Forest Inventory
and Analysis
National
Program
Relationship of an Indicator of Lichen Species Richness to Air Pollution and
Climate Variables Across the United States
Mark J. Ambrose1, Randall S. Morin2, and Susan Will-Wolf3
1North
Carolina State University
Dept. of Forestry & Environmental Resources
NCSU Box 8008
Raleigh, NC 27695 USA
mambrose@fs.fed.us
2USDA
Forest Service, Northern Research Station
Forest Inventory and Analysis
11 Campus Blvd., Suite 200
Newtown Square, PA 19073 USA
rsmorin@fs.fed.us
3FIA
Lichen Indicator Advisor – Cooperator
University of Wisconsin–Madison
Department of Botany
430 Lincoln Dr., Madison, WI 53706 USA
swwolf@wisc.edu
Objectives
¾To investigate the relationship between Lichen S, defined as the number of lichen species found on a standard FIA lichen
plot, and air pollution and climate variables.
¾To determine whether Lichen S, the simplest and most broadly applicable of the FIA P3 Lichen Indicator metrics, has
potential as an indicator of pollution and climate change across the country.
¾To determine whether Lichen S is consistent as an indicator at the scale of ecological regions within a large region.
Methods
We used publicly available FIA lichen, tree, and plot characteristics data (FIA Program Tools and Data. Online databases. http://fia.fs.fed.us/toolsdata/data/), collected using standard FIA field protocols.
Figure 6. Scatterplots for pollution variables: A. East - Lichen S vs.
NO3; B. East - NO3 vs. latitude; C. West - Lichen S vs. NH4; D. West
- NH4 vs. latitude.
Figure 7. Relationship of Lichen S to climate: A. Lichen S vs. precipitation; B. Lichen S
vs. avg. min. January temperature; C. Lichen S vs. avg. max. July temperature; D. Lichen
S vs. precipitation; E. Lichen S vs. avg. min. January temperature; F. Lichen S vs. avg.
max. July temperature.
Results, continued
Figure 1. The FIA lichen plot is the green
shaded area between subplots.
Figure 2. The density of plots with lichens data included in this project varies
by state.
We analyzed data from 1399 plots in the West and 1087 plots in the East representing 12 years of lichens sampling (Fig. 2). Total plot BA and hardwood
and softwood BA were taken from the FIA data. The three climate variables, precipitation, maximum July temperature, and minimum January temperature, come
from the PRISM Climate Source model (Daly and Taylor 2000). The three air pollution variables, SO42-, NO3-, and NH4+ come from a deposition model (Coulston
and others 2004). Air pollution is generally much lower in the West than in the East (Tables 1 & 2, Fig. 4).
We analyzed data separately for the East and West regions. Within East or West, we then analyzed the same data subdivided into ecoregion groups (Fig. 3)
for comparison with the entire region. We graphed the data, calculated Spearman rank correlations, and used stepwise linear regression. Scatter plots for variables
within East or West (Fig. 5, 6, 7) illustrate non-linear relationships and those where a variable represents a ‘limit’ on Lichen S; neither is easily detectable with
simple correlation or linear regression.
WEST:
¾ Lichen S is most highly correlated with geographic and climate variables (Table 2). Lichen S is higher to the northwest, at lower elevations, and
where precipitation is higher (Table 2, Fig. 3).
¾ NH4+ deposition is the air pollution variable most highly correlated with Lichen S.
¾ Unlike in the East, there is no large-scale
Table 1. Range of Lichen S and environmental explanatory variables and Spearman rank correlations
geographic pattern to pollution. Air pollution is
among all variables for the Eastern U.S.
related to scattered hot spots of agriculture and
Correlations
urbanization, and spread is affected by mountain
Min.
Max.
Variable
Range
Lichen
Tree SW
HW
Elev. Lat. Long. Precip. Jan.
ranges (Jovan and McCune 2006).
July SO4 NO3 NH4
S
BA
BA
BA
temp.
ECOREGION-SCALE ANALYSES:
¾No consistent pattern was found in the
relationship of Lichen S to environmental
variables at the ecoregion scale in either the East
or West.
¾In the East, in three ecoregion groups Lichen S
was highly correlated with air pollution variables,
while in the other two, Eastern Deciduous Forest
and Southeastern Forest (Fig. 3), correlations
were very weak.
¾ In the West, Lichen S was correlated with some
environmental variables in most ecoregions, but
the patterns varied among ecoregions (e.g., in
some regions Lichen S was positively correlated
with air pollution and in others, negatively
correlated; significant variables vary by
ecoregion).
¾In many ecoregions, indicators based on lichen
species composition (McCune and others 1997;
Geiser & Neitlich 2006 in press; Will-Wolf and
others 2006) may be more useful than Lichen S.
Figure 3. Ecoregion groups used for analysis. Printed numbers are average
Lichen S, the number of lichen species per plot, for each ecoregion. Groups
are based on Bailey’s ecoregion divisions and provinces.
Results
¾ Correlations among environmental variables differ from West to
East (Tables 1 & 2).
¾Many environmental explanatory variables are more highly
correlated with each other than any are with Lichen S.
¾ Average Lichen S is slightly higher in the East, but overall
variability and maximum Lichen S are both higher in the West (Fig.
3, Tables 1 & 2).
¾ Lichen S was, at most, weakly related to stand variables (total
BA, HW BA, and SW BA) at all scales and in all regions.
Figure 4. Average pollution values by ecoregion.
temp.
0 to 33
1.000
0.097
0.152
0.044
0.021
-0.186
-0.190 -0.387 -0.463 -0.363 -0.052
Elevation (m)
0 to 1660
0.097
1.000
0.247
-0.088
-0.178
-0.406
-0.553
Latitude (dd)
30.64 to 49.18
0.152
0.247
1.000
0.436
-0.554
-0.966
-0.828 -0.439 -0.129 -0.034
0.032
0.126 -0.063
0.044 -0.088
0.436
1.000
0.070
-0.409
-0.513 -0.101 -0.001 -0.553
0.111
0.225 -0.048
483 to 1919
0.021 -0.178 -0.554
0.070
1.000
0.514
0.311
0.058 -0.119 -0.426
0.032
0.107 -0.039
Longitude (dd)
-67.11 to -96.95
Precipitation (mm)
Min. Jan. temp. (˚C)
0.211
0.297
0.186
0.166 -0.124
0.066 -0.218
0.230
-24.3 to 5.0
-0.186 -0.406 -0.966
-0.409
0.514
1.000
0.873
0.397
0.093
0.050 -0.044 -0.109
Max. July temp. (˚C)
21.9 to 33.9
-0.190 -0.553 -0.828
-0.513
0.311
0.873
1.000
0.248
0.007
0.172 -0.079 -0.079 -0.027
SO4 (kg/ha/yr)
4.32 to 28.98
-0.387
0.211 -0.439
-0.101
0.058
0.397
0.248
1.000
0.883
0.507
0.028 -0.422
0.329
5.7 to 22.51
NO3 (kg/ha/yr)
-0.463
0.297 -0.129
-0.001
-0.119
0.007
0.883
1.000
0.593
0.081 -0.384
0.350
0.7 to 4.32
-0.363
0.186 -0.034
-0.553
-0.426
0.050
0.172
0.507
0.593
1.000 -0.015 -0.383
0.250
Tree BA (ft²/acre)
22.1 to 579.49
-0.052
0.066
0.032
0.111
0.032
-0.044
-0.079
0.028
0.081 -0.015
1.000
0.266
SW BA (ft²/acre)
0 to 290.30
0.166 -0.218
0.126
0.225
0.107
-0.109
-0.079 -0.422 -0.384 -0.383
0.266
1.000 -0.498
HW BA (ft²/acre)
0 to 579.49
0.230 -0.063
-0.048
-0.039
0.039
-0.027
0.584 -0.498
NH4 (kg/ha/yr)
-0.124
0.093
0.039
0.329
0.350
0.250
0.584
1.000
Table 2. Range of Lichen S and environmental explanatory variables and Spearman rank correlations
among all variables for the Western U.S.
Correlations
Long.
Precip.
Min.
Jan.
temp.
Max.
July
temp.
NO3
NH4
Tree
BA
SW
BA
0.487
-0.485
0.490
0.358
-0.140
0.152
-0.196
-0.329
0.255
0.208
0.155
-0.400
0.714
-0.303
-0.773
-0.335
-0.162
0.326
0.209
-0.069
0.001
-0.190
-0.400
1.000
-0.418
0.499
0.009
-0.488
0.018
-0.329
-0.497
0.175
0.244
-0.190
0.714
-0.418
1.000
-0.560
-0.751
0.012
0.132
0.552
0.441
-0.222
-0.156
-0.167
-0.522
Variable
Range
Lichen
S
Elev.
Lat.
Lichen S
0 to 45
1.000
-0.528
0 to 3,740
-0.528
1.000
31.42 to 48.96
0.487
-103.1 to -124.41
-0.485
Elevation (m)
Latitude (dd)
Longitude (dd)
Precipitation (mm)
171 to 4,731
SO4
HW
BA
0.490
-0.303
0.499
-0.560
1.000
0.327
0.093
-0.231
-0.379
0.379
0.306
0.177
Min. Jan. temp. (˚C)
-22.6 to 4.7
0.358
-0.773
0.009
-0.751
0.327
1.000
0.349
0.105
-0.277
-0.134
0.116
-0.003
0.305
Max. July temp. (˚C)
15.7 to 38.3
-0.140
-0.335
-0.488
0.012
-0.522
0.349
1.000
-0.117
-0.066
0.346
-0.260
-0.304
0.124
SO4 (kg/ha/yr)
0.18 to 7.53
0.152
-0.162
0.018
0.132
0.093
0.105
-0.117
1.000
0.727
0.218
0.099
0.049
0.147
NO3 (kg/ha/yr)
0.38 to 5.9
-0.196
0.326
-0.329
0.552
-0.231
-0.277
-0.066
0.727
1.000
0.571
-0.014
-0.034
NH4 (kg/ha/yr)
0.0 to 2.14
-0.329
0.209
-0.497
0.441
-0.379
-0.134
0.346
0.218
0.571
1.000
-0.161
-0.211
Tree BA (ft²/acre)
1.51 to 650.24
0.255
-0.069
0.175
-0.222
0.379
0.116
-0.260
0.099
-0.014
-0.161
1.000
0.893
0.102
SW BA (ft²/acre)
0 to 650.24
0.208
0.001
0.244
-0.156
0.306
-0.003
-0.304
0.049
-0.034
-0.211
0.893
1.000
-0.243
HW BA (ft²/acre)
0 to 242.67
0.155
-0.190
-0.190
-0.167
0.177
0.305
0.124
0.147
0.048
0.132
0.102
-0.243
1.000
0.048
0.132
¾ Simple linear regression models did not explain Lichen S very well at large scales:
Best models: East: r2 = 0.307, significant variables: NO3-, elevation, NH4 +, min. Jan. temp., SW BA, total BA.
West: r2 = 0.404, significant variables: elevation, latitude, total BA, max. July temp., NH4+.
¾ Need to explore other model forms. Scatter plots of data suggest non-linear relationships.
Conclusions
EAST:
¾ Air pollution variables are all highly correlated (Table 1).
¾ Lichen S was most highly correlated with NO3- deposition (Table
1, Fig. 6A). The areas of lowest Lichen S and highest deposition
occur at approx 41˚N latitude (Figs. 5A & 6B). This area is also
one that has been heavily cleared for agriculture and urban
development and otherwise affected by human activity.
¾Lichen S was not strongly correlated with any geographic or
climate variable.
Literature Cited
REGRESSION ANALYSES:
Lichen S
Figure 5. Lichen S vs. geographic variables: A. Lichen S vs.
latitude; B. Lichen S vs. elevation; C. Lichen S vs. latitude; D.
Lichen S vs. elevation.
Coulston, J.W.; Riitters, K.H.; Smith, G.C. 2004. A Preliminary Assessment of Montreal Process Indicators of Air Pollution for the United States. Environmental Monitoring and Assessment. 95: 57-74.
Daly, C.; Taylor, G. 2000. United States Average Monthly or Annual Precipitation, Temperature, and Relative Humidity 1961-90. Arc/INFO and ArcView coverages. Corvallis, OR: Spatial Climate Analysis
Service at Oregon State University (SCAS/OSU).
Geiser, L.H.; Neitlich, P. 2006 [in press]. Air pollution and climate gradients in western Oregon and Washington indicated by epiphytic macrolichens. Environmental Pollution.
Jovan, S.; McCune, B. 2006. Using epiphytic macrolichen communities for biomonitoring ammonia in forests of the Greater Sierra Nevada, California. Water, Air, and Soil Pollution 170: 69–93.
McCune, B.; Dey, J.; Peck, J.; [and others]. 1997. Regional gradients in lichen communities of the Southeast United States. Bryologist 100: 145-158.
Will-Wolf, S.; Geiser, L.H.; Neitlich, P.; Reis, A. 2006. Forest lichen communities and environment - How consistent are relationships across scales? Journal of Vegetation Science 17: 171-184.
¾ Lichen S, the number of lichen species found on an FIA plot, does indicate a 'response of lichen communities' to
underlying environmental factors at large scales, BUT
¾ The environmental variables used here do not necessarily map perfectly to the underlying environmental factors
affecting lichens.
¾ Lichen S is related to pollution (and/or other human influences) in both East and West.
¾ Lichen S is most strongly related to geographic/climate variables in the West.
¾ In the East, pollution (and related human influences) are more strongly correlated with Lichen S than any other
environmental variable.
¾ Lichen S has some potential as an indicator of pollution and/or climate change both at the large scale and for some of
the ecoregions we investigated, BUT
¾ Complex relationships among variables affecting lichen species richness present challenges to any operational use of
lichen S as an indicator of pollution and/or climate change.
Acknowledgements
This project was funded in part by the U.S. Department of Agriculture Forest Service under Cooperative Agreement Forest Health Assessment SRS 05-CA-11330146-180 with
North Carolina State University (an equal opportunity provider) and under Cooperative Agreement SRS 05-CA-11330145-163 with University of Wisconsin-Madison.