Environmental Pollution 112 (2001) 163±169
www.elsevier.com/locate/envpol
Relationship between epiphytic lichens, trace elements and gaseous
atmospheric pollutants
H.F. van Dobben a,*, H.Th. Wolterbeek b, G.W.W. Wamelink a, C.J.F. Ter Braak c
a
Alterra Green World Research, PO Box 47, 6700 AA Wageningen, The Netherlands
Interfaculty Reactor Institute, Delft University of Technology, Mekelweg 15, 2629 JB Delft, The Netherlands
c
Centre for Biometry Wageningen, CPRO-DLO, PO Box 16, 6700 AA Wageningen, The Netherlands
b
Received 26 February 1999; accepted 18 March 2000
``Capsule'': Lichens were shown to be e€ective accumulator biomonitors for heavy metals in The Netherlands.
Abstract
A study was conducted to determine the joint e€ect of gaseous atmospheric pollutants and trace elements on epiphytic lichens.
We used our data to test the hypothesis that lichens are generally insensitive to toxic e€ects of trace elements, and can therefore be
used as accumulator organisms to estimate concentrations of these elements in the environment. In a ®eld study in The Netherlands
the abundance of epiphytic lichen species was estimated, and their supporting bark was collected. Concentrations of a range of trace
elements were determined in the bark, and concentrations of atmospheric trace gases were estimated at the sites of collection.
Multivariate statistics were used to determine the relation between the abundance of the species and pollutant concentrations.
Atmospheric SO2 and NO2 appeared to be the most important factors determining lichen biodiversity. Nearly all species were
sensitive to these compounds. The e€ect of the other trace elements was very slight; only Sb had a signi®cantly negative e€ect on the
abundance of a few species. It is concluded that lichens can safely be used as accumulator organisms in pollution studies, provided
that concentration in lichen thalli re¯ect atmospheric concentrations. # 2001 Elsevier Science Ltd. All rights reserved.
Keywords: Lichen; Bark; Trace element; Air pollution; Sensitivity; Multivariate statistics
1. Introduction
Lichens are generally considered as useful organisms
to monitor air quality. Numerous studies have been
undertaken in which lichen biodiversity, or the
presence/absence of indicator species, was used as a
measure for air quality (Seaward, 1992). Information on
the pollutants involved and their working mechanisms is
scarce, however. Most authors implicitly (e.g. Barkman,
1958) or explicitly (e.g. De Wit, 1976) state SO2 as the
main cause for the decline of lichens in polluted areas,
but others claim an additional sensitivity to NO2 (Nash,
1976), O3 (Sigal and Nash, 1983), NH3 (Van Dobben
and De Bakker, 1996), ¯uoride (Nash, 1971), heavy
metals (Folkesson and Andersson-Bringmark, 1988) or
air pollutants in general (Nylander, 1866; Herzig et al.,
1989; Garty et al., 1993).
* Corresponding author. Tel.: +31-317-477936; fax: +31-317424988.
E-mail address: [email protected] (H.F. van Dobben).
Besides indicating air quality by their presence or
absence, lichens have also been used as accumulator
organisms in studies on atmospheric trace element pollution (Lippo et al., 1995; Sloof, 1995; Loppi and
Bargagli, 1996). In these studies the elemental concentrations in lichens are considered to re¯ect atmospheric concentration or deposition, irrespective of
possible toxic e€ects. In fact an implicit hypothesis is
made in this type of study, namely that trace elements
do not have toxic e€ects on lichens (Purvis, 1996). If the
elements under consideration would kill the monitoring
organisms they would be absent in the most polluted
spots, and estimated concentrations would become
biased towards lower values.
The aim of our study was to determine the joint e€ect
of gaseous atmospheric pollutants and trace elements
on epiphytic lichens. A second aim was to test the
implicit hypothesis that trace elements do not have toxic
e€ects on lichens. We recorded epiphytic lichen vegetation, bark chemical composition and atmospheric concentrations of SO2, NO2 and NH3 at 123 sites along a
0269-7491/01/$ - see front matter # 2001 Elsevier Science Ltd. All rights reserved.
PII: S0269-7491(00)00121-4
164
H.F. van Dobben et al. / Environmental Pollution 112 (2001) 163±169
number of east±west and north±south transects through
The Netherlands. We assumed that atmospheric concentrations of trace elements are re¯ected by their
concentrations in bark (cf. Wolterbeek et al., 1996). A
multivariate statistical approach was used to determine
the joint e€ect of atmospheric chemistry and bark
chemistry on the composition of the lichen vegetation.
