Torrey Botanical Society
Diversity Patterns in Forest Vegetation of the Wenatchee Mountains, Washington
Author(s): Roger del Moral
Source: Bulletin of the Torrey Botanical Club, Vol. 99, No. 2 (Mar. - Apr., 1972), pp. 57-64
Published by: Torrey Botanical Society
Stable URL: http://www.jstor.org/stable/2484199 .
Accessed: 27/08/2013 19:16
Your use of the JSTOR archive indicates your acceptance of the Terms & Conditions of Use, available at .
http://www.jstor.org/page/info/about/policies/terms.jsp
.
JSTOR is a not-for-profit service that helps scholars, researchers, and students discover, use, and build upon a wide range of
content in a trusted digital archive. We use information technology and tools to increase productivity and facilitate new forms
of scholarship. For more information about JSTOR, please contact [email protected].
.
Torrey Botanical Society is collaborating with JSTOR to digitize, preserve and extend access to Bulletin of the
Torrey Botanical Club.
http://www.jstor.org
This content downloaded from 128.193.8.24 on Tue, 27 Aug 2013 19:16:29 PM
All use subject to JSTOR Terms and Conditions
B U L L E T I N
O F
T H E
T O R R E Y
B O T A N I C A L
VOL. 99, No. 2, pp. 57-64
C L U B
MARCH-APRIL1972
Diversitypatternsin forestvegetationof the
Wenatchee Mountains, Washington'
Roger del Moral
Department of Botany, Universityof Washington,Seattle, Washington 98195
DEL MORAL,R. (Dept. Botany, Univ. Washington, Seattle, 98195). Diversity patterns
in forest vegetation of the Wenatchee Mountains, Washington. Bull. Torrey Bot. Club
99: 57-64. 1972.-Patterns of species richness, dominance and Brillouin diversity are
described for mature forests on serpentine and nonserpentine soils. Where continuous
forest canopies can develop, increased dominance and reduced habitat heterogeneity
result in lower diversity.Where conditions prevent a continuous canopy from developing,
dominance is less pronounced and the habitat is more heterogeneousleading to increased
diversity.Within either a continuous or an open forest, increased environmentalrigor
reduces diversity.A model is presented which details the interactions among rigor, competition, dominance, and habitat heterogeneity.
The subject of species diversity has
spawned innumerablehypotheses,models,
and conjectures concerningthe causes of
and factorsinfluencingthis aspect of communitystructure.There are few data with
which to test or develop hypothesesconcerningplant species diversity (cf. Monk
1967, Auclair and Goff1971). How species
diversityis regulatedis a questionthat derives froma desire to understandcommunitystructureand processmoreclearly. In
this paper I will reportthe results of my
descriptionsof vascular plant species diversityand use these data to develop a
modelregardingits regulationwithinland
plant communities.
Species diversitymay be controlledby
any of several factors acting singly or in
combination.Pianka (1966) discusses six
hypotheseswhichmightaccount for latitudinal diversitygradients,and several papers on the subject of diversityregulation
appear in Woodwelland Smith (1969). Hypothesesbased on habitatheterogeneity,
environmentalrigor,and competitionmay be
mostrelevantto understandingpatternsof
diversitywithina small region.Ecological
or evolutionarytime, predation, and productivityare thoughtto bear less directly
on this problemin terrestrialcommunities.
Time is probablyan importantvariable in
dealing with differe.nees
between regions
but of no great significancein the current
study.A major purpose of this paper is to
investigatethe effecton diversityof gradients of moistureand temperature,habitat
heterogeneity,and competition in plant
communities.
Auclair and Goff (1971) cite twelve
common diversity iindices ranging from
simple richness to slope (equitability vs.
dominance) measuressuch as the ShannonWiener and Simpsonindices.I have chosen
to investigatethe behaviorof richnessand
two equitability measures. The Brillouin
(1956) equitability index (H) derives
frominformationtheoryand is influenced
both by the number of species and their
relative abundance. It was chosen in preference to the Shannon-Wiener function
(H') because, unlike the latter,it is interpreted as a propertyof the sample rather
than as an estimate of a property of a
population from which a random sample
has been obtained (Pielou 1966). The
Simpson index (C) was ehosenas an additional equitabilityor "evenness" statistic.
