Estimating Epiphyte Abundance in High-Elevation Forests of Northern British Columbia

7 Pages·1999·1.4 MB·

by  Campbell, Jocelyn

Selbyana 20(2):261-267. 1999.
ESTIMATING EPIPHYTE ABUNDANCE IN HIGH-ELEVATION FORESTS
OF NORTHERN BRITISH COLUMBIA
JOCELYN CAMPBELL
University of Northern British Columbia, 3333 University Way, Prince George,
British Columbia, V2N 4Z9. E-mail: [email protected]
SUSAN K. STEVENSON
Silvifauna Research, 101 Burden St., Prince George, British Columbia, V2M 2G8.
E-mail: [email protected]
DARWYN S. COXSON*
University of Northern British Columbia, 3333 University Way, Prince George,
British Columbia, V2N 4Z9. E-mail: [email protected]
ABSTRACT. We characterized the distribution patterns of arboreal lichens in a wet temperate spruce-fir
(Picea engelmanniilAbies lasiocarpa) rain forest in east-central British Columbia. Using the single rope
technique for canopy access, we employed the "clump method" to estimate the abundance of arboreal
lichens in three functional groups: Alectoria, Bryoria, and foliose lichens. Verification was achieved through
a regression of those estimates with lichen biomass measured on a subset of branches. We used a probability
proportional to prediction sampling scheme such that the probability that a branch would be selected for
measurement was proportional to its estim.ated lichen abundance. Used in this way, the clump m.ethod is a
reliable estimator of lichen abundance that can be used to accurately describe the distribution of lichen
biomass in tree crowns.
Key words: canopy, epiphytes, lichens, high elevation forest, Engelmann spruce, subalpine fir
RESUMEN. Hemos caracterizado la distribuci6n de muestras de Ifquenes arb6reos en un abeto-pino (Picea
engelmanniilAbies lasiocarpa), humedo, tempi ado bosque al este de British Columbia central. Usando el
metodo "single rope," tecnica de cuerda simple, aplicamos el metodo de grupo para estimar la abundancia
de Ifquenes ab6reos en tres grupos funcionales; Alectoria, Bryoria y liquenes frondoso. La compr6bacion
fue realizada a traves de una regresi6n de estimaciones con Ifquenes mediante una prueba aparte (sub
sample) obtenido a traves de probabilidad proporcional a pruebas pronosticas. La evaluaci6n de la com
binaci6n de esas tecnicas muestra que el metodo grupo (clump method) es un estimador seguro de Ifquenes
abundantes que se puede usar para producir un selecci6n datos objectivo que describe con exactitud la
distribuci6n de epifita Ifquenes de biomasa.
INTRODUCTION vious from the ground (Pike et al. 1975) and an
inventory of the lower canopy and trunk pro
The study of epiphytic lichen biomass in the vides an incomplete view of the total ecosystem
upper canopy of high elevation Engelmann (McCune 1990). Thus, as much effort is often
spruce-subalpine fir stands has previously been dedicated to accessing the canopy as is devoted
limited to studies of felled trees (Edwards et al. to studying the communities within the canopy
1960) and studies of lichens within a few meters (Moffett and Lowman 1995). The rewards of
of the forest floor (Stevenson 1979, Van Daele such effort are found when the canopy organ
and Johnson 1983, Detrick 1985, Rominger et isms may be studied with as much detail as ter
al. 1994). This is perhaps due to the difficulties restrial species. This, in tum, increases our con
involved with accessing the communities in the fidence in the reliability of estimation methods
upper canopy coupled with the intimidating task at each level in the canopy.