The analysis focused on determining the e€ect of trace
element in bark after accounting for the e€ects of gaseous pollutants and a number of ecologically relevant
variables (tree species, tree diameter, distance to the
coast).
2. Materials and methods
2.1. Sampling
Table 1
Minimal, mean and maximal concentrations in bark found in our
data, in ppm on a dry weight basisa
Element
Min.
Mean
Max.
As
Br
Ca
Cd
Ce
Co
Cr
Cs
Fe
Hg
K
La
Na
Ni
Sb
Sc
Se
Sm
Th
Zn
NH4
NO3
SO4
pH
1.98E-01
6.10E+00
7.94E+03
2.25E-01
1.37E+00
2.35E-02
5.50E+00
6.46E-02
1.40E+01
5.13E-02
2.81E+02
6.62E-01
3.17E+02
2.91E+00
9.00E-01
1.17E-01
2.45E-01
7.43E-02
1.55E-01
5.91E+01
3.77E+00
2.66E-01
6.05E+00
3.85E+00
1.14E+00
1.98E+01
1.62E+04
2.58E+00
4.32E+00
1.34E+00
1.32E+01
2.70E-01
1.88E+03
2.16E-01
1.34E+03
2.05E+00
9.34E+02
1.03E+01
2.15E+00
4.11E-01
6.24E-01
2.77E-01
3.81E-01
1.97E+02
1.64E+01
1.28E+00
2.32E+01
4.66E+00
4.60E+00
6.76E+01
4.67E+04
1.04E+01
1.61E+01
4.28E+00
4.83E+01
1.04E+00
6.94E+03
6.35E+00
3.31E+03
6.99E+00
2.90E+03
3.37E+01
2.69E+01
1.30E+00
1.68E+00
3.65E+00
1.37E+00
7.81E+02
5.43E+01
7.52E+00
9.23E+01
5.40E+00
The sampling stations consisted of rows of 10 trees,
distributed along six transects through The Netherlands
(see Wolterbeek et al., 1996, for a map). There were
123 sampling stations, on the following tree species:
Quercus robur L. (65 stations), Populuscanadensis
Moench (43 stations), Salix alba L. (®ve stations) and
Ulmushollandica Miller (10 stations). Sampling near
farms, villages and industrial sites was avoided. The
lichen vegetation of the sampled trees mostly belonged
to the alliance Xanthorion parietinae Ochsn. 1928. Sampling was carried out in the period August±October
1990 (for details, see Van Dobben, 1993). The abundance of all lichen species present on the trunks of the
trees from the base up to a height of 2 m was estimated
and scored on a six-point scale (Van Dobben, 1993).
Most species were identi®ed in the ®eld but individuals
that were not readily recognisable were sampled for
later identi®cation. Nomenclature follows Aptroot et al.
(1999). Bark ¯akes of ca. 5 mm thickness were cut from
the three middle trees of each station at a height of ca.
1.5 m. These samples were air-dried and ground to a
grain size of <1 mm by sieve-milling.
Data were obtained from the Dutch Air Quality
Monitoring Network (Anonymous, 1990). SO2 and NO2
were estimated as means of hourly measured concentrations (SO2 April±September 1989, NO2 June 1989±May
1990) at monitoring stations, followed by interpolation
(Van Egmond et al., 1978) of the concentrations at the
sample points. Mean NH3 concentrations were estimated on a 55-km2 grid basis using the 1988 emission
data and the atmospheric transport and deposition
model TREND (Asman and Van Jaarsveld, 1990).
2.2. Bark analysis
2.4. Statistical analysis
Analytical procedures are described in detail by Wolterbeek et al. (1996). Concentrations of NO3, NH4 and
SO4 were determined in aqueous extracts using colorimetric methods. pH was measured in the same extract.
Lead was determined by graphite-furnace atomic
absorption spectrometry (AAS) after digestion in
HNO3/HClO4/HF. A large number of other elements
(As, Br, Ca, Cd, Ce, Co, Cr, Cs, Fe, Hg, K, La, Na, Ni,
Sb, Sc, Se, Sm, Th and Zn) were determined by Instrumental Neutron Activation Analysis (Bode and De Bruin,
1990). All these elements will be further denoted as trace
elements. The mean, minimal and maximal contents per
element are summarized in Table 1.
Redundancy analysis (RDA; Ter Braak, 1994) was
used to determine the relation between atmospheric
chemistry, bark chemistry and the composition of the
lichen vegetation. The analysis was carried out using
the program Canoco 4.0 (Ter Braak and Smilauer, 1998).
