Covariance of lichen and vascular plant floras

21 Pages·1999·7 MB·English

by  J P Bennett

RHODORA,
Vol. 101, No. 907, 277-297, 1999
pp.
COVARIANCE AND
OF LICHEN VASCULAR
PLANT FLORAS
James Bennett
P.
Biological Resources Division, U. Geological Survey,
S. Institute for
Environmental
Studies, University of Wisconsin, 504 Walnut
St.,
WI
Madison. 53705
M. Wetmore
Clifford
Department
of Plant Biology, University of Minnesota, 1445
MN
55108
Paul,
St.
abstract. The geographic relationships among taxonomic groups im-
are
portant to study to determine patterns of biodiversity and whether or not
associations occur between large groups, birds and vascular This
e.g., plants.
study was undertaken to determine relationships between higher plants and
lower plants, specifically vascular plant and lichen floras in nine national
No
parks of the Great Lakes region. significant relationship was found be-
tween vascular plant floras and lichen floras in this area, which spans 1200
km
longitudinally, or between an additional 19 areas from North America
km km
that were less than 1000 2 in area. For areas larger than 1000 2 however,
,
a significant positive relationship existed for 33 areas that span one to ap-
km
proximately 150 million 2 The ratio of numbers of vascular plants to
.
lichens appeared to average just over 6 across the 33 areas. In the Great
28-30%
Lakes parks, between of either the vascular plant or lichen species
were singletons (occurring in only one park), but the parks that contained the
most singletons were not congruent: Isle Royale had the most singleton
li-
chens, while Indiana Dunes had the most vascular plant singletons. Fewer
lichen species (2%) than vascular plants (4%) occurred in all nine parks.
some between two
Latitude appeared to explain of the variation the groups:
vascular plants decreased with increasing latitude, while lichens increased.
Key Words: lichens, vascular plants, species-area curves
floras,
Mex-
The number North America, north of
of lichen species in
3600 and Egan
ico, thought to be approximately (Esslinger
is
number approaches
1995), while the of vascular plant species
22,000 (Kartesz 1994), a ratio of about six vascular plant species
have been
one Although general, ratios like this
to lichen species.
Hawksworth
used to make broad biodiversity estimates (e.g.,
made more
Such can be useful the geographic
1991). estimates if
pattern of covariance of higher taxa can be ascertained. In recent
covariance of higher taxa have been con-
years studies of spatial
ducted order determine patterns of biodiversity hotspots, or
in to
277
278
Rhodora
[Vol. 101
among
the relationships taxa, if any exist (Faith and Walker 1996;
Many
Gaston
1996). factors can
affect spatial covariance, includ-
ing scale, environmental and
factors, biological properties of the
taxa. Several authors have pointed out and
that lichen vascular
may
plant
diversities not track each other in general due
to
cli-
mate (Huston
1994) and habitat diversity (Galloway
1992; Gilbert
Another
1977). study found
that lichen biodiversity decreased
significantly with an increase
in vascular plant cover (Pharo and
we
Beattie
1997). In addition, suspected
that the effects of
air
may
pollutants also be a
factor affecting lichen and vascular plant
greater
air
This
and
Lakes
region of the United
States.
The
lichen
floras of nine
national parks
in the Great Lakes
region of
the north central United have
States been
well
studied
in recent years and
are considered be
to fairly complete. The
parks include Apostle
Islands National Lakeshore
Wisconsin
in
(APIS), Cuyahoga
Valley
National
Recreation Area Ohio
in
(CUVA),
Grand
Portage Monument
National
Minnesota
in
(GRPO),
Indiana Dunes
National Lakeshore
in Indiana (INDU),
Royale
Isle National Park (ISRO)
and
Pictured Rocks
National
Lakeshore
Michigan
in (PIRO),
Croix
St. National Scenic River
n^
in Minnesota and Wisconsin iSACNV SWnina rw-o m„
r
Lakeshore
Michigan
in
(VOYA:
Minnesota
i Fi
affinities
associated
geographically
within homogeneous
the
Great Lakes/
The
km
1200
km
The
2
vascular
plant
.
segregated two
into groups by
multivariate
This was
study
also undertaken
determine
to same
if the groups
""
< ° **"~"
*
°
lichen
flic floras.