This index measures the concentrationof
dominanceand is not stronglyaffectedby
richness.
The Wenatchee Mountains form a
south-easttrendingspur of the Washington Cascades (latitude 47? 25' N). Geologically complex,theseruggedmountainsare
clothed with a distinctivepattern of vegetation intensifiedby the mosaic of different
edaphic types. Kruckeberg (1969) described the general featuresof the vegetation, and a detailed gradient analysis of
1 This work was supported in part by National
Science Foundation contract GB-20492 anid in part
by Graduate School Research Funds of the University of Washington. I am grateful to R. G.
Cates, G. Kunze, J. L. Main, and R. R. Richardson
for assistance in various stages of this study. C. L.
Hitchcock and A. R. Kruckeberg identified immature individuals and annotated my identifications in difficultgroups.
Received for publication January 27, 1972.
57
[THE
BULLETIN
for January-February (99: 1-56) was issued May 18, 1972]
This content downloaded from 128.193.8.24 on Tue, 27 Aug 2013 19:16:29 PM
All use subject to JSTOR Terms and Conditions
58
BULLETIN
OF
THE
TORREY
the vegetation upon which this study is
based is in preparation (del Moral, unpubl.).
Methods. SAMPLING AND CLASSIFICATION.
I sampled the vegetationof the Teanaway
River drainage and some surrounding
slopes during the summers of 1969 and
1970 by direct gradient analysis (Whittaker 1967). Parent materials in this region are diverse but fall into two major
types. Peridotite and serpentiniteare ultramaficsoils of low fertilitywhich support structurallydepauperate vegetation.
Vegetation samples from both of these
parent materialshave been pooled and are
considered to come from serpentine (S)
soils. Sandstone,quartz diorite,and greensto.neproduce predominantlyacidic soils
which are more fertile than those from
ultramaficmaterials. Samples from acidic
parent materials will be consideredto be
fromnonserpentine(NS) soils.
I sampled over 75 0.1 ha plots on serpentine soils between 1190 and 1770 m
above sea level and over 125 0.1 ha plots on
nonserpentinesoils between 850 and 1770
m such that each plot was vegetationally
homogeneousand mature,withno evidence
of recent disturbance. For each sample,
density, basal area, and intercept cover
were determinedfor trees and shrubs,and
frequencyand cover for herbs in 25 1-m2
plots. The plots were placed in the landscape pattern to sample representative
combinationsof solar exposure and elevation.
Thirty stands, representingthe range
of vegetationalvariationin each 200-melevation band, were grouped followingprocedures outlined by Whittaker (1960).
Each stand was assigned a nmoisture
index
value on the basis of exposureand topography (cf. Whittaker,1967) and the stands
groupedaccordingto this index. The modal
index value for each species was determined fromwhich a weightedmoistureindex was calculated for herbs and trees
separately. Stands for which the two indices did -notagree were discarded. This
procedure resulted in an ordination of
stands called a weightedmoisturetransect
which was divided into five classes, designated A (mesic) to E (xeric) with five
stands per class.
On NS soil there are fivebands, designated band 1 (low elevation) to band 5
(high elevation), while on S soil there are
threebands, designatedbands 3 to band 5.
BOTANICAL
CLUB
[VOL.
99
A series of seven continuous-recording
thermographsoperating fromMay to October,1971 confirmedthat mean temperatures declined with elevation and that the
time between spring thaw and fall snows
declinedregularlywith elevation.
DIVERSITY
CALCULATIONS. The
clusters
of fivestands yield pooled diversity(Hr)
values. These stands treated individually
and then averaged yield mean diversity
(H) values.
Calculations are based on cover values
obtained from a 50-m interceptfor trees
and arboreseentshrubsand fromthe average cover in 25 1-m quadrats for small
shrubs,forbsand graminoids.I calculated
the Brillouin index,
N!
1
= Ilog,
H H=N
no N,l!
, nl, ! . .. nt,
withN the total leaf cover and n the cover
of each of s species for each of the 200
stands. The table of Lloyd et al. (1968)
was used for calculatingH.