of estimating the biomass, which, in some areas, Methods of assessing epiphytic lichen bio
exceeds 3000 kg/ha (Edwards et al. 1960). mass range from obtaining biomass by tediously
Access to the canopy is an integral part of dissecting and weighing each sample to tradi
studying epiphytic communities. The mass of tional cover estimation techniques using quad
epiphytes on the upper branches is often not ob- rats and transects. Direct biomass sampling is
too time-consuming to be used in an inventory
study in which the entire community Il).ust be
* Corresponding author. quantified. Traditional methods that are gener-
261
262 SELBYANA Volume 20(2) 1999
ally used in terrestrial plant studies are rapid but rounded by an unspecified number of shorter
are not easily applied or verified in a three di layers, these clumps physically separated one
mensional environment (Boucher and Stone from another;
1992, McCune 1990). Quicker, more specialized b) disturbance vectors--eligible trees were lo
techniques were therefore necessary to describe cated in undisturbed forest cover, at least two
arboreal lichens in this study. We describe the tree lengths (ca. 60 m) from man-made open
use of the "clump method," an estimation meth ings (roads, harvest openings);
od that was originally developed to quantify li c) safety concems--eligible trees were not in
chen biomass in the lower canopy (Stevenson immediate proximity to standing dead trees
1979), to visually assess lichen abundance (snags) which would pose a hazard to re
throughout tree crowns. We also evaluate the ef search personnel and were otherwise deemed
fectiveness of verifying our estimates with a safe to ascend (no heart rot, major cracks or
sampling scheme in which the branches with breaks in the main trunk, etc.);
high lichen abundance are sampled more inten d) ecological analogues-all trees were located
sively than those with low lichen abundance. within the same subzone of the subalpine
spruce-fir forest zone, on south-facing slopes
MATERIALS AND METHODS with less than < 10° inclination, between
1450 and 1475 m elevation a.sJ, and greater
Study Area
than 2 m in height.
We conducted the research in the Bowron Eight groups of trees were randomly selected
River drainage near Pinkerton Mountain in east from within the 4 ha study area that met the
central British Columbia. The study area above noted criteria. A set of 55 trees within
(53°38'N, 121°25'W) is approximately 90 kID these clumps made up the sample tree popula
ESE of Prince George on a gentle south-facing tion. Seven additional trees were also included
slope at 1450-1475 m in the Engelmann Spruce in the data set, these representing individual
Subalpine Fir Zone (Meidinger and Pojar 1991). trees that had a solitary growth habit (not grow
The climate in the study area is cold and wet. It ing in clumps) and were located within 30 m of
is within the range of woodland caribou (Ran the above selected clumps. All selected trees
gifer tarandus caribou), a species at risk in east were rigged, climbed, and the amount of lichen
central British Columbia that depends on the li was estimated.
chens Bryoria spp. andAlectoria sarmentosa for
winter forage. Canopy Access
The coniferous forest is dominated by subal
We reviewed the existing canopy access tech
pine fir (Abies lasiocarpa) and Engelmann
niques (Denison 1973, McCarthy 1988, Perry
spruce (Picea engelmannii), many of which
grow in close association with one another, and Williams 1981, Ring and Winchester 1996,
Thcker and Powell 1991) and judged that most
forming clumps separated by small natural open
were more complicated than was necessary to'
ings. Mature trees were 20-30 m tall with char
achieve a safe and reliable working environment
acteristically narrow canopies. Branches were 2-
in our study area. We selected the single rope
4 m long but decrease in length with increasing
access technique described by Perry (1978) as it
height. They were downwardly sloped and usu
was the most suitable for the requirements of our
ally spaced in clusters of four or more approxi
research in the subalpine canopy where the nar
mately every 0.5-1.0 m up the trunk. The den
row crowns and dense branches created prob
sity of branches increased markedly with in
lems in maneuvering around the trunk.
creasing height (Campbell 1998).
Using a single rope for ascending and de
scending from canopy proved to be quite suc
Sample Selection
cessful. However, it was not possible to make
The study area is the site of long-term re detailed observations in high elevation trees
search on the effects of partial cutting on various from the main climbing line. The brarich density
resource values and ecological processes, in made it difficult to reach every limb. A second
cluding caribou habitat and epiphyte dynamics. ary system (lanyard) commonly used by tree
Study trees were randomly selected from a pool climbers was employed. It allowed us to move
of representative trees. To be included in this around to all parts of the tree while remaining
population, trees had to meet the following cri securely attached to the trunk.
teria:
Epiphyte Biomass Estimation Methods
a) canopy architecture-trees were located in
characteristic groups (clumps) of trees, com Epiphytes were divided into three functional
prising a taller stratum or layer of trees sur- groups: Alectoria, Bryoria, and foliose lichens.