All bark chemical variables except pH were expressed in
ppm on a dry weight basis and logarithmized. Air quality
variables were expressed in mg mÿ3 and not transformed.
The other variables considered were: (1) tree species,
determined as the contrast oak versus other tree species
(cf. Wolterbeek et al., 1996), and entered as a dummy
variable (1=oak, 0=other); (2) tree diameter (DBH in
cm, untransformed); and (3) distance to the coast (in km,
a
Means are geometric means except pH.
2.3. Air pollution
H.F. van Dobben et al. / Environmental Pollution 112 (2001) 163±169
untransformed, cf. Van Dobben and De Bakker, 1996).
Interaction terms were not considered.
The importance of the explanatory variables was
determined by stepwise selection. In each step the `extra
®t' was determined for each variable, i.e. the increase in
regression sum of squares over all species when adding a
variable to the regression model. The variable with the
largest extra ®t was then included, and the process was
repeated until no variables remained that could signi®cantly improve the ®t. The statistical signi®cance of
the e€ect of including a variable was determined by
means of a Monte Carlo permutation test (for details,
see Ter Braak and Smilauer, 1998).
The results of the multivariate analysis were visualised
by means of biplots. A biplot attempts to optimally
represent the joint e€ect of the environmental variables
on all species in a single plane (Ter Braak, 1995). In our
biplots, species are indicated by their abbreviated names
and environmental variables by arrows. Arrows can also
be drawn from the origin to each species' point located in
the centre of its name (actually these arrows were not
drawn to avoid overcrowding of the plot). The cosine of
the angle between each pair of arrows (species±species,
species±environment, or environment±environment) is a
measure for the correlation between that pair of variables (sharp angles indicate positive correlations, obtuse
angles indicate negative correlations). In general, species
and environmental variables with the longest arrows are
best represented in the biplot. Further detail on the
interpretation of biplots is given by Ter Braak (1995).
In this study the results of a single analysis are presented in two biplots. In a ®rst biplot the e€ect is shown
of all environmental variables that signi®cantly contribute to the ®t of the model. This biplot appeared to
be strongly dominated by the e€ect of a few variables
that explain about half of the variance explained by the
full model. In this biplot the e€ect of the less important
variables (that still signi®cantly contribute to the ®t of
the model) is blurred by these few variables. Therefore,
a second biplot was drawn after accounting for the
e€ect of the most dominant variables (technically this
was achieved by declaring these variables as covariables
in CANOCO). Finally, the e€ect of the trace elements
that most strongly a€ected the species (as shown by the
multivariate analysis) was analysed in detail by (univariate) logistic regression of the presence/absence of
the most common species on the trace element concentrations, after accounting for the e€ect of all other
relevant variables. The logistic regression was carried
out by the program GENSTAT 5.3 (Payne et al., 1993).
3. Results
In total, 72 species were found in the 123 sampling stations. The mean number of species per
165
sampling station was 18. The result of the stepwise
selection in RDA is given in Table 2. The full model (i.e.
containing all variables that signi®cantly contribute to
the ®t) explains ca. 40% variance, about half of which is
explained by the air quality variables SO2 and NO2.
Tree species and bark pH each explain ca. 4% variance,
and small additional percentages variance are explained
by tree DBH, atmospheric NH3 and the bark chemical
factors Br, Sb, As, Ca, and NH4. Bark concentrations of
SO4, NO3 and the remaining trace elements included in
the analysis did not have a signi®cant e€ect on the
abundance of the lichen species, and there was no signi®cant e€ect of distance to the coast.
Fig. 1 is the biplot representing the e€ect of all variables in the model of Table 2. The species shown are
those for which this model explained >10% of the variance in their abundance. The biplot clearly shows the
dominant e€ect of SO2 and NO2 (mainly represented
along the horizontal axis). All species except Lecanora
conizaeoides Nyl. ex Cromb. are negatively correlated
with these variables which therefore strongly a€ect general species richness. A second important source of variation is represented along the vertical axis, which is
mainly determined by atmospheric NH3, bark pH and
bark Ca on the lower side, and oak which works in the
opposite direction. The so-called `nitrophytic' species
(e.g. Physcia spp. and Xanthoria spp., cf. Van Dobben
and Ter Braak, 1999) have a low position relative
to this axis, whereas the `acidophytic' species (e.g.
Hypogymnia physodes (L.) Nyl., Evernia prunastri (L.)