Whittaker (1965) distinguishesamong
several richnessattributes.The numberof
species withina habitat is alpha (m), while
beta (13) measuresthe rate at whichspecies
are replaced along some particular gradient. Whereas x expressesniche differentiation, and : expresses habitat differentiation, they tend to vary together (Whittaker1960).
I calculated /3 as if it were a halfchange value analogous to radioactive deis:
cay rates. The formuLla
logio a - log1iz
log1o2
of commuwhere a is the mean coefficient
nity [CC = Sab/(Sa + Sb - Sab) where Sa
and Sb are the numberof species in samples a and b, and Sab is the number of
species commonto both] between samples
adjacent on a transect,and z is the similarity of samples from the extremes calculated as if the rate at which similarity
declines were uniform for the transect
(Whittaker 1960).
For each group of five samples, I also
calculated H (mean diversity) and H,
(pooled diversity). These values give a
better interpretationof the general patternsof diversityand formthe basis of the
discussion.
Simpson's index, C = E (nilN/)2, where
i,= 1
This content downloaded from 128.193.8.24 on Tue, 27 Aug 2013 19:16:29 PM
All use subject to JSTOR Terms and Conditions
1972]
MORAL:
FOREST VEGETATION
OF THE
WENATCHEE
MOUNTAINS,
WASHINGTON
59
mesic sites and A. lasiocarpa, Ceanothus,
and Arctostaphylosin the xerie sites.
Serpentinevegetationwas divided into
threebands withthe lowerboundarydetermined by the limits of serpentinein the
studyarea and the upper boundaryreflecting the local treeline. Band 3 is dominated
by A. lasiocarpa, Rhododendron,and V.
membranaceumin mesic sites and Pseudotsuga, Agropyron spicatum, and Arctostaphylosin the xeric sites. Dominants in
Band 4 are similarto band 3. Band 5 dominants include A. lasiocarpa, Rhododendron, Dodecatheon jeffreyi,and V. membranaceum in mesic sites and Pinus albicatlis, Agropyron,Anenome drummondii,
and Arctostaphylosin the xeric sites. A
more complete descriptionof this vegetaGROWTH-FORM SPECTRA. Shifts in the tion is in preparation.
growth-formcomposition along various
OF ALPHA DIVERSITY. Values
GRADIENTS
gradients were investigated to ascertain of alpha diversity are summarized for
how one aspect of physical structure pooled vegetationsamples on the two soil
changed with diversity. I distinguished types (Table 1). The mean number of
four growth-forms:trees, shrubs, forbs, species per stand (S), the total numberof
and graminoids.Because theybehave some- species in the group (Se), the mean Brilwhat differently,
the shrubs were subdi- louin ilidex of the group (H), and the
vided into arborescent shrubs and low pooled Brillouin index (Hp) are shownfor
shrubs.
each grouLp.
The following general trends can be
Results and discussion. DESCRIPTION OF
VEGETATION BANDS. The vegetation on non- seen (Table 1). Maximum richness,SS, on
serpentinesoils is divided into five eleva- nonserpentinesoils oceurs on the wettest
tion bands, the lowest occurring at the sites, except at low elevation where very
limitsof undisturbedvegetationat around rich meadow communitiesare absent. In
1200 m and the highesthaving an upper the more mesic types,maximumS, occurs
boundarynear timberline.Band 1 is dom- at mid-elevationbut shiftsto low-elevation
inated by Abies grandis,Achlys triphylla, in the drier types. These trends suggest
and Rubus parviflorusin the mesic sites that decreasing available moisture limits
and Pseudotsuga menziesii,Ceanothutsvel- richness.Reduced richnesswith increased
utinrts,
and Carex geyeriin the xeric sites. elevation suggests that a shortergrowing,
(Nomenclaturefollows that of Hitchcock, seasoniis also a factorregulatingdiversity
et al. 1955, 1959, 1961, 1964, and 1969.) anidis an aspect of climaticrigor. On serBand 2 dominantsinclude Abies grandis, pentine soils S, increaseswith elevation at
Pachistima myrsinites, and Vacciniutm both the wet and at the xeric ends of the
membranacetmin mesic sites and vegeta- spectrum,buLtnot in mesic sites. On both
tionsimilarto band 1 in xeric sites.Balnd 3 wet and dry serpentinesites tree cover dedominants include Abies lasiocarpa, V. ereases with elevationwhich seems to permembranace
enit,and Clintonia uniflorain mitmore total species.
