Evidence of climate trend effects on vegetation
patterns in the US Pacific Northwest—
Response of epiphytic lichens from 1993-2009
Linda Geiser1, Sarah Jovan2, Doug Glavich1
1US
Forest Service, Pacific Northwest Air Resource Management
Program, PO Box 1148, 17 Corvallis, OR 97339
2US
Forest Service, Pacific Northwest Research Station, Forest Inventory
and Analysis program, (FIA) 20 SW Main, Suite 400, Portland, OR 15
97205
Introduction
• The FIA Air Quality and Climate
Indicator is designed to assess
status and trends in air pollution
and climate affecting the nation’s
forests and to provide evidence of
ecological effects from these trends.
• Uses lichen community composition
• Which shifts as different species
reach their tolerance limits.
• Shows response before effects on
less sensitive flora can be detected.
Lichens are symbiotic organisms
consisting of a fungus and a
photosynthetic green algal +/or
blue-green bacterial partner.
Unique morphology and physiology
underlies special sensitivity of lichens
• Epiphytes are little influenced by soil nutrition –no
roots, aerial location. Rely on atmospheric
deposition for nutrients and moisture.
• No barrier cuticle or guard cells…moisture,
nutrients, and pollutants are absorbed passively
over entire surface of the lichen.
• Dehydration concentrates pollutants
• Precipitation, if clean, leaches pollutants
• Dynamic equilibrium , fast response.
• Require wetting and drying for nutrient exchange
between symbionts.
• Species are differentially adapted to hot, dry vs.
cool, wet climate regimes and to atmospheric
deposition of nutrients and acidity.
Every lichen is adapted to specific
atmospheric deposition levels & chemistry
Lichens of low fertility,
acidic environments.
©Erin Cooper
Lichens of high nutrient, more
alkaline environments
… so lichen species composition can be used to
pinpoint a site along regional air pollution gradients.
In Oregon’s Willamette Valley, oligotrophic species thrive in clean air.
And eutrophic species thrive where
nutrient availability is greater
Every lichen is also adapted to a
specific range of climatic conditions…
…
© Sharnoff
As climate becomes warmer the
composition of coastal epiphytes
changes dramatically
…therefore lichen community
composition can be used to pin-point a
site along regional climate gradients.
Data collection
•
•
•
•
0.4 ha plot on the systematic national P3 grid (1 plot/96,000ac)
Surveys up to 2 hours
Collect sample of each species detected
And rate abundance:
1 = Infrequent (< 3 thalli)
2 = Uncommon (4-10 thalli)
3 = Common (>10 thalli; covers < 50% of
all boles and branches)
4 = Abundant (>10 thalli covers > 50% of
all boles and branches)
• ID by regional taxonomic expert
Lichens sampled in shaded area
The FIA AQ & Climate Network
Jovan & McCune. 2005. Eco Apps 15:
1712-1726.
Jovan & McCune. 2006. Air, Water,
Soil Poll. 170:69-93.
McCune et al. 1997. Bryologist 100:
145-158.
Geiser & Neitlich. 2007. Env. Poll.
145:203-218.
Western PNW model
Round 1: 1993-2001
1416 plots were surveyed
by:
• FIA (23 km grid)
• USFS PNW regional Air
Program (5.4 km grid)
Round 2: 2003-2009
350 Air program plots were
re-surveyed
• Providing data for this
trends analysis
Geiser, L.H.; Neitlich, P.N. 2007. Env Poll145: 203-218.
Non-metric multi-dimensional scaling (NMS) ordination
Coast Distance
2: Climate
AxisAxis
2
Elev
Continentality
Longitude
Lichen %N
%RH
Mean T
Min Dec T
Axis 1:Axis
Air 1
Quality
Poll?
0
1
Western
PNW
model
Two major
influences :
•air quality
•climate
scored as distance
along Axes 1 & 2
Geiser & Neitlich. 2007.
Env. Poll. 145: 203-218.