CAMPBELL ET AL.: EPIPHYTE ABUNDANCE IN HIGH-ELEVATION FORESTS 263
The Alectoria functional group was composed
almost entirely of Alectoria sarmentosa, al
though Usnea spp., which are extremely rare in
the study area, were included if they were pre
sent. The Bryoria functional group comprised
Nodobryoria oregana and a number of Bryoria
species, primarily Bryoria capillaris, B. fremon
tii, B. fuscescens, B. glabra, and B. pseudofus
cescens. The foliose lichen functional group was
dominated by Hypogymnia imshaugii, H. me
taphysodes, H. occidentalis, H. physodes, H. tu
bulosa, Parmelia sulcata, Platismatia glauca,
and Tuckermannopsis platyphylla.
Those groups were selected because of their
functional and ecological importance in the
study area. Caribou use both Alectoria and Bryo
ria, but prefer the latter (Rominger et al. 1996).
Foliose nCI[lell
The two fruticose groups differ in their distri
bution in the canopy, and are expected to re (1.5 g)
spond differently to forestry activities that alter
the structure of the canopy. Foliose lichens con
taining green algal symbionts are abundant in
the study area, but are not known as important
forage items for wildlife. Epiphytic bryophytes
and cyanolichens were not included because
they are rare or absent in our study area. Crus
tose lichens were excluded because of time con
straints.
The abundance of each functional group on
each branch was estimated using the "clump FIGURE 1. The standard clump and card used for
method," a technique that facilitates estimation lichen biomass estimation at Pinkerton Mountain. The
of biomass on a three dimensional substrate clump of fruticose lichen (Alectoria and Bryoria) is
(Stevenson 1979). The lichen on a given branch 2.5 g, of which 30% is Bryoria and the card of foliose
lichen weighs 1.5 g.
was compared to a standard unit of lichen of
known size and weight; biomass was estimated
according to how many multiples of the standard made by the same researcher to minimize inter
unit that it represented (FIGURE 1). For both fru observer error and to maintain the high level of
ticose lichen functional groups (Alectoria and consistency required with this technique. As
Bryoria) a 2.5 g clump of lichen was used as a well, observers calibrated their estimates
standard and the total biomass of fruticose li throughout the field season by practicing togeth
chen was estimated as multiples of that weight. er. Accuracy and precision were tested by com
Alectoria and Bryoria were subsequently parti paring the estimates to dissected samples ob
tioned into separate functional groups by giving tained by sampling with a probability of selec
an estimate of the percentage of the total fruti tion that was proportional to estimated biomass.
cose lichen that was made up of Bryoria. A sim
ilar method was used for the foliose lichen func
Verification of Estimates
tional group. The three dimensional clump of
fruticose lichen was modified to a "card" of fo Because biomass sampling is time-consum
liose lichen of known dimensions. The standard ing, we aimed at a sample size of 5% of the
was an 8.5 X 8 cm surface which was the equiv branches on each sampled tree. Random sam
alent of approximately 1.5 g of foliose lichen. pling with a relatively small sample size is un
The foliose lichen on each branch was then as likely to adequately represent the branches with
sessed according to how many multiples of that high lichen biomass, because the frequency dis
surface area (card) it represented. In many cases tribution of lichen biomass on branches is
foliose lichen grew in horizontal layers, forming strongly skewed right. We used Probability Pro
mats. In such cases, the number of cards re portional to Prediction (3P) sampling (Cochrane
quired to make the mat was estimated by the 1977), a method similar to Probability Propor
observer. tional to Size (PPS) sampling as described by
All biomass estimations in one tree were Pike et al. This provided better representation of
264 SELBYANA Volume 20(2) 1999
branches with high biomass loading, as com RESULTS
pared to a strictly random sampling of canopy
branches. Branches were chosen for inclusion in Our measurements of lichen biomass were
the sample in direct proportion to the estimated based on a combination of the clump method
and 3P sampling. The standard errors of estimate
biomass of lichen found on them. The resulting
(SEM) indicate that there was some variability
sample thus has a greater proportion of branches
in the difference between estimated and mea
with a large amount of lichen than would occur
sured biomass. However, the strong coefficients
through random sampling.
of determination for the regressions between es
Branch selection was achieved in three steps.
timated and measured biomass (FIGURE 2) and
First, the number of clumps of fruticose lichen
large effect sizes for each functional group
on each branch of a tree was estimated, as de
(Alectoria: R2 = 0.68, F = 2.12, SEM = 6.41;
scribed above. Second, the sample branches for Bryoria: R2 = 0.76, F = 3.14, SEM = 19.63;
that tree were selected, using a method borrowed foliose lichen: R2 = 0.61, F = 1.58, SEM =
from timber cruising (Dilworth and Bell 1971).