the mesic sites and A. grandis, ArctoIt canibe seen that Hp declineswith instaphylos nevadensis, and Pachistima in creasingaridityon bothsoil types.Oninonthe xeric sites. In band 4, dominaantsin- serpentinesoils Hp peaks in band 2 and
clude Abies amabilis, Rutbsts lasiococcuts, thendeclinieswithelevation.On serpentine,
and Rhododendron albiflorurmin mesic HI is generally highest at high elevation,
sites and Pseudotsuga, Pteridirnt aqutli- althoughmoistureclass B is an exception.
num, and Pachistima in xeric sites. In I inferthat because tree dominanceon serband 5, dominants include A. aniabilis, peiitine in the drier and wettestportions
Rhododendron,and V. membranaceumin of the mnoisture
gradient declines with eleC is the concentrationof dominance,and
othernotationsare as above,was calculated
fromcoverdata. This aspect of equitability
aids in the interpretationof diversitypatterns.
CANOPY DOMINANCE. The use of individual cover estimatesmakes it possible to
estimatethe degreeof shadingby treesand
shrubs.This methodcounts overlaps twice,
though usually only a single species was
responsiblefor cover. I assume that the
percent canopy determineslight interception and thus below groirndbiomass,which
approximatesthe intensityof soil resource
utilization.By determiningthe effectsof
changingdegrees of canopy dominanceon
diversity,the effectsof competitionon diversitymay be suggested.
This content downloaded from 128.193.8.24 on Tue, 27 Aug 2013 19:16:29 PM
All use subject to JSTOR Terms and Conditions
6"
OF THE
TORREY
~~~~~~~~~~~~~~~~~BULLETIN
BOTANICALJ
't
00r-o
C- <=
o
e
c :0 cl a ] c0:
~to c;l It -t
19
CS1
-
cs
~
so
0
t0
o
Co
m
cqc0
o
C:
C;i
C;1
m m cq
In
a;]
00
cS
so
c]
a
m
Co r-
O
cl
c0:
C)
e:
C)
C)1
t:s
'I
In
'I
ei
tC40
C)~~~~~~~~~C
E
l"
~~~~~~~til
to co
X
eo.~~~~~~~~t
S -11cl
06 ei 1 11
to LO to
o
12
o~~~~~~~~~~~~0
m
0 e:
m
i~~~~~~~~~~~~~~~~~l
'tmm
C> S- M
mm
cq
OC%eo
cq
C.
"ItesC)C)
C)
X~~~~~~~~~~~'
"I
C;1 CQ
0 C;
(C
ct~~~~~~~~~~c
lo
ca
4->
oo
d
,?
(:I
C
m
cl
e
o
l
co
e: 4
C:t0
C;i CA cl
"Il
eeOie
cq o
eNsO4
',
C
n~0
O
C)ls
C0 'I1
cle:
ll Cs
t
O
Pt
cs
com
co
4P
b2
C)
3
f
C;i
cl
lo
cl
eo
4
se
c
c
cli cq m
.sq
oensm
n
s
s
Co
C)
e:in L-
4
L(St-
r-o
r4
C>c
C> C>
t
o:eo:
aoo
m t-
e~~~~~~~~~~C
~~~~~~~~~~~~cI
pX
;1
O
m~~~~~~~~~~~~~~~~~~~~~~~~~~C
C)~N
cz
t-
tmo
C.0 m
t o t-
S
ce
c;] asc;i
It
e
C>
m
LnHo
H 00
n C.
C: C;
n
t
C)
acs~C1
0:
o0t-C0'
~ ~~~~~~~~
o lrao~ c;] co
O~~~~~~~~~~~I
2~~~~~~~~~~~'
o
mcj
Ce:
00o
c<]
c;i cs 14 cKi c;] ai
X~~~~~~~~~~0
pH
o
- t-
r--
c<]
:ce
IVOL.
CLUB
m
o
:
=L
-Oir- r
P, r-
?
C)
D
ccHtAmY
C)
C)
L:o
- C
ee
t~~~~~~~~~~~~~~~~~~~~~
L- cL(-.0)me:S?
eee co Lo C) e
=;
eW o
This content downloaded from 128.193.8.24 on Tue, 27 Aug 2013 19:16:29 PM
All use subject to JSTOR Terms and Conditions
99
1972]
MORAL:
FOREST VEGETATION
OF THE
WENATCHEE
vation,the resultantdecline in competition
permitsa betterdeveloped understory.In
the dry serpentinesites, S8 and H also increase. These effectsapparently dominate
any temperatureeffectswhich are evident
on nonserpentinesoils. Serpentinesoils are
more diverse (Hp = 3.815) than nonserpentine soils (Hp = 3.384) at the same elevais not statistictions. While this difference
ally significant(P < .10) it suggestsa true
biological difference. Higher elevation
xeric serpentinesites are morediversethan
xeric nonserpentinesites (P < .05).
The mean Brillouin index (IH) does not
vary as much as Hp and as a consequence,
trendsare less evident.Nonserpentinesamples in the "'B'", "'C'", and "'D'" moisture
classes at the highestelevationare the least
diverse of any in their respective class.
Table 2. Concentration of dominance' for
pooled samples of the vegetation in the Wenatchee
Mountains along a moisture gradient from most
mesic (A) to most xeric (E).
Soil material
elevation (m)
Nonserpentine
850-1040
1140-1190
1190-1370
1370-1550
1550-1770
Serpentine
1190-1370
1370-1550
1550-1770
Moisture
B
C
.22
.06
.06
.05
.08
.20
.08
.09
.08
.20
.08
.05
.06
.04
.12
.13
A
class
D
E
.19
.09
.14
.13
.22
.13
.13
.12
.21
.10
.20
.10
.16
.23
.08
.19
.12
.19
.13
.12
.05
.21
.16
.08
'Computed from Simpson's index C using coverage data.
This suggests that environmentalharshness acts to limit this aspect of diversity.
On serpentinethere is a trend (P < .10)
towards lower diversityin B and C and
greaterdiversityin D and E. This suggests
that where the forest canopy is uniform,
a short growingseason limits diversityof
shrubsand herbs but that a discontinuous
canopy promotesincreased diversityin the
lower strata. Trends in HI between substratessubstantiatethe greaterdiversityon
serpentineshown by the Hp values. Habitat heterogeneityis thus suggested as a
regulatoryfactor.
Simpson's index (C) was calculated for
each group of stands (Table 2). Only general trends are evident. On nonserpentine,
dominancedecreaseswithelevationin most
mesic sites, is high at both low elevation
MOUNTAINS,
61
WASHINGTON
Table 3. Percent tree and shrub canopy determined as the sum of individual species' intercepts.
Mean for nonserpentine plots is 58.3%, for serpentinieplots it is 33.8%.
Elevation band
Nonserpentine
1
2
3
4
5
Serpentine
3
4
5
A
Moisture class
C
B
D
E
67.4 68.0 62.8 53.9
52.5 49.0 60.6 52.1
37.2 49.2 68.1 60.5
49.5 52.4 57.7 75.8
36.6 69.0 82.2 52.4
74.7
60.6
56.0
57.1
53.4
43.1
32.0
32.6
27.7
26.6
12.5
51.2
49.7
49.3
30.3 43.3
28.8 22.2
32.5 31.8
and high elevationbut low in intermediate
sites in classes B and C, and is irregular
in the remainingsites. On serpentine,dominance increasesin class B sites where forest canopies are fairly well developed and
decreases in the driest sites wherethe forests are depauperate. Dominance increases
with increasing soil aridity on serpentine
and to somedegreeon nonserpentine.There
is a decline of dominance on both substrateswith increasingelevation.
Dominance is more pronouncedon nonserpentine soil than on serpentine soil,
leading to lower H. While over-allrichness
is slightly greater on nonserpentinesoil
than on serpentine, the latter soil has
slightly greater richness in comparable
moisture classes per given altitude. The
way in which Simpson's index varies suggeststhat it is largelya functionof canopy
dominance (Table 3). Canopy dominance
Table 4. Rate of replacement (beta diversity)1
of species in composite transects along moisture
and elevation gradients in the Wenatchee Mounitains.