The
management
of
natural area
preserves
sometimes
is
fo-
cused on
a select group of
organisms,
mammals,
e.g.,
trees, rare
plants,
birds, or
butterflies.
often
It is implicit management
in the
of
these groups
that the management
of one
group
will also
sat-
This
groups
—
Wetmore Co
Bennett and
1999] variance of Floras 279
VOYA
CUVA
NORTH
Map
Figure of nine Great Lakes national park units for which floras
1.
were used in this study. Park codes are explained in the introduction.
purposes
and
and
ales thei
management
may assumed more
spective however, be that if there are vas-
it
more
be This study
cular plants in an area there will also lichens.
was undertaken in order to determine the validity of this as-
no
sumption by attempting to prove the hypothesis that there is
and
between vascular plant lichen floras.
relationship
AND METHODS
MATERIALS
parks
summer
during the
(Wet
parks every year
particular
280 Rhodora
[Vol. 101
Numbers
Table of and
collections, collecting days, collection
1 localities
.
Lakes
names.
Number
of
Collecting Collection
Park Days
Collections
Localities
APIS
1497
21 28
CUVA
304
11 31
GRPO
373
3 6
INDU
371
14 24
ISRO
5246
53 72
PIRO
1231
11 25
SACN
2327
88 77
SLBE
847
10 27
VOYA
8028
80
128
more
1983b,
1985, 1988a, 1988b,
1989, 1990a, 1990b, 1991,
Wetmore
1992b; and
Bennett 1996
[data are available on
the
Internet at www.ies.wisc.edu/brd]).
Collection each
localities in
park were
selected to include and
all habitats vegetation
types.
The
collection were
localities distributed over the entire park, and
vegetation types were
studied multiple At
times. each
locality,
all
groups of lichens were
collected
(fruticose, foliose, squamulose,
and crustose) and was
sufficient time spent examining
sub-
all
At
each
strates. locality species found were
all
collected to pro-
vide abundance
relative
estimates, even though same
the
species
might have been found A
previous summary
at
localities. of
the
collecting shown
efforts Table
is in Vascular
plant and
1 floras
.
methods were
described
previously
(Bennett
1996a).
The
nomenclature
of each
lichen was
flora updated
to the
fifth
checklist (Egan
1987) so
as to standardize the names. The
all
names
were
then entered into a computer
spreadsheet program
(MICROSOFT
EXCEL,
WA)
Microsoft
Corp.,
Seattle, and
the
presence/absence
of each
species recorded
a
in each
field for
park.
Various and
were
tallies sorts performed
in order perform
to qual-
ity control procedures and
rank
to species by
frequencies. This
MINI
(Minitab
between
variables
linear
correlation
coefficients.
Similarities between nark
fl.
—
Bennett and Wetmore
1999] Covariance of Floras 281
Table 2. Area, numbers of lichen and vascular plant species, average Jac-
card similarity based on lichens and vascular and average
plants, distance
from
the other parks for nine Great Lakes national parks. See introduction
park names.
for
Number
of Average
Similarity
Latitude
Species
(decimal Vascular Average
Area
degrees Vascular Lichen Plant Distance
Park (km2 N) Lichen Plant (%) (%) (km)
)
APIS 6605 47.000 285 509 39 35 388
CUVA
3136 41.283 66 855 26 842
18
GRPO
287 47.967 182 279 30 23 383
INDU
5203 41.625 62 1399 19 28 638
ISRO
54,140 48.000 554 698 31 37 400
PIRO
25,545 46.567 245 723 36 39 415
SACN
26,459 46.000 282 1165 34 33 443
SLBE
23,663 44.875 182 928 33 37 426
VOYA
55,955 48.500 458 603 35 35 566
similarity index, which measures the proportion of park pairs
where species are present in both parks in the Details are
pair.