7.53) showed that the clump method coupled
For each branch, a random number was drawn
with 3P sampling for verification was an effec
between 1 and an upper limit. The formula for
tive tool for assessing lichen biomass at the tree
choosing that top random number, T, was:
level.
There were no significant differences in the
T = Sx/s quality of the estimates on branches measured
early versus late in the season or on branches
where T was the top random number, Sx was the with low versus high abundance (R2 difference
sum of the estimates of total fruticose lichen for test, P < 0.05; Fox 1997). That indicates that
the entire tree, and s was the expected sample the clump method can be effectively applied
size, which was 5% of the branches on the tree. with little practice time throughout the canopy.
The estimate of fruticose lichen on each branch It was, however, necessary to "calibrate" the es
was then compared to a random number less timates of the observers at the beginning of the
than or equal to the top random number. The field season, based on side-by-side comparisons
branch was selected for sampling if the estimat using standardized lichen clumps of known
ed biomass was greater than or equal to the ran weight, to ensure consistency within the re
dom number. search team.
Sometimes, because of the skewed frequency The application of the clump estimation tech
distribution of the lichen estimates, T was small niques in sub-alpine spruce-fir forest at Pinker
er than one or more of the largest single esti ton Mountain demonstrates a strong height re
mates on a tree. In that case, the formula was lated vertical zonation for each of the three func
modified so that T was doubled and each branch tion groups (FIGURE 3). Members of the Bryoria
on that tree was given two chances to be sam functional group reach greatest abundance in the
upper canopy, up to 30 g/branch, at heights
pled. If a branch was selected by both random
above 15 m in the canopy. Alectoria thalli, in
numbers then it was counted twice in the cal
contrast show greatest abundance in the lower
culations (Stevenson 1979). The third step was
canopy, ca. 5 m above the forest floor. Foliose
to return to the tree and remove all the lichen
lichens are represented are represented at all
from the sample branches.
height stratums, although reaching greatest
Natural log (Bryoria + 1) and log (Alectoria
abundance in the lower canopy.
+ 1 and foliose + 1) transformations were ap
plied to the measured and estimated lichen abun
dance values, to which a constant of 1 had been DISCUSSION
added to avoid taking the logarithm of zero.
The single rope technique was successfully
Simple regression models were calculated for
applied in the high elevation forests of east cen
each functional group to describe the relation
tral British Columbia. The equipment was rela
ship between the measured and estimated lichen
tively inexpensive and readily available from
on the 5% sample. The resulting (In-In or log many outdoor equipment stores, could be carried
log) regression equations were used to calculate and easily assembled by one person, and afford
a corrected lichen biomass estimate for each ed rapid access.
functional group on each branch on each study The success of the access techniques allowed
tree, including the trees that had not been sub us to apply the clump method confidently to
jected to destructive sampling. Estimates were each branch individually. Stevenson (1979) ap
converted back to g/branch by taking the anti plied the clump method to both highly visible
log/ln of the output of the regression equation. and poorly visible branches and found that co-
CAMPBELL ET AL.: EPIPHYTE ABUNDANCE IN HIGH-ELEVATION FORESTS 265
en 2.0 •
en
'" • 25
E •
i0D-_5 1.5 • • •
.2~u ._0~ 1.0 • • • • • 20
-~.:O 0>> g
"0 0 0.5 •
.:'e~"nJ, d • • • E
::;; 0.0 'iOjjJ 15
:x:
0.0 0.5 1.0 1.5
Estimated Alectoria Biomass (Log g/branch) ..r::
0c
!!! 10
III
-= Alectoria
•
Bryoria
• 5 == Foliose
•
0
0 5 10 15 20 25 30 35
Lichen Biomass (g/branch)
FIGURE 3. Height related vertical zonation of li
2 4
Estimated Bryoria Biomass (Ln gibranch) chen biomass at 1 m class intervals (Mean and Stan
dard Error) for Alectoria, Bryoria, and Foliose func
tional groups in subalpine spruce-fir forest at Pinkerton
Mountain, B.C.
2.0,------------------,
IccInnI -. • . . ••
E vented the evaluation of the very top of each
.2 _ 1.5
1Il.<: ...... - tree. However, since the branches in this region
~ g • II.' •• ••
o ~ 1.0 were primarily young, most had a diameter be
~~ • lei • : ••••• low our study cut off (2 cm) and would not have
~,.,3 0.5 •• •• • been included in the sample. Thus, this omission
::IcQ;In;I) 0.0 • • • • eosntliym saltiegsh. tIlny aafdfdeicttieodn ,o suirn coev ethrael lv laiscth mena jboiroimtya ossf
lichen in that portion of the tree was Bryoria,
0.0 0.5 1.0 1.5
the impact of excluding these branches on Alec
Estimated Foliose Biomass (Log g/branch)
toria and foliose lichen was negligible.
FIGURE 2. Regression lines for the clump method We found that the clump method was well
as an estimator of lichen abundance. Measured lichen
suited for the lichen community structure of our
(Y) is the biomass of the dissected and weighed sam
study site in two ways. It allowed rapid esti
ples obtained through 3P sampling, and estimated li
mations of all three lichen functional groups on
chen (X) is the abundance of lichen estimated using
the clump method on the same sample as above. Re each branch. This was necessary both because
gression statistics were performed on log and In trans this was an inventory study and because our
formed data. A. Alectoria: Log Y = 0.0755 + 0.992 scheme for verifying and correcting our esti
log X (R2 (N = 94) = 67.9%). B. Bryoria: Ln Y = mates required that the entire population be as
0.790 + 1.l02 In X (R2 (N = 94) = 76.1 %). (C) Fo sessed. Also, we found that even though esti
liose: Log Y = 0.l36 + 0.969 log X (R2 (N = 94) = mates could vary by as much as two orders of
61.3%).
magnitude, we were able to use a single refer
ence clump for fruticose lichens. We were also
able to use a single card as a standard for foliose
efficients of variation ranged from 15% to 85%. lichens in this ecosystem in which the branches
Coefficients were lowest for branches that were were small and easily seen. Although o,ur eco
most easily observed and not concealed by oth system required only one size standard, a refer
ers. Combining the clump method with tree ence clump can be tailored in size to suit the
climbing techniques ensured that every branch community, and several reference clumps of dif
in this study was highly visible. ferent sizes can be used if necessary. This flex
The single rope technique provided near total ibility makes the method applicable to a wide
access to the tree crown. As with other canopy range of ecosystems.
studies (Pike et al. 1977), safety concerns pre- The efficacy of the clump method as an esti-
266 SELBYANA Volume 20(2) 1999
mator varied among our lichen functional with the sub-sampled biomass. The ratio method
groups. There are two reasons why this variation is more suitable for intensive studies of one or
might exist. First, the standard clump was large a few trees, because correction ratios are calcu
and therefore more easily compared to lichen in lated for the whole population that is sampled
large amounts. Bryoria makes up the bulk of with probability proportional to prediction. In
epiphyte abundance and is therefore more ac studies that involve climbing, that population
curately estimated using a large standard than would normally be a single tree. It would be
either of the less abundant functional groups. impractical to estimate abundance on the entire
Second, the 3P sampling was calculated based population of trees, select the sample branches,
on the estimates of total fruticose lichen (Alec and return to the trees to collect the samples.
toria and Bryoria combined), the majority of Thus, the number of brandIes sampled on each
which is Bryoria. In essence, then, the selection individual tree would have to be sufficient to
of samples was based on the sum of estimates achieve a reasonable confidence interval when
of Bryoria and therefore has a larger proportion using ratios. By using regression statistics, we
of branches heavily laden with Bryoria than were able to pool the data from all our study
those with large loads of either Alectoria or fo trees, and apply a single regression for each
liose lichen. The larger coefficient of determi functional group to all branches.
nation for Bryoria reflects this bias. Sampling The efficacy of these clump estimation tech
according to Bryoria abundance was, however, niques is demonstrated by the fine resolution
still a reasonable estimator of both Alectoria (R2 data obtained for height related vertical zonation
= 67.9%) and foliose (R2 = 61.3%). It appears of lichen functional groups in subalpine forest
advantageous to select branches for sampling ac at Pinkerton Mountain. Lichens in the Bryoria
cording to the most important or most abundant functional group reach their greatest abundance
attribute in the context of the study (Le., caribou
in the upper canopy, in marked contrast to that
forage lichen).
of the Alectoria functional group lichens. Al
By estimating lichen abundance on all branch
though Alectoria and Bryoria spp. lichens have
es, measuring lichen abundance on some
previously been grouped together within an
branches, and using the relationship between es
Alectorioid functional group (e.g., McCune
timated and measured values to correct the es
1990) our present findings demonstrate quite dif
timates, we arrived at a reasonably accurate and
ferent ecological affinities for these two genera,
objectively collected data set. The variation in
at least within the context of these moist-cool
the relationship between estimated and mea
subalpine forests.
sured biomass, which was reflected in the SEM,
This differential response of these two lichen
is a weakness in any epiphytic lichen estimation
groups may reflect changes in both substrate
technique that occurs because it is not logisti
characteristics and operating environments for
cally possible to directly measure every branch
lichen growth, especially parameters of canopy
in the population. Our technique resulted in a
microclimate. Goward (1998) hypothesizes that
relationship between measured and estimated li
Bryoria lichen communities cannot tolerate ex
chen biomass that was comparable to those re
tended wetting episodes, and thus are favored in
sulting from many of the existing methods. The
upper canopy aspects. We are presently profiling
coefficients of determination from the regression
canopy microclimate in context of measures of
equations calculated for the three functional
lichen physiology to verify the applicability of
groups varied from 0.76 for Bryoria to 0.61 for
foliose lichens. McCune (1990) estimated lichen this hypothesis. A more detailed examination of
biomass using a variety of estimators. Resulting canopy structural variables for this site is pro
coefficients of determination were 0.58 < R2 < vided in Campbell (1998).
0.91, 0.30 < R2 < 0.96 and 0.33 < R2 < 0.91 In summary, the combination of the clump
using cover classes, maximum thallus length, method, 3P sampling, and regression statistics
and maximum thallus width, respectively. Al allowed us to characterize lichen abundance and
though our estimator was not as closely related distribution on a large number of trees. These
to the actual lichen biomass as in the best cases techniques may be effectively applied where de
with previously used estimators, it is uncertain scriptions of lichen biomass at the stand level
whether cover classes and maximum measures are required. In the subalpine spruce-fir forests
for each functional group would be easily ap of the present study area these measures show
plied in ecosystems where Alectoria and Bryoria that Alectoria and Bryoria lichens occupy dif
thalli are abundant and intertwined (Stevenson ferent niches within the canopy, and suggest that
and Enns 1993). traditional groupings of these genera into a com
Regressions, rather than ratios (Pike 1981), mon Alectoriod functional group may not be ap
were used in this study to compare the estimates propriate in these cool wet ecosystems.
CAMPBELL ET AL.: EPIPHYTE ABUNDANCE IN HIGH-ELEVATION FORESTS 267
ACKNOWLEDGMENTS Meidinger, D. and J. Pojar. 1991. Ecosystems of Brit
ish Columbia. B.c. Min. Forests, Victoria.
Moffett, M.M. and M.D. Lowman. 1995. Canopy ac
We thank K. Jordan, J. Clement, J. McGann,
cess techniques. Pp. 3-26 in M.D. Lowman and
and T. Rambo for assistance with canopy access
N.M.Nadkarni, eds. Forest Canopies. Academic
and data collection; J. Marsh for species deter
Press, San Diego.
minations; and S. Wawryszyn, S. Gibson, K. Perry, D. 1978. A method of access into the crowns
Stanko, P. Byman and C. Ford for biomass sort of emergent canopy trees. Biotropica 10(20): 155-
ing. Project funding was provided by the Natural 157.
Sciences and Engineering Research Council of Perry, D. and J. Williams. 1981. The tropical rainforest
Canada, Forest Renewal British Columbia, and canopy: a method providing total access. Biotro
the British Columbia Ministry of Forests. pica 13(4): 238-285.
Pike, L.H. 1981. Estimation of lichen biomass and pro
duction with special reference to the use of ratios.
Pp. 533-552 in D.T. Wicklow and G.C. Carroll,
LITERATURE CITED
eds. The Fungal Community: Its Organization and
Role in the Community. Marcel Dekker, New
Boucher, v.L. and D.E Stone. 1992. Epiphytic lichen York.
biomass. In G.c. Carrol and D.T. Wicklow, eds. Pike, L.H., W.C. Denison, D.M. Tracy, M.A. Sher
The Fungal Community: Its Organization and wood and EM. Rhoades. 1975. Floristic survey of
Role in the Ecosystem, 2nd ed. Marcel Dekker, epiphytic lichens and bryophytes growing on old
New York. growth conifers in W. Oregon. Bryologist. 78(4):
Campbell, J. "Canopy Research in North-Central Brit 389-402.
ish Columbia: An Exploration of Lichen Com Pike, L.H., R.A. Rydell and W.e. Denison. 1977. A
munities." M.Sc. thesis, Univ. N. British Colum 400-year-old Douglas Fir tree and its epiphytes:
bia. Prince George, British Columbia, Canada, biomass, surface area and their distributions. Ca
1998. nad. J. Forest Res. 7: 680-699.
Cochran, W.O. 1977. Sampling Techniques, 3rd ed. Rin, R. and N.N. Winchester. 1996. Coastal temperate
John Wiley, New York. rainforest canopy access systems in British Co
Denison, W.C. 1973. Life in tall trees. Sci. Amer. 228: lumbia, Canada. Selbyana 17: 22-26.
74-80. Rominger, E.M., L. Allen-Johnson and J.L. Oldemeyer.
Detrick, R.W.T. "Effects of Fire and Logging on Ar 1994. Arboreal lichen in uncut and partially cut
boreal Lichen Availability to Caribou." M.Sc. subalpine fir stands in woodland caribou habitat,
thesis, Univ. Idaho, Moscow, Idaho, 1985. northern Idaho and southeastern British Columbia.
Dilworth, J.R. and J.E Bell. 1971. Variable Probability Forest Ecol. Managem. 70: 195-202.
Sampling-Variable Plot and Three-p. Oregon Rominger, E.M., C.T. Robbins and M.A. Evans. 1996.
State University Book Stores, Corvallis. Winter foraging ecology of woodland caribou in
Edwards, R.Y., J. Soos and R.W. Ritcey. 1960. Quan northeastern Washington. J. Wild!. Managem.
titative observations on epidendric lichens used as 60(4): 719-728.
food by caribou. Ecology 41(3): 425-431. Stevenson, S.K. 1979. Effects of selective logging on
Fox, J. 1997. Applied Regression Analysis, Linear arboreal lichens used by Selkirk caribou. B.C.
Models, and Related Methods. Part II Linear Min. For. and Min. Environ. Fish and Wildlife Re
Models (Statistics). Sage Publications, London. port No. R2. Victoria.
Goward, T. 1998. Observations on the ecology of the Stevenson, S.K. and K.A. Enns. 1993. Quantifying ar
lichen genus Bryoria in high elevation conifer for boreal lichens for habitat management: a review
ests. Canad. Field Naturalist 112: 496-501. of methods. R.C. Min. For. IWIFR-42. Victoria.
McCarthy, B. 1988. A method of access into the Tucker, G.E and 1.R. Powell. 1991. An improved can
crowns of subcanopy and canopy trees. Canad. J. opy access technique. N. J. Appl. Forest. 8(1): 29-
Forest Res. 18: 646-649. 32.
McCune, B. 1990. Rapid estimation of abundance of Van Daele, L.J. and D.R. Johnson. 1983. Estimation
epiphytes on branches. The Bryologist 93(1): 39- of arboreal lichen biomass available to caribou. J.
43. Wilill. Managem. 47: 888-890.