Ach.) have a high position.
Table 2
Result of stepwise selection of variables using redundancy analysis
(RDA)a
Variable
Extra ®t
Cumulative ®t
Signi®cance
Atm. SO2
Atm. NO2
Bark pH
Oak
Bark Br
Bark Sb
Bark As
Bark Ca
Bark NH4
Atm. NH3
Tree DBH
15.7
5.4
4.1
4.5
2.2
1.7
1.7
1.6
1.3
1.0
0.9
15.7
21.1
25.3
29.7
32.0
33.7
35.4
37.0
38.3
39.3
40.2
***
***
***
***
***
***
***
**
**
*
*
a
Variables are given in the order of inclusion. The extra and
cumulative ®t are given as percentages relative to the total sum of
squares over all species (comparable to the percentage explained variance in univariate regression). Number of observations: 123; total
number of species: 72. Signi®cance was determined by Monte Carlo
permutation using 999 random permutations. Di€erences in the
cumulative ®t and the sum of the extra ®t and the cumulative ®t in the
preceding row are due to rounding errors.
*0.01<P40.05.
**0.001<P40.01.
***P40.001
166
H.F. van Dobben et al. / Environmental Pollution 112 (2001) 163±169
Fig. 1. Correlation biplot showing the relations between species (names) and environmental variables (arrows), based on the model in Table 2. See
text for further explanation. Eigenvalues: 0.216, 0.093 and 0.029 for the ®rst (horizontal), second (vertical) and third axis (not shown). Only species
are shown for which the model explains >10% of its variance. Explanation of species names: Art.rad, Arthonia radiata (Pers.) Ach.; Buel.gris,
Buellia griseovirens (Turner & Borrer ex Sm.) Almb.; Cal.lut, Caloplaca luteoalba (Turner) Th. Fr.; Can.aur, Candelariella aurella (Ho€m.) Zahlbr.;
Can.ref, Candelariella re¯exa (Nyl.) Lettau; Can.vit, Candelariella vitellina (Ho€m.) MuÈll. Arg.; Can.xan, Candelariella xanthostigma (Ach.) Lettau;
Cet.chl, Cetraria chlorophylla (Willd.) Vainio; Chaen.f, Chaenotheca ferruginea (Turner & Borrer) Mig.; Clad.sp, Cladonia species; Cli.grif, Cliostomum grithii (Sm.) Coppins; Dim.pin, Dimerella pineti (Ach.) Vezda; Ever.pru, Evernia prunastri (L.) Ach.; Hypoc.sc, Hypocenomyce scalaris (Ach.)
M. Choisy; Hyp.phy, Hypogymnia physodes (L.) Nyl.; Hyp.tub, Hypogymnia tubulosa (Schaerer) Havaas; L.carp, Lecanora carpinea (L.) Vainio;
L.chlar, Lecanora chlarotera Nyl.; L.coniz, Lecanora conizaeoides Nyl. ex Crombie; L.disp, Lecanora dispersa (Pers.) Sommerf.; L.expal, Lecanora
expallens Ach.; L.pulic, Lecanora pulicaris (Pers.) Ach.; L.symm, Lecanora symmicta (Ach.) Ach.; Lec.eleo, Lecidella elaeochroma (Ach.) M. Choisy;
Lepr.inc, Lepraria incana (L.) Ach.; Op.niveo, Opegrapha niveoatra (Borrer) J.R. Laundon; P.acet, Parmelia acetabulum (Necker) Duby; P.exasp,
Parmelia exasperata De Not.; P.lacin, Parmelia laciniatula (Flagey ex H. Olivier) Zahlbr.; P.revol, Parmelia revoluta FloÈrke; P.saxat, Parmelia saxatilis (L.) Ach.; P.subau, Parmelia subaurifera Nyl.; P.subru, Parmelia subrudecta Nyl.; P.sulc, Parmelia sulcata Taylor; Pert.alb, Pertusaria albescens
(Hudson) M. Choisy & Werner; Pert.ama, Pertusaria amara (Ach.) Nyl.; Pert.coc, Pertusaria coccodes (Ach.) Nyl.; Ph.orbic, Phaeophyscia orbicularis (Necker) Moberg; Phl.arg, Phlyctis argena (Sprengel) Flotow; Ph.adsc, Physcia adscendens (Fr.) H. Olivier; Ph.caes, Physcia caesia (Ho€m.)