Air Quality and Climate scores
Round 1: 1993-2001
Geiser, L.H.; Neitlich, P.N. 2007. Air pollution and climate gradients in western Oregon and Washington
indicated by epiphytic macrolichens. Environmental Pollution 145: 203-218.
%
30
Air scores and all available measures of
atmospheric N deposition are correlated
20
10
0
-1
Total Inorg N kg/ ha
Linear Fit
0.75
0.5
1
0.25
% Eutrophs
0
0
-0.25
-0.5
-1
CMAQ total N
1
2
1
Bivariate Fit of Mean Air S core By Mean NH4 mBivariate
g/L 95-98
Fit of % Eutrophs By V21s_Air
Mean Air Score
AirScore Leverage Residuals
Le verage Pl ot
0
Air Score
3
4
5
6
7
NADP Wet N
-0.75
0
8
.025
.05
.075
100
90
80
70
60
50
40
30
20
10
0
% Eutrophs
-1
.1
Mean NH4 mg/L 95-98
0
Air Score
1
(ugFitm -3 y-1)
Bivaria teTotal
Fit of
Air N
Score
NP laglaP<.0001Bivariate Fit of Ai r Score By Am bient fine parti culate NLinear
Inorg
kg/haBy
Leverage,
Bivariate Fit of % oligotrophs By V21s_Air
0.75 Linear Fit
0
-1
Lichen % N
0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1 1.11.21.31.41.5
NPlagla
0.5
Mean Air Score = -0.481793 + 12.61532 Mean NH4 mg/L 95- 98
90
Summa ry of Fit
0.25
70
RSquare
0RSquare A dj
Root Mean Square Error
-0.25
Mean of Response
Obs ervations (or Sum Wgts)
-0.5
0.738111
0.694463
0.242337
0.072493
8
% Oligotrophs
Air Score
Air Score
100
Linear Fit
1
60
50
40
30
20
IMPROVE N
10
Analysi s of V ariance
-0.75Source
DF Sum of Squares Mean Square
0
.2
.4
.6
.8
1
Model
1
0.9931078
0.993108
Ambient
f
ine
par
ticulate
N
(ug
m-3
y
-1)
Error
6
0.3523639
0.058727
C. Total
7
1.3454716
Linear Fit
% Oligotrophs
80
F Ratio
0
-1
16.9105
Prob > F
0
Air Score
Linear Fit
1
0.0063deposition in western
Geiser
Linear Fitet al. 2010. Lichen-based critical loads for atmospheric nitrogen
Pa rame ter Estima tes
Linear
Fit and Washington forests,
100
Linear
FitEnv. Poll. 158: 2412-2421.
Oregon
USA.
Summa ry of Fit
RSquare
0.533013
90
Ter m
Estimate Std Error t Ratio Prob>|t|
Air Score = -0.758189 + 1.9060603 Ambient fine partic ulate N (ug m-3 y-1)
80
Intercept
-0.481793 0.159716
-3.02 0.0235
Summa ry of Fit
70
Mean NH4 mg/L 95-98
12.61532 3.067751
4.11 0.0063
60
RSquare
0.933582
ophs
Air Score = -0.828166 + 1.3865437 NPlagla
Bivariate Fit of % Eutrophs By V21s_Air
Climate Change Anticipated effects
PREDICTED CHANGE:
• Mean annual temperature to ↑1.5 to 3.2° C by
2040.
POTENTIAL RESPONSES:
• A 1o C temp Δ ↓probability of finding some
lichens 2-10 fold.
• Most biodiversity contributed by rare species
→Concern for local/regional extirpations
→ Especially coastal (maritime) and subalpine
(high elevation) species.
10/14/2010 02:24 PM
o
Data Table=Climate Category R1=maritime,Climate
Category R1=marit
Cli mate
Cate
gory
R1=
lowla
nd
-10-9
-8
-7
-6
-5
-4
-3
-2
-1
0 1 2 3 4
-10-9 -8 -7 -6
Distributions
10/14/2010 02:24 PM
anMin DeCategory
cTem pR1=low
1988-1998
C
Me anMin
Dela
DataMe
Table=Climate
land,Climate
Category
R1=low
Mean min Dec Temp C at round 1
lichen plots (1988-1998)
Cli mate Cate gory R1= montane
Distributions
10/14/2010 02:24 PM
Maritime
Me anMin De
Data Table=Climate Category R1=montane,Climate Category R1=mont
10/14/2010 02:24 PM
Distributions
-10-9 -8 -7 -6Category
-5 -4 -3 -2
-1 0 elevation,Climate
1 2 3 4
-10-9
-8 -7 R1=
-6
Data Table=Climate
R1=high
Category
Cli mate
Cate gory
R1=p
high
elevation
Me anMin
De cTem
1988-1998
C
Me anMin De cTem p 1988-1998 C
elev ation
Lowland
Me anMin De
Distributions
Cli mate
-10-9Cate
-8 -7gory
-6 -5 R1=
-4 -3montane
-2 -1 0 1 2 3 4
-10-9 -8 -7 -6
Montane
Me anMin
Me anMin De
10/14/2010
02:24De
PMcTem p 1988-1998 C
Data Table=Climate Category R1=montane,Climate Category R1=mont
Cli mate
Cate
-10-9
-8 -7gory
-6 -5 R1=
-4 -3mari
-2 -1time
0 1 2 3 4
-10-9 -8 -7 -6
Distributions
Hi Elevation
10/14/2010 02:24 PM
anMin DeCategory
cTem pR1=maritime,Climate
1988-1998 C
Me anMin
De
DataMe
Table=Climate
Category
R1=marit
-10-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4
-10-9 -8 -7 -6
Distributions
Me anMin De cTem p 1988-1998 C
Me anMin De
M
Category
R1=maritime,Climate
-5 -4 -3 -2
-1 0 1 2 3 4 Category
-10-9
-8 -7
-6 -4
-5 -3
-4 -2
-3 -1
-2 -1
-10-9
-8 R1=maritime
-7
-6 -5
0 0
1 21 32 43 4
Change in mean annual min Dec T (C) at
lichen plots, 1988-98 vs. 1999-2008
Cli mate Cate gory R1= mari time
-10-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4
10/14/2010 02:24 PM
cTem
1988-1998
C
Me
anMin
DeR1=
c Te -3
mp-21999-2008
ry
R1= p
mari
time
Cli mate
Cate
nd0 1 2 3C 4 Category
-10-9
-8gory
-7 -6Category
-5 -4lowla
-1
-10-9
-8 -6
-7 -5
-6 -4
-5 -3
-4 -2
-3 -1
-2 0
-1 10 21 32 43 4 -1
Data Table=Climate
R1=maritime,Climate
R1=maritime
-10-9
-8 -7
M
10/14/2010
02:24 PM
Distributions
Category R1=maritime,Climate
R1=maritime
Data Category
Table=Climate
Category R1=low land,Climate Category R1=low land
Me anMin
De cTem
1988-1998
C
MeCate
anMin
DeR1=
c Telowla
mp 1999-2008
C
Cli mate
Cate gory
R1= p
mari
time
Cli mate
gory
nd
Distributions
10/14/2010 02:24 PM
10/14/2010 02:24 PM
oC
oC
Maritime
+
0.7
-5 -4 -3p-21988-1998
-1 0 1 2 C
3 4 MeMe
-10-9
-8De
-7 De
-6 c
-5+Te
-40.1
-3
-21999-2008
-1 0 1 C
2 3C 4 MeLowland
cTem
anMin
mp
anMin
cTem
p
1988-1998
anMin
De
c
Te
mp
1999-2008 C
Data Table=Climate Category R1=maritime,Climate
R1=maritime
Data Category
Table=Climate
Category R1=low land,Climate Catego
Distributions
Cli mate
Cate gory R1= high elevation Distributions
ry R1= montane
I999-2008
10/14/2010
02:24
-10-9
-8 PM
-7 De
-6 cTem
-5 means
-4 -3p-21988-1998
-1 0 1 2 C
3 4 MeMe
-10-9
-8De
-7 De
-6 c
-5 Te
-4pmp
-3
-21999-2008
-1 0 1 C
2 3C 4 Me
Me
anMin
anMin
anMin
cTem
1988-1998
M
Data Table=Climate Category R1=high elevation,Climate Category R1=high
Cli mate Cate gory R1= high elevation
Category
R1=montane,Climate
Category
R1=montane
-5 -4 -3 -2
-1 0 1 2 3elev
4 ation
-10-9
-8
-7
-6 -4
-5 -3
-4 -2
-3 -1
-2 -1
-10-9
-8 -7
-6 -5
0 0
1 21 32 43 4 -10-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4
Cli mateI988-1997
Cate gory means
R1= montane
10/14/2010 02:24 PM
Distributions
Data Table=Climate Category R1=high elevation,Climate C
10/14/2010 02:24 PM
cTem
1988-1998 C
Me
anMin
De
c
Te
mp
1999-2008
C 4 Me
Me
anMin
De
cTem
p
1988-1998
C
anMin
De
c-6Te
mp
1999-2008
C
ry
R1= p
montane
Cli mate
Cate-8gory
R1=
time
ation
-10-9
-7 -6Category
-5 -4mari
-3 -2
-1 0 1 2 3elev
-10-9
-8
-7 -5
-5 -3
-4 -2
-3 -1
-2 0
-1 10 21 32 43 4 -1
Data Table=Climate
R1=montane,Climate
Category
R1=montane
-10-9
-8 -7
-6
-4
M
10/14/2010
02:24 PM
Distributions
Distributions
Category R1=montane,Climate
Category
R1=montane
Data Table=Climate Category R1=maritime,Climate Category R1=maritime
MeMe
anMin
De cTem
pmari
1988-1998
C C Me
Me anMin
De cTem
1988-1998 C
anMin
DeR1=
c Te
mp
1999-2008
Cli mate
Cate gory
R1= p
montane
Cli mate
Cate
gory
time
Distributions
10/14/2010 02:24 PM
10/14/2010 02:24 PM
-5 -4 -3p-21988-1998
-1 0 1 2 C
3 4 Me
-10-9
-8De
-7 De
-6-4
-5-3
-4p-2
-3
-21999-2008
2 43C 4 Me
-10-9
-8 -7
-6
-5
-1
0-1 10 21 3C
-10-9
-8 -7 De
-6 -5
-4 -3
-2 1999-2008
-1 0 1 2 3C4
cTem
Me
anMin
c
Te
mp
cTem
1988-1998
anMin
c Te
mp
Data anMin
Table=Climate
Category
R1=montane,Climate
Category
R1=montane
Data Table=Climate
Category R1=maritime,Climate Categ
Distributions
Distributions
Cli mate
Cate
gory
nd
-10-9
-8 -7
-6
-5
-1
0-1 10 21 3
-10
-10-9
-8
-7 De
-6R1=
-5 -4lowla
-3p-21988-1998
-1 0 1 2 C
3 4 Me
-10-9
-8De
-7 De
-6-4
-5-3
-4p-2
-3
-21999-2008
Me
anMin
cTem
Me
anMin
c
Te
mp
anMin
cTem
1988-1998
C2 43C 4 Me
10/14/2010 02:24 PM
o
o
land
-5 -4 -3 -2 -1 0 1 2 3Data
4 Table=Climate
-10-9
-8 -7
-6 -4
-5 -3
-4 R1=low
-3 -1
-2 -1
0
4 -10-9R1=low
-10-9
-8 -7
-6Category
-5
-2
0 land,Climate
1 21 32 43 Category
-8 -7 -6
-5 -4 -3 -2 -1 0 1 2 3 4
Montane + 1.2 C
Distributions
High elevation +2.0 C
Cli mate Cate gory R1= lowla nd
10/14/2010 02:24 PM
Me anMin
cTem
p -3
1988-1998
C
Me anMin
De
c-6Te
mp
1999-2008
C
Cli mate
Cate-8De
gory
-10-9
-7 -6 R1=
-5 -4montane
-2 -1 0 1 2 3Data
4 Table=Climate
-10-9
-8 -6
-7Category
-5 -3
-4 R1=low
-3 -1
-2 0
-1land,Climate
4 -1
-10-9
-8 -7
-5 -4
-2
10 21 32 43 Catego
=high elevation,Climate
Category
-10-9
-8 -7 R1=high
-6 -5 -4 -3 -2 -1 0 1 2 3 4
Change in mean annual min Dec T (C) at -10
lichen plots, by climate zone, 1988-2008
Cli mate Cate gory R1= montane
988-1998 C
Me anMin De c Te mp 1999-2008 C
means for R2 High Elevation plots
10/14/2010 I998-2008
02:24 PM
Data Table=Climate Category R1=montane,Climate Categ
1 0 1 2 3 4
Distributions
Me anMin De cTem p 1988-1998 C
-10-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4
Me a
la nd
-10-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1 2 3 4
=low land,Climate CategoryI988-1997
R1=lowmeans
landfor R1 Montane Plots
-10
What actually
happened: PRISM
temperatures
1988-2008
• Warming rates ↑ with
elevation
• Low elevations, valleys
and the coast range did
not warm
• No sites became cooler
What actually
happened: Lichens
1993-97 vs. 2003-07
Lichen community based
climate scores also showed
greatest increases at
highest elevations.
Like temperature, lichen-indicated warming
increased with elevation
Climate change direction
indicated by lichen communities
(1993-2009)
N
Mean
elevation
(m)
much warmer
5
813
warmer
34
705
no change detected
248
583
cooler
37
539
much cooler
12
470
What actually
happened:
Lichens
But lichen-community based
climate scores show
cooling—where PRISM mean
min Dec temps have not
changed.
What’s happening?
But what actually happened?
Poll?
Coast Distance
Axis 2: Climate
Elev
Continentality
Longitude
Lichen %N
?
%RH
Mean T
Min Dec T
Axis 1: Air Quality
0
1
Could shifts in other
climate-axis variables
be influencing lichen
response?
What actually
happened: % RH
1988-2009
• Humidity ↓ in the high
elevation Cascades
• Humidity ↑ in the lower
elevation Cascades &
Coast Ranges
Comparing climate & lichen responses
Mean Min Dec Temp
% Relative humidity
• warmer temps
• less humidity
Lichen-based climate scores
• warmer and less humid
R2 climate scores were correlated with
climate variables
10/5/2010 09:45 AM
Data Table=MegaDB2
Fit Y by X Group
Bivaria te Fit of Clim S core R2 By Ele v ft
2
1.5
1.5
0.5
0
-0.5
1.
1
• Data from Round
2 (2003-2009).
• Low climate0.5scores → low elevations
0
with warm, humid
climates,
• High climate-0.5scores → high elevations
-1 climates.
with cool, dry,
r2 adj = 0.72
-1
-9 -8 -7 -6 -5 -4 -3 -2 -1 0 1
2 3 4 5
0
MeanMin Dec Temp 1999-2008 C
ClimBivaria
S core R2
By of
MeClim
anMin
De cR2
TeBy
mpMe
1999-2008
C
Bivaria te Fit of Clim S core R2 By Ele v ft
te Fit
S core
an%RH 1999-2009
2
Linear
Fit
2
Linear Fit
1
Summa
ry of Fit
RSquare
0.5
RSquare Adj
Root
0 Mean Square Err or
Mean of Response
Obs
-0.5 ervations (or Sum Wgts )
r2 adj = 0.23
0.722646
0.721731
0.264458
0.550792
305
Clim Scor eR2
Clim Scor eR2
1.5
1.5
Clim Scor eR2 = 0.2995646 - 0.1858041 MeanMin Dec Temp 1999- 2008 C
1
0.5
0
-0.5
Analysi s of V aria nce
-1
Source 65 DF 70
Sum of Squares 80
Mean Square
F Ratio
85
90
-6 -5 -4 -3 -2 -1 0 1 2 3 4 755
Model
1
55.213705
1999-2009 55.2137 789.4683
MeanMin Dec Temp 1999-2008 Mean%RH
C
Error
303
21.191165
0.0699 Prob > F
C. Linear
Total Fit 304
76.404869
<.0001
Linear
Fitter Estim ates
Pa rame
Clim Scor eR2 = 3.8408658 - 0.0435131 Mean% RH 1999-2009
0.
