
6.4
8.2
11.4
3.2
7.6
13.0
2.2
0.2
1.2
1.4
6.1
13.6
1.6
5.6
0.5
0.5
0.2
0.8
0.6
0.4
0.4
0.2
0.4
1.6
0.9
from principal
indicate
components
absence.
species
.5 m
analysis
1m
of the separate
2m
and combined
Combined
I
II
I
II
I
II
I
II
54.7%'
15.7%
40.(
33.6%
47.6%
25.3%
46.5%
27.0%
.957
.090
.780
.624
.902
.934
-.258
.056
.802
-.618
.773
.235
.802
.264
.955
.259
.340
.035
.006
.042
.021
.104
.007
.099
-.474
.056
.078
.009
.025
.050
.138
.029
.075
.064
.071
.359
.211
.025
.016
.038
.001
.001
.019
.006
.000
.002
.040
.017
.008
.005
.030
.014
-.003
.007
.011
.007
.041
.013
.013
.002
.004
.050
.004
.000
.009
.007
.006
.009
-.388
-.446
-.012
.082
.006
.009
.008
-.007
-.004
.002
-.002
.005
.002
.004
.003
.002
-.015
.141
-.005
.003
.006
.006
.011
.025
.016
.009
.004
.012
1 percent variation
explained
2 abbreviated
species
epiphyte
name
0.7
4.1
1.4
0.4
Dum.
2. Eigenvectors
levels data. Dashes
2.3
0.5
1.4
1.4
(Schrad). Dumort
(Hedw.) Warnst.
Table
0.4
2.4
B.S.G.
(Hedw.)
(Dum.)
1.6
0.4
0.4
Kop.
B.S.G.
(Hedw.)
Brid.
SCA BOL2
DIC FUS
BAZ DEN
LEP REP
HYPCIR
PLA POR
PTI CAL
RHI GLA
PLA UND
ISO STO
HYL SPL
CEP LUN
LOP HET
RHY LOR
2 m*
ex
(Kindb.)
Plagiothecium
Isothecium
stoloniferum
lm
7.4
0.2
californicum
Tsuga
heterophylla
=
(n
22)
ex Gott.
(Torrey
et al.) Trev.
Lepidozia
.5m
2m
Pseudotsuga
0.5
1.3
151
n
The results
and Dicranum
fuscescens.
a
exerts
that
microclimate
stronger influence
suggest
the distribution
of
in controlling
than bark-type
bolanderi
epiphytes.
vari?tes
Canonical
analysis
CVA gave results that were, in general, consistent
those of PCA; however, grouping the data into
categories proved to be instructive.
height-species
on the
The ordination
of quadrat group-centroids
with
A
D?
a &?
4
A A da
A
A
A
Fig.
I. Ordination
Symbols
(Tsuga
of
trees
on
components
analysis
code: a (Thujaplicata),
principal
first two canonical axes is shown in Figure 2. Four
aspects of the results are noteworthy. First, the axes
account
for 79% of the variation
among group
thus may be considered
and
successful
in
means,
in
the
data.
summarizing
major
relationships
is
of Tsuga heterophylla
Second, the distinctiveness
5-3
D
A
AAAAAD
AAA
?AA
D
the first
of
O
AA
two axes
the combined
D (Pseudotsuga
derived
levels
menziesii),
by
data.
O
heterophylla).
Principal
components
somewhat
since
over-emphasized
the
means
being influenced by a few trees growing along
ravine
the microenvironment
edge where
level sample
the
lowest
different.
Third,
are
the
is
of
because
analysis
Table 2 contains the eigenvectors pertaining to the
Since the
first two axes for each PCA performed.
results from separate analyses were considered to be
reasonably similar, only those from the PCA using
averaged data are discussed. In fact, this ordination
seemed to clarify trends noted when the sampling
scatter
The
separately.
heights were analysed
menziesii
appears distinctive
Pseudotsuga
of the combined
influence of a few rare (primarily
forest floor) species in this group (Table 1). Fourth,
this ordination, moreso
than the PCA's, demon
strates
variation
substrate
related
to changing
height
and
conditions.