Moisture-gradientbeta,diversity:
Elevation band
Soiltype
Low
1
2
3
Nonserpentine
1.21
1.19
1.04
Serpentine
-
-
1.50
Elevation-gradientbeta diversity:
Mesic
A
4
High
5
1.50
2.13
1.84
Moisture class
B
C
D
1.01
Xeric
E
1.34 2.38 1.81 1.37 1.25
Nonserpentine
0.65 0.45 1.5
1.22 2.00
Serpentine2
1 Computed using density data; units are the
number of times 50% of the species are replaced.
elevational gradient is only 0.5
2Serpentine
that of nonserpentinesoils. If this rate of replacement were extrapolated over a similar range,
values would be approximately double.
This content downloaded from 128.193.8.24 on Tue, 27 Aug 2013 19:16:29 PM
All use subject to JSTOR Terms and Conditions
62
BULLETIN
OF THE
TORREY
is morepronouncedon nonserpentinethan
on serpentine.It tends to decrease with
elevation both on dry sites and very wet
sites, and, on serpentine, often declines
with droughtat any given elevation.
BETA DIVERSITY.Patterns of /8diversity
are reportedin Table 4. On nonserpentine
soil, /3increases with increasingelevation.
This is interpretedas resultingfroma responseto a moreopen forest.On serpentine
0 increasesbetweenband 3 and 4, but declines sharply in band 5. This results because micrositesin band 5 are not as dis-
BOTANICAL
[VOL. 99
CLUB
Within a moistureclass a regular patternof /3diversityoccurs along a gradient
of decreasing temperatureon nonserpentine soil. In the wettestsite : is low. Species here respond largely to high moisture
levels, a condition that does not change
dramatically with elevation. In class B
communities,,3 is high, and then declines
as conditionsbecome more xeric. Where 3
is high, foreststructurechanges considerably withelevation.Under such conditions,
both temperatureand moistureconditions
are changing with elevation at a specific
Table 5. Growth-formspectra of the flora in various coenoclines in the Wenatchee Mountains.
Deciduous
indeX2
Trees
Arborescent
shrubs
Small
shrubs
Forbs
13
11
13
15
10
11
15
14
12
120
100
106
11
10
10
53.5
53.5
47.2
12
13
13
9
9
11
15
11
10
8
7
11
9
11
11
71
82
68
82
67
7
6
6
8
5
56.7
59.0
51.5
50.0
50.0
10
10
8
10
5
4
13
9
11
84
82
88
6
8
8
45.5
29.2
30.0
Moisture
Nonserpentine
Wet (A) (146)
(B) (113)
(C) ( 95)
(D) ( 95)
Dry (E) ( 85)
13
12
11
11
8
14
11
12
11
10
12
12
10
10
8
99
73
56
56
52
8
5
6
7
7
53.8
51.4
51.5
50.0
50.0
Moisture
Serpentine
Wet (A) (116)
(B) (108)
(C) ( 90)
(D) ( 91)
Dry (E) ( 67)
10
10
9
9
8
6
10
3
6
2
11
13
10
8
5
80
69
60
62
46
9
6
8
6
6
18.5
37.8
22.7
34.8
20.0
Parent material
Nonserpentine(174)1
Serpentine (145)
N.S. (3 bands) (152)
Elevationitransect
Nonserpentine
Low (108)
(127)
Middle (107)
(120)
High (100)
Serpentine
Middle (123)
(114)
High (119)
Graminioids
parentheses refer to the total number of species present per group.
Percent of all trees and shrubs that are deciduous species.
1 Numbers in
2
tinctly differentiated as those at lower
elevations, all sites being relatively xeric
except wet meadows. Many xerophytic
species occurring throughout the coenocline
in band 5 are more restricted to band 4.
For example, in band 4 Jtniperts com-
munis, Arctostaphylos nevadensis, and
A4gropyron spicatum are confined to the
xeric portion of the gradient while in band
5 they occur throughout. The pattern of $
diversity appears to be influenced largely
by changes in tree and shrub canopy and
thus ultimately by en-vironmentalrigor.
relative position on the moisture gradient.
On serpentine the elevation gradient is
not extensive and to conmpare serpentine
and nonserpentine coenoelines, values for
the former must be doubled. Values for ,
are highest in the driest and lowest in class
B sites. This reflects mininmal structural
changes with elevation in the mesic sites,
while the driest sites change from tree vegetation that is merely depauperate to a
truly desolate condition.
These values for /3 in nloisture gradients are low compared to those of Whit-
This content downloaded from 128.193.8.24 on Tue, 27 Aug 2013 19:16:29 PM
All use subject to JSTOR Terms and Conditions
1972]
MORAL:
FOREST VEGETATION
OF THE
taker (1960) in the Siskiyou Mountains.
This may reflect the effects of glaciation
in these mountains compared to the more
southernly unglaciated surfaces. Time is
required for recolonization, soil maturation, and habitat differentiation.These low
values suggest that time itself is one factor in explaining diversity differences between regions (cf. Sanders 1969, Whittaker 1969).
RICHTRANSECT
AND
GROWTH-FORMS
NESS.
Growth-form spectra and transect
richness are important aspects of community structure (Table 5). The percentage
of the trees and shrubs that are deciduous
is shown for each group. Deciduous trees
and shrubs tend to be restricted to fewer
elevation bands; although they represent
about 50% of the total tree and shrub flora,
at a given elevation they may constitute a
considerably smaller fraction.
A lnumber of trends is evident. While
174 species were encountered on nonserpentine soils and 145 on serpentine, there
are 152 species on nonserpelntine soils at
the same elevation as the serpentine samples, and richness does not differ greatly
betweeni the two substrates. The distributioniof species among growth-formsis similar, but on nonserpentine soil the tree and
shrub classes contain more deciduous species. No evergreen coniferous species found
on nonserpentine soils avoids serpentine,
although some, such as Abies grandis, are
rare.
On nonserpentine soil the .number of
trees and large shrubs declines with elevation and the number of deciduous species
is reduced from 23 to 14. This suggests
that shorter growing seasons favor herbaceous plants over woody plants and the
evergreen habit in the latter. On serpentine, arborescent shrubs show a pronounced
decline with elevation. Forbs are more numerous on serpentine than off,and overall
richness in bands 3 aind 5 is greater on serpentine. Increasing forb richness with elevation is a predictable consequence of decreasing canopy dominance, if competition
for light is a significant factor in diversitv.
On both soils increasingly xeric habitats result in declining richness in all
groups except graminoids. The proportion
of deciduous species on serpentinie is lowest at the moisture extremes and is very
much lower than on nonserpentine soil.
These trends suggest that environmental
WENATCHEE
MOUNTAINS,
WASHINGTON
63
rigor regulates species richness and that
drought is an importantinfluence.
Conclusions. Diversity changes described here are controlled primarily bv
the degree of environmentalrigor which
in turn controlsthe developmentof trees
and shrubs. Trees and shrubs are capable
of preempting incoming light and they
thereforeexert significantcontrol on the
structureof the remainderof the commuiiity. The demarcation between "harsh"
and "mild" climatesis determinedby the
minlimummildness needed to support a
elosed tree canopy.
In a rigorous environment,productivity is low and biomass reduced. The reduction of tree biomass leads to open comnmutiities and permits forbs and grasses to
contribute more to the overall biomass.
This results in greater evenness. In such
a community,there is considerablehabitat
heterogeineitygenerated by patterns of
light and shade, zones of root occupation,
and levels of competitiveintensity. This
providesmicrositesin which
heterogeneity
less prominentplants may survive. Thus
the richness componentis also enhanced.
At the same time withina harsh environmenit,increased harshness selects for increasingly generalized species (Levins
1968) which accompanies decreasing diversity.
conditionsperIn a mild environment,
mit the developmentof forestswith considerable biomass. In these closed communities, taller species preempt a large proportionof the available light which results
in strong dominanceand is reflectedin a
loweringof the evennesscomponentof diversity.If the canopy is closed, sunflecks
will be minimal an-d the rooting volume
will be betterfilled.Under such conditions,
miero-habitatheterogeneitywill be minimized, and richnesswill be low. Counterinigthis trend is the ability of more specialized species to occur in anl increasingly
mild environment(Levins 1968).
This model predicts that diversity is
maximized at that degree of harshness
whichpreventsa closed canopy but is sufficientlymild to permit considerable specialization. Overall diversitywill be high
in very mild environments(NS, class B,
low elevation). Diversity will decline as
rigor increases until the forestcanopy nio
longerremainsintact (NS, class C and D,
mid-elevations).Diversityjumps as barsh-
This content downloaded from 128.193.8.24 on Tue, 27 Aug 2013 19:16:29 PM
All use subject to JSTOR Terms and Conditions
64
BULLETIN
OF
THE
TORREY
ness increases and stands become open
(NS, class C and D, low elevation; 8, class
B, low elevation). As harshnesscontinues
to increase, diversityonce again declines
(NS, class E, low elevations; 8, class D
and E, low elevation).
This model predicts that fewer species
can be sustained in habitats rendered
homogeneousby their physical or biotic
structural features than in habitats that
are more heterogeneous.Iln the former
there is a greater probabilityof competitive exclusion than in the latter. Recent
successionalstudiesin forestsindicate that
diversitydeclines as a result of the closing
of the canopy (Auclair and Goff 1971)).
This result supports the above prediction.
The model does not considerwhat factors governthe absolute numberof species
in a particular region. Rather, it deals
with relative changes within a region.
Other mechanisms which may influence
the absolute number of species available,
such as differencesin ecological or evolutionary time, predator intensity,or productivity,do not seem to be involvedwith
the local scale pattern of diversity.These
mechanisms may be involved ill setting
absolutelimitsto diversity.
Literature Cited
1971. Diversity
relations of upland forests in the western
Great Lakes Area. Amer. Natur. 105: 499528.
BRILLOUIN, L. 1956. Science and Information Theory. Academic Press, New York.
HITCHCOCK, C. L., A. CRONQUIST, M. OWNBEY, and
J. W. THOMIPSON.1955. Vascular Plants of
the Pacific Northwest. Part V: Compositae
(1955). Part IV: Ericaceae through Campanulaceae (1959). Part III: Saxifragaceae
through Ericaceae (1961). Part II: SalicaAUCLAIR, A. N., and F. G. GoFF.
BOTANICAL
CLUB
[VOL.
99
ceae to Saxifragaceae (1964). Part I: Ferns
Univerthrough Monocotyledons (1969).
sity of Washington Press, Seattle.
A. R. 1969. Plant life cin serpentine
KRUCKEBERG,
and other ferro-magnesianroeks i-i northwestern North America. Syesis 2: 15-114.
LEVINS, R. 1968. Evolution in Changing Environments. Princeton Univ. Press. Princeton,
New Jersey.
LLOYD, M., J. H. ZAR, and J. R. KARR. 1968. On
the calculation of information-theoretical
measures of diversity. Amer. Midl. Nat.
79: 257-272.
MONK, C. D. 1967. Tree species diversity in the
eastern deciduous forest with particular
reference to north central Florida. Amer.
Natur. 101: 173-187.
PIANKA, E. R. 1966. Latitudinal gradienitsin species diversity: A review of coneepts. Amer.
Natur. 100: 33-46.
PIELOU, E. C. 1966. Shannon's formula as a measurement of specific diversity and its use
and misuse. Amer. Natur. 100: 463-465.
SANDERS, H. L. 1969. Benthic marine diversityand
the stability-timehypothesis,p. 71-81. In:
Diversity and Stability in Ecological Systems. G. M. Woodwell and H. H. Smith
Brookhaven Symposium No. 22,
(eds.)
Brookhaven National Laboratory.
WHITTAKER, R. H. 1960. The vegetationi of the
Siskiyou Mountains, Oregon and Washington. Ecol. Monogr. 30: 279-338.
1965. Dominance and diversity in
land plant communities. Science. 147: 250260.
1967. Gradient analysis of vegetation. Biol. Rev. 42: 207-264.
1969. Evolution of diversity in
plant communities,p. 178-196. In: Diversity and Stability in Ecological Systems.
G. M. Woodwell and H. H. Smith (eds.).
Brookhaven Symposium No. 22, Brookhaven
National Laboratory.
WOODWELL, G. M. and H. H. SMITH, eds. 1]969.
Diversity and Stability in Ecological Systems. Brookhaven Symposium No. 22,
Brookhaven National Laboratory.
This content downloaded from 128.193.8.24 on Tue, 27 Aug 2013 19:16:29 PM
All use subject to JSTOR Terms and Conditions