-0.
-
1000 2000 3000 4000 5000 6000 7000
Elev ft
Bivaria te Fit of Clim S cor
2
Linear Fit
Linear Fit
1.5
Clim Scor eR2 = -0.129624 + 0.0002429 Elev ft
Summa ry of Fit
RSquare
RSquare Adj
Root Mean Square Err or
Mean of Response
Obs ervations (or Sum Wgts )
r2 adj = 0.66
0.662637
0.661679
0.289994
0.545156
354
Clim Scor eR2
1
Bivari
Clim Scor eR2
2
Clim Scor eR2
Clim Scor eR2
Bivaria te Fit of Clim S core R2 By Me anMin De c Te mp 1999-2008 C
1
0.5
0
-0.5
Analysi s of V aria nce
L
Line
Clim
Su
R
R
R
M
O
An
-1
-1
Source
DF Sum of Squares Mean Square
F Ratio
S
0
1000 2000 3000 4000 5000 6000 7000
65
70
7
Model
1
58.143161
58.1432 691.3878
M
Elev ft
Mean%R
Error
352
29.601899
0.0841 Prob > F
E
C. Total
353
87.745061
C
Linear Fit
Linear Fit<.0001
Linear Fit
Pa rame ter Estim ates
Linear Fit
Pa
What actually happened: α-Diversity
Climate Score
Shift Category
R1 αR2 αΔ in αdiversity diversity diversity
% Δ in αdiversity
N
much cooler
14
20
19
-1
-3
cooler
42
20
19
-1
-3
no change
270
22
26
4
18
warmer
38
19
24
5
27
much warmer
5
14
21
7
53
α-diversity = mean species richness per survey site
N = number of sites
Conclusions
1. The composition of epiphytic lichen communities are highly
sensitive to climate and are used by FIA to indicate climate.
2. A preliminary analysis of 10-year trends in lichen community
composition provides initial evidence of climate-driven effects
on species composition & diversity of PNW lichens.
3. Since lichen monitoring began in 1993, mean min Dec
temperatures have clearly increased in the mid and upper
elevation Cascades, but not in the Coast Range, major river
valleys or other low elevation sites.
4. Trends in lichen-community based climate scores are consistent
with temperature trends data: species composition shifts
indicate greatest warming effects at the highest elevations.
Conclusions
5. ‘Cooler’ lichen-based climate scores in the Coast Range and
low elevation Cascades, where temperature has not
decreased, may be related to increases in relative humidity.
6. Both mean species richness at survey sites and landscape
level diversity increased with warming temperatures and
decreased with cooling temperatures.
7. With regard to other plant communities, these results imply
that managers can expect different responses and different
response rates across the landscape with differing directions
and rates of change in climate variables.
Acknowledgements
We thank!—
• Our many dedicated field biologists who
tramped through the brush and surveyed
lichens
• The lichenologists who assisted with
identifications, especially Jim Riley and Pekka
Halonen.
• Our office staff for digitizing our data,
especially Anne Ingersoll and Larissa Lasselle
• Our cooperators on the Gifford Pinchot, Mt.
Hood, Willamette, Umpqua, and Siuslaw
national forests and the Columbia River Gorge
National Scenic Area who helped with
logistics, field support, and funding
• The USFS PNW Region Air Resource
Management Program and USFS PNW
Research Station/FIA Program for major
funding and support of this work.