The two ma-
given Bennett (1996b). correlation of similarity
in
PC-ORD
(McCune
was with Mantel's with and
trices tested test
Mefford
STREET
were measured
and using
tances latitudes
Lorme, Freeport, ME), a desktop computer mapi
and were
program. Straight line distances latitudes calculated us-
ing the approximate park centroids for park locations. Distances
were rounded the nearest five kilometers.
to
RESULTS
A
shown
summary of each park is in Table 2. Isle
floristic
Royale had number of lichen species, while Indiana
the greatest
Dunes had the lowest. The greatest and lowest numbers of vas-
however, were Indiana Dunes and Grand
cular plant species, at
Based on the lichen floras, seven of the
Portage, respectively.
40%
on between 30 and similar to the
parks were, average, all
Dunes and Cuyahoga
Indiana Valley,
other parks, except for
20%
which were almost similar to all other parks. This is not too
away from
surprising, given that on average they are farther all
on
Overall similarity based vascular plant floras
the other parks.
282
Rhodora
[Vol. 101
pattern, except for Grand Portage
lower Cuyahoga
similarity, than
Indiana Dunes.
Some
of were
the variables
significantly correlated (Table
3):
lichen numbers and park area (Figure and average
lichen sim-
2),
and
average
ilarity distance (Figure or were
3) latitude the stron-
Number
gest relationships. of
lichens increased with
latitude,
while vascular numbers
plant decreased Numbers
(Figure
of
4).
and
lichens vascular plants appeared be
to negatively
correlated,
but the probability of
the correlation occurring by chance was
high Park
(0.33). similarities based on and
lichens vascular
plants
were comparable =
0.05
Vascular
(/> 7). plant similarity signif-
1
icantly increased with area and did not appear
related to latitude,
similarity increased with and
latitude
related to area.
matrix
ilarity
of
pairs the nine parks shown Two
is in Table 4. park
pairs, Isle
Royale and
Voyageurs, and
Apostle
Islands and Rocks
Pictured
were 50%
greater than
similar, while
the overall average was
31%. The
comparable
distance matrix (Table shows Cuy-
5)
that
ahoga
Valley and Voyageurs km
are just over 1200
while
apart,
Grand
Portage and
Royale
km
Isle are the
closest pair only 60
at
The
apart. similarities for these same were 11% 30%
pairs and
When
respectively (Table
36
4). all pairs of and
similarities
dis-
were
tances
plotted against each
other (Figure was
there a
5),
clear negative
relationship, even though maxima
the and minima
did not
correspond.
all
The
overall average of
vascular 32%
plant was
similarities
(see
Table
Bennett
in 1996b), and only one
park
1 Royale
pair. Isle
and
Pictured Rocks, was more 50%
than
The
similar.
similarities
and
The
plant similarity matrix was
significantly pc
lichen = =
similarity matrix
(Mantel 6.3718,p"=
r 0.9151,
t
0.0000), suggesting
that these parks show
comparable
degrees
and
patterns of
similarity based on
both
vascular and
plant
lichen
floras.
A
cluster analysis of
the lichen
flora Jaccard
similarity matrix
did
not reveal any
significant groupings
because
the
similarities
an
matrix of
species
all in the nine parks
»
SO
NO
sO
03
3
9
Table 3. Pearson's linear correlation coefficients for variables given in Table 2. Significance levels of coefficients are indicated H
3
***
** Number
by preceding asterisks: * (0.05), (0.01), and (0.001 of species. CL
). '
Average
Vascular Average Lichen Vascular
Plant
Area Latitude Lichens Plants Similarity Similarity
i
0.5702
Latitude
***0.9124 •0.7968
Lichens
1
-0.1671 0.6985 0.3715
Vascular plants
1
**
Average 0.5214 0.8379 0.6210 0.4850
lichen similarity
Average
vascular plant
0.7054
0.3882 0.5783 0.0440 0.6628
similarity
**
Average distance -0.2143 0.7685 0.4796 0.4002 0.8083 0.4541
3
c
7
00
Rhodora
284 101
[Vol.