21
".?-M
diagram for this analysis is shown in Figure 1.
The first axis explains about 47% of the variance
and is strongly influenced by the degree of Scapania
cover (increasing
from left to right).
bolanderi
tree
Considerable
overlap among
species is evident.
The second axis explains an additional 27% of the
variance and primarily displays a cover gradient of
to
Dicranum fuscescens
(increasing from bottom
.5-M
A 2"M
"l-M
"2-M
along the
species tend to be positioned
the
in
order, Thuja plicata
Pseudotsuga
gradient
A group of Tsuga
menziesii
Tsuga heterophylla.
trees growing along the ravine edge
heterophylla
which are exposed to increased light levels and river
toward the top of the second
spray are positioned
top). Tree
axis.
that
In general,
the PCA results demonstrate
is continuous
in the epiphytic vegetation
variation
related to cover changes in Scapania
and mainly
2-M
on the first two
of quadrat group centroids
Fig. 2. Ordination
axes derived
vari?tes
code: a
by canonical
analysis.
Symbols
m (Pseudotsuga
(Thuja plicata),
menziesii),
(Tsuga hetero
phylla).
152
THUJAPLICATA
SCA BOL
coordinates
Principal
analysis
Scatter
of epiphytes
obtained
diagrams
by
analysing the data from each tree species separately
with P-Co-A are presented in Figure 3. Measures
of
ordination efficiency, expressed as the proportion of
variance explained by the first two axes, range from
to 51% with Thuja
menziesii
37% with Pseudotsuga
plicata. In comparison with the quadrat ordinations
obtained with PCA and CVA, the P-Co-A
results
among epiphytes based on their
display associations
distributional
patterns; thus, the closeness of species
in the ordinations
reflects the degree to which they
were
in quadrats.
recorded
together
growing
DIC |FUS
-.8
PSEUDOTSUGA
MENZIESII
ISOSTd
Further insight into the strengths of the associations
may be gained by comparing the three sets of results
giving particular attention to the five most common
the leading
bryophyte
species. Scapania bolanderi,
occurs
dominant,
community
nearthecenter
of each
ordination; other species are arranged peripherally
with S. bolanderi
in accord with their associations
each other. Bazzania
and with
and
denudata
in cover with
reptans both decline
Lepidozia
a relatively
strong
increasing height,
exhibiting
on all tree species; Hypnum
association
circinale
DIC FUS
shows
higher
the opposite
its cover at
trend, increasing
levels, and is thus well separated from the
group.
preceding
Dicranum
fuscescens
varies
in the ordinations,
and no distinct
positionally
are
in
the
data (Table 1).
indicated
height preferences
It occurs commonly on both Pseudotsuga
menziesii
it tends to be
where
and Tsuga heterophylla,
associated with the higher levels group of species; on
LEP REP
CEPLUN
-.6
TSUGA HETEROPHYLLA
SCA BOL
from
it is markedly
Thuja plicata,
separated
rare
Ptilidium californicum
(both relatively
species)
in quadrats.
owing to their mutual exclusiveness
In general, the species ordinations
suggest that
among epiphytes are similar
patterns of association
on different
tree species. Microclimatic
variation
seems to be the overriding
influence in determining
where
wLEPREP
epiphytes
grow.
Discussion
Variation
PTI CAL
u-.7
of epiphyte
species on the first two axes
coordinates
by
analysis. Variance
explained
axis I, 26.6%, axis II, 24.7%; Pseudotsuga
axis
I, 17.9%, axis II, 16.8%; Tsuga heterophylla,
Fig. 3. Ordinations
derived by principal
axes: Thuja plicata,
menziesii,
I, 21.3%,
axis
axis
II, 19.2%.