FuÈrnrohr; Ph.stel, Physcia stellaris (L.) Nyl.; Ph.tene, Physcia tenella (Scop.) DC.; Ph.ente, Physconia enteroxantha (Nyl.) Poelt; Ph.gris, Physconia
grisea (Lam.) Poelt; Ps.furf, Pseudevernia furfuracea (L.) Zopf; Pyr.que, Pyrrhospora quernea (Dickson) KoÈrber; R.farin, Ramalina farinacea (L.)
Ach.; R.fast, Ramalina fastigiata (Pers.) Ach.; X.cand, Xanthoria candelaria (L.) Th. Fr.; X.par, Xanthoria parietina (L.) Th. Fr.; X.polyc, Xanthoria
polycarpa (Ho€m.) Th. Fr. ex Rieber.
The general relation between the occurrence of epiphytic lichens and environmental conditions as shown
by this biplot is very similar to the relation reported by
Van Dobben and De Bakker (1996) on the basis of an
independent dataset. There are two main directions of
variation: (1) species poor versus species rich, represented along the horizontal axis and mainly determined
by the presence of the atmospheric pollutants SO2 and
NO2; and (2) nitrophyte-dominated versus acidophytedominated, represented along the vertical axis and
mainly determined by factors related to bark pH. Fig. 1
shows that the abundance of the nitrophytic species is
positively correlated with bark pH and Ca concentration, and with atmospheric ammonia (which alkalises
bark, cf. Van Dobben and De Bakker, 1996), and
negatively correlated with oak (which has a naturally
acid bark). For the acidophytes these relations are viceversa. The biplot indicates that Br concentration is
positively related to species richness, and that the e€ect
of Ca is similar to the e€ect of pH. It does not, however,
yield clear indications on the e€ects of the two other
trace elements included (As and Sb).
The second biplot (Fig. 2) is based on the same model
but with the e€ect of SO2 and NO2 accounted for. In
this biplot species are shown for which the model (after
adjusting for the e€ects of SO2 and NO2) explained
>3% of their variance. In this plot pH has become the
variable with the most important e€ect (mainly represented along the horizontal axis), together with
Ca concentration in bark, and atmospheric NH3. The
H.F. van Dobben et al. / Environmental Pollution 112 (2001) 163±169
167
Fig. 2. Correlation biplot showing the relations between species (names) and environmental variables (arrows), based on the model in Table 2 after
accounting for the e€ects of SO2 and NO2. Eigenvalues: 0.100, 0.026 and 0.020 for the ®rst (horizontal), second (vertical) and third axis (not shown).
Only species are shown for which the model explains >3% of its variance. See Fig. 1 for an explanation of the species names.
horizontal axis now separates the nitrophytic species
from the acidophytic ones, like the vertical axis in Fig.
1. The vertical axis is now mainly determined by tree
DBH and Br concentration. Species that preferably
occur on mature trees tend to have positions at the top
of Fig. 2 (e.g. Parmelia acetabulum (Necker) Duby,
Ramalina fastigiata (Pers.) Ach.), whereas pioneer species tend to be at the base of the plot (e.g. Xanthoria
polycarpa (Ho€m.) Th. Fr. ex Rieber, Lecanora symmicta (Ach.) Ach.). In this plot the e€ect of Sb concentration seems to be comparable to the e€ect of low
pH because of its high score on the horizontal axis (i.e.
its arrow pointing in a direction opposite to the `pH'
arrow). However, As has a very short arrow and its
e€ect is therefore poorly represented in the plot.
The results of the univariate logistic regression are
shown in Table 3. Out the 35 species with more than 10
occurrences, six are signi®cantly (P40.05) correlated
with the Sb concentration (®ve negatively, one positively). None of the species is signi®cantly negatively
correlated with the As concentration, and only one is
signi®cantly positively correlated with this element.
The signi®cantly positive correlations with As and Sb are
found for the same species (L. conizaeoides), which is
also the only species whose occurrence is positively correlated with the gaseous pollutants SO2 and NO2. This
species therefore seems to be truly multi-resistant. In
contrast, the species that appear to be most sensitive to
Sb (Xanthoria parietina (L.) Th.Fr. and Physconia grisea
(Lam.) Poelt) do not have extreme sensitivities to either
SO2 or NO2 (cf. Van Dobben and Ter Braak, 1999).