600
500 Lichens = 0.7817 area + 54.262
R2
= 0.8325
o
400
<D
Q-
300
JQ 200
E
3
100
100 200 300 400 500 600
Area (km
)
Figure 2. Relationship between number of lichen species and area for
nine Great Lakes national parks, and fitted linear regression line.
some
groups At
interesting (Figure the highest
6). similarity level,
80%,
about four parks (Apostle Islands and Pictured Rocks, and
Cuyahoga Valley and Indiana Dunes) were grouped two
into pairs
36%
were themselves only about Cuyahoga
that similar. Valley
Dunes
and Indiana appeared be two were
to parks that unrelated
50%
which
floristically to the other seven, appeared be about
to
was
similar overall. Isle Royale, however, less similar to the other
six northern parks, in spite of the high individual Jaccard simi-
Two
larity with Voyageurs. groups of three parks each had sim-
0.45
*
0.4
'«
0.35
03
I
0.25
§
JC
u 0.2
0.15
Similarity = -0.0004 distance + 0.4997
0.1
2
R
3 = 6533
0.05
300 400
500 600
700
800 900
Average
distance (km)
Figu
similarity
age distance from
the other parks for the nine Great Lakes
national parks,
and
fitted linear regression
line.
—
Wetmore
Bennett and Covariance of Floras 285
1999]
1600
= +
Vascular plants -88.418 degrees 4841
1400
R2
=
0.4878
w 1200
1000
A
800
.Q 600 Lichens = 48.801 degrees - 1975
E
R2
= 635
400
200
40 42 43 44 45 46 47 48 49
41
Latitude (degrees North)
A
Lichens Vascular plants
Figure Relationships between number of vascular plant and lichen spe
4.
cies and latitude for nine Great Lakes national parks, and fitted linear re
gression
lines.
ilarities
shore of Lake Superior and one at the northern end of Lake Mi-
Pictured Rocks, and Sleeping Bear
chigan (Apostle Islands,
of Minnesota parks (Grand
Dunes), and the other consisted three
Voyageurs, and Croix).
Portage, St.
of 698 lichen species in 162 genera were found in the
Totals
The most
when were aggregated. frequent
nine parks the floras
were found nine parks (14 species,
in
those that all
species,
i.e.,
2% included Arthonia caesia,
or of lichen species found),
all
Cladonia
Cladina rangiferina, chlo-
Candelariella efftorescens,
of nine national
(%) matrix for lichen floras
Table Jaccard similarity
4.
parks
SACN SLBE
CUVA GRPO INDU ISRO PIRO
APIS
CUVA
14.3
GRPO
43.7 13.2
INDU
43.8 14.0
16.4
ISRO 30.0 9.0
44.7 9.0
PIRO 40.9 15.0 38.2
57.3 13.1
SACN 36.4 37.2
20.0 33.7 19.0
43.5
45.7 39.8
SLBE 42.8 20.4 31.9 25.1 25.8
VOYA 41.7 45.7 30.6
59.1
35.6 11.1
49.8 11.0
286
Rhodora
[Vol. 101
Table Distance matrix
5.
fi kilo-
meters between park
centroids.
CUVA GRPO
APIS INDU ISRO PIRO SACN SLBE
CUVA
950
GRPO
145 975
INDU
670 470 745
ISRO
185 930 60 725
PIRO
340 690 300 560 250
SACN
125 980 275 620 320 435
SLBE
425 530 300 390 410
185 480
VOYA
260 1210 260 920 320 560 310 685
rophaea,
C. coniocraea,
C. cristatella, Flavoparmelia
caperata,
LT eprari* a •"* t • Mm e»lai nelmim *a —
finkii, subaurifera, Parmelia
and
rudecta.
Scoliciosporum
chlorococcum.
Thirty percent of
lichens (210
spe-
cies) occurred in only one Most
park. of
these (129) occurred
in
Royale. For
Isle the vascular an
plants, aggregated of 2102
flora
species in 691 genera was found
over
all the parks, with 81 (4%)
occurring
in all nine parks. Twentv-eieht Dercent of thp v^n.hr
0.70
0.60
Vascular =
•• plants -0.0002 distance + 0.4251
R2
=
0.2612
0.50
0.40
•c
IV
E
0.30
I
v>
0.20
0.10
Lichens = -0.0004
distance
2
R
=
0.00 0.5898
500
1000
1500
Average
distance
(km)
Figure
similarities
pairs
Lakes