changes
plicata,
in microclimate
and
bark
charac
results
in continuous
compositional
in the epiphytic
of Thuja
communities
teristics
Pseudotsuga
menziesii,
and
Tsuga
hetero
phylla at Capilano Canyon. These findings are in
studies
general agreement with those of comparable
of epiphytes done elsewhere (Beals, 1965; Jesberger
& Sheard,
1973; Szczawinsky,
1953). Concerning
153
bark factors, the physical nature of the substrate
appears to exert a stronger influence on epiphytes
than does its chemistry (average bark pH's of 4.4,4.2
for ten trees each of T.
and 4.4 were determined
P. menziesii,
plicata,
and
T. heterophylla)',
however,
accounts
bark chemistry
for the more
probably
the three
marked
observed
between
differences
conifers of this study and other tree species in the
area
as
such
Taxus
and
brevifolia
macro
Acer
is up to
phyllum. The bark of Pseudotsuga menziesii
30 cm thick on mature trees, is deeply fissured and
in plates when kept wet. The accumula
exfoliates
tion of decomposing
organic matter in fissures near
the tree base favours establishment
by bryophytes,
especially species such as Rhizomnium
glabrescens,
undulatum,
Plagiothecium
Hylocomium
splen
loreus which commonly
dens, and Rhytidiadelphus
occur on fallen logs. In this respect, the tree base
an extension
the forest
of
floor
represents
environment.
bark
is
Tsuga heterophylla
deeply
furrowed with flat, scaly ridges; exfoliation appears
to occur
only
on
trees
to continuous
exposed
river
spray. The bark of Thuja plicata is thin and fibrous,
and may be peeled off in long, vertical strips; the bark
surface is relatively smooth, with little fissuring.
Observations
suggest that the bark of all three
remains
moist for considerable periods after
species
wetting.
The
to variation
of epiphytes
in
sensitivity
of
to
factors
the
bark
be
appears
physical
mainly
an
growth form dependent. Dicranum fuscescens,
erect,
moss,
acrocarpous
tuft-forming
may
have
difficulty establishing on the relatively smooth bark
of Thuja plicata. A similar explanation may account
for
the decline
in cover
of
the
sub-erect,
tuft-forming
hepatic, Scapania bolanderi, at heights above 1m on
Thuja plicata. Also, forest floor species which often
colonize
the bases of Pseudotsuga
menziesii
and
occur
on
Tsuga heterophylla,
rarely
Thuja plicata.
In
the
contrast,
denudata
pleurocarpous
and
mat
Lepidozia
moss,
forming
reptans,
Hypnum
hepatics,
for
the growth
of many
the
tree
base
receives a
bryophytes.
splayed
of the total precipitation
than
greater proportion
parts of the bole which
higher
undoubtedly
wider
ranges in light and moisture
experience
conditions. Of the 14 epiphyte species recorded, the
show a decline in cover with increasing
majority
microenvironment
Also,
circinale and Isothecium
height; however, Hypnum
show
the
stoloniferum
opposite trend well beyond
the upper limit of sampling at 2 m. Both of these
species grow closely appressed
this growth form
Presumably,
more xeric conditions
and
surface at higher levels. Species
erect
to sub-erect
tufts,
bolanderi,
Dicranum
tree,
as
are
the
abundant
prostrate,
including
and
fuscescens,
are more
glabrescens,
to the bark surface.
is an adaptation to the
nearly vertical bark
which grow in loose
near
mat-forming
Scapania
Rhizomnium
the base
of
the
hepatics,
and Ptili
Lepidozia
reptans, Bazzania denudata,
dium californicum.
The
results also
show variation
to
related
on
a
microclimatic
scale
the
within
changes
larger
trees
study area. A group of Tsuga heterophylla
the
ravine
above
the
river
have
growing along
edge
been identified as supporting a distinctive epiphytic
assemblage. This appears to be caused by a different
set of microenvironmental
conditions
(river spray,
increased light) which,
in particular,
favours the
as
of
such
Dicranum
and
growth
species
fuscescens
bolanderi.
Scapania
The ordination methods
used have provided an
efficient summarization
of a complex body of data
with minimal
loss. Major properties of
information
the structure
and
inherent
of the
variability
have been exposed.
Several
epiphytic vegetation
for more
directions
detailed
of
the
study
between
com
relationships
epiphytes,
epiphytic
and
munities,
microenvironmental
are
factors
indicated.
Bazzania
and
the appressed,
seem
to be
circinale,
less sensitive to differences
in bark-type.
It seems likely that epiphytes are also responding
to a light-moisture
complex
gradient
operating
the bole. Microclimate
along
may change
sig
nificantly over a vertical distance of a few meters.
near
the ground,
Consistently
high humidity
maintained
the
shelter
afforded
partly by
by
neighbouring
understorey plants, creates a suitable
References
J. E., 1968. In:W. H. Mathews
for
(ed.) Guidebook
field trips in southwestern
British Columbia.
geological
Dept.
of Geology,
U. B. C, Report No. 6.
J. J.,
1958. Phytosociology
Barkman,
and
of
ecology
Van Gorcum,
cryptogamic
epiphytes.
Assen, Netherlands.
J. J., 1973. Synusial approaches
to classification,
Barkman,
In: R.
Armstrong,
H. Whittaker
Dr. W.
Junk,
of Vegetation
(ed.) Handbook
The Hague,
p. 435-491.
Science.
Vol.
5.
154
of some corticolous
1965. Ordination
Beals, E. W.,
cryptogamic
16: 1-8.
Oikos
Wisconsin.
in south-central
communities
lichens with blue-green
R. T. T., 1975. Canopy
algae: a
56:
source
rain forest. Ecology
in a Columbian
nitrogen
1176-1184.
3rd
for MIDAS.
Fox, D. J. & Guire, K. E., 1976. Documentation
Forman,
Univ. of Michigan.
Research
edition. Statistical
Laboratory,
on Pseudotsuga
distributions
L. P., 1975. Cryptogam
Gough,
Boulder
in the Front Range,
and Abies
menziesii
lasiocarpa
78: 124-145.
Colorado.
Bryologist
County,
325-338.
Harris, G. P., 1971. The ecology
on oak and birch
zonation
of corticolous
in South
in a virgin
lichens.
Devon.
I. The
J. Ecol.
59:
431-439.
1973. Flora of the Pacific
C. L. & Cronquist,
A.,
Hitchcock,
of Washington
Northwest.
Press, Seattle.
University
R. G., 1969. An ecological
G. R. & Kazmierski,
Hoffman,
study
men
and lichens on Pseudotsuga
of epiphytic
bryophytes
on the Olympic
I. A descrip
Peninsula,
Washington.
72: 1-19.
tion of the vegetation.
Bryologist
T. & Odani,
1957. The daily compensation
N.,
Hosokawa,
in a beech forest. J.
ranges of epiphytes
period and vertical
ziesii
Ecol.
45: 901-915.
H., 1964. Causality
T., Odani, N. & Tagawa,
Hosokawa,
in forests with
of corticolous
distribution
species
to the physioecological
reference
approach.
of the
special
67:
Bryologist
396-411.
Japan.
bryophyte
159-352.
communities
in
185-201.
1970.
E.,
Canada.
Lichens
Bryologist
on
tremuloides
Populus
73: 557-578.
in
on the distribu
K. A., 1964. Preliminary
observations
Kershaw,
lichens inWales.
tion and ecology of epiphytic
Lichenologist
2: 263-276.
V.
Krajina,
relations
interspecific
phytosociological
Rev. Bryol. Lichenol.
communities.
phyte
zones
J., 1959. Bioclimatic
series 1: 1-47.
Botanical
U.B.C.,
G. & Schofield,
Canyon Park, North
Krause,
W.
B.,
Vancouver,
in British
Columbia.
1977. The moss
flora of Lynn
British Columbia.
Syesis 10:
97-110.
Syesis 9: 317-354.
1970. Vegetation
Scott, G. A. M.,
Part
II. Epiphytic
Fiordland.
vegetation
30-50.
on
the northern
S.,
on Secretary
studies
Island,
and
ground
cryptogamic
New Zealand
J. Bot. 8:
slopes.
Sheard, J. W. & Jonescu, M. E., 1974. A multivariate
analysis of
of lichens on Populus
the distribution
in west
tremuloides
central Canada.
77: 514-550.
Bryologist
1961. Epiphytische
Sj?gren,E.,
Moosvegetation
Acta Phytogeographica
der Insel ?land.
summary).
Slack, N. G.,
Stotler,
1976. Host
North
America.
in Laubw?ldern
Suecica
44 (english
of bryophytic
specificity
epiphytes
J. Hattori
Bot. Lab. 41: 107-132.
R. & Crandall-Stotler,
and hornworts
in
1977. A checklist
of the
B.,
of North
America.
Bryologist
80: 405-428.
Stringer, P. W. & Stringer, M. H. L., 1974. A quantitative
study of
of Winnipeg,
in the vicinity
corticolous
bryophytes
77: 551-560.
Manitoba.
Brylogist
and
1953. Corticolous
A.,
Szczawinski,
lignicolous
plant
in the forest associations
of the Douglas-fir
communities
forest
on Vancouver
Island.
and Biology,
U.B.C.
V. & van der Maarel,
Westhoff,
In R. H. Whittaker
approach.
Ph.D.
thesis,
Department
Botany
tion Science.
Yarranton,
lichens
Vol.
1977. On
analysis
Pike,
Accepted
5. Dr. W.
E.,
1973. The
Braun-Blanquet
of Vegeta
(ed.), Handbook
Junk, The Hague,
G. A., 1972. Distribution
and succession
on black spruce near Cochrane,
Ontario.
75: 462-480.
L. H.,
1971. The role of epiphytic
lichens and mosses
in
and nutrient cycling in an oak forest. Ph. D. thesis,
production
of Biology,
Univ. of Oregon.
Dept.
in epiphytic
bryo
43: 207-217.
in the quantitative
Scott, G. A. M., 1971. Some problems
ecology
New Zealand
of bryophytes.
J. Bot. 9: 744-749.
for biologists.
statistical analysis
Seal, H. L., 1964. Multivariate
London.
Methuen,
the interpretation
of principal components
in ecological
contexts.
34: 191-197.
Vegetatio
in vegetation
research. 2nd
Orl?ci, L, 1978. Multivariate
analysis
ed. Dr. W. Junk, The Hague.
Nichols,
in western
W. B., 1976. Bryophytes
of British Columbia.
III.
Schofield,
and distributional
Habitat
information
for selected mosses.
liverworts
1973. A quantitative
J. A. & Sheard, J. W.,
study and
Jesberger,
lichen communities
in
of corticolous
multivariate
analysis
Can. J. Bot. 51:
the southern boreal forest of Saskatchewan.
M.
Jonescu,
west-central
conifers
1977. A400-year-old
Pike, L. H., Rydell, R. A. & Denison, W. C,
fir tree and its epiphytes:
surface area, and
biomass,
Douglas
their distributions.
Can. J. For. Res. 7: 680-699.
eastern
1960. The epiphytic
Z.,
Bot. Lab. 22:
J. Hattori
Iwatsuki,
M. A. &
Sherwood,
of epiphytic
lichens
in the
Rasmussen,
L., 1975. The bryophytic
epiphyte vegetation
forest Slotved Skov, Northern
Jutland. Lindbergia3:
15-38.
L. & Hertig,
of
Rasmussen,
J., 1977. Statistical
investigation
of latent root and
1966. Some distance
J. C,
Gower,
properties
53:
used inmultivariate
vector methods
analysis. Biometrika
of cryptogams
Hale, M. E., 1952. Vertical distribution
33: 398-406.
forest of Wisconsin.
Ecology
L. H., Denison,
W. C, Tracy, D. M.,
F. M.,
1975. Floristic
Rhoades,
survey
on old-growth
and bryophytes
growing
78: 391-404.
Oregon.
Bryologist
Pike,
8.1.1981.
p. 617-726.
of epiphytic
Bryologist
of