4. Discussion
Our results clearly show the order of importance of
the factors determining epiphytic vegetation on wayside
trees in The Netherlands. Most important are the toxic
atmospheric pollutants SO2 and NO2. Nearly all species
decrease with increasing concentrations of these compounds, which therefore strongly negatively a€ect species diversity. This phenomenon is the basis of a large
number of bioindicator studies using lichens (e.g. Barkman, 1958; Skye, 1968; LeBlanc and De Sloover, 1970;
De Wit, 1976; Anonymous, 1995; Van Dobben and Ter
Braak, 1999). Second in the order of importance are the
ecological factors that show a natural variation, such as
bark pH, tree species and tree DBH. Rather than
determining general species diversity, these factors
determine the species composition of the vegetation. In
the third place are the trace elements. Most of the elements tested in this study did not have a signi®cant
e€ect, with the exception of Br, Ca, Sb and As. Of these
four, Ca and Br probably do not have toxic a€ects, and
can be considered as `normal' ecological factors with a
natural variation.
168
H.F. van Dobben et al. / Environmental Pollution 112 (2001) 163±169
Table 3
Sign and signi®cance of the regression coecients for the terms As and
Sb of the logistic regression equation with modela
Species
n
As
Sb
Lecanora carpinea (L.) Vainio
Lecanora conizaeoides Nyl. ex Cromb.
Lecidella elaeochroma (Ach.) Choisy
Phaeophyscia orbicularis (Necker) Moberg
Physconia grisea (Lam.) Poelt
Xanthoria parietina (L.) Th.Fr.
43
63
73
52
14
89
0
+1
0
0
0
0
ÿ1
+1
ÿ1
ÿ2
ÿ1
ÿ2
a
SO 2+NO 2+NH 3 +oak+DBH+Ca+Br+NH4 +pH+[As or
Sb], with: SO2, NO2, NH3: concentrations of SO2, NO2, NH3 in air, in
mg mÿ3; oak; dummy variable; 1=oak, 0=other tree; DBH, tree diameter at 1.50 m above ground level, in cm; As, Sb, Ca, Br, NH4,
logarithmised concentrations in bark of As, Sb, Ca, Br, NH4, in ppm;
pH, bark pH. The sign of each entry is the sign of the regression
coecient and its magnitude denotes signi®cance determined on the
basis of t-values: 2=0.001<P40.01; 1=0.01<P40.05; 0=P>0.05.
Only species are given with >10 occurences and a signi®cant
(P40.05) e€ect of at least one of the elements As or Sb. Number of
observations: 123; total number of species: 72; number of species with
>10 occurrences: 35.
Ca is positively correlated with bark pH (r=0.34),
and both have a very similar e€ect on the lichen vegetation (Fig. 1). Apparently bark pH is to a certain
degree in¯uenced by the presence of Ca-containing buffer substances, which most probably originate from
calcareous soil dust. Br is strongly negatively correlated
with distance to the coast (r=ÿ0.61). This element is
present in seawater in relatively high amounts (Weast et
al., 1987), and probably has to be considered as just an
indicator for oceanic in¯uence. It is therefore uncertain
whether there is a direct in¯uence of Br. Previous studies have shown a general decrease in species diversity
at greater distances from the coast (Van Dobben and
De Bakker, 1996). In the present study there was no
signi®cant in¯uence of distance to the coast, but a positive correlation between the occurrence of most lichen
species and Br concentration (Fig. 1). Apparently the
joint e€ect of the intercorrelated variables Br and distance to the coast is best explained by the former one.
Their joint e€ect may be interpreted as either a direct
e€ect of sea-spray ions (i.e. Br or correlated elements),
or as the result of a climatic gradient with Br as an
indicator for oceanic conditions. A climatic e€ect with
increasing species diversity in more oceanic conditions
was hypothesized by Van Dobben and De Bakker
(1996).
The e€ects of As and Sb probably come about though
direct toxicity, as these elements are known to be toxic
to numerous organisms (Gebel, 1997). However, the
limited e€ect of the trace elements considered in this
study indicates that there is only a slight risk that
lichens are killed by extreme concentrations. Sensitivity
could only be shown for a limited number of species,
and to a limited number of elements. We therefore
conclude that epiphytic lichens can safely be used as
accumulator organisms in pollution studies, provided
that concentration in lichen thalli do re¯ect atmospheric
concentrations. This conclusion is supported by studies
in which extreme concentrations (up to several per cents
on a dry weight basis) of potentially toxic elements have
been recorded in apparently healthy lichen thalli (e.g.
Lambinon et al., 1964; Shimwell and Laurie, 1972;
Brown and Beckett, 1984; Purvis, 1984), or where lichen
communities with a high biodiversity are reported from
substrates with high concentrations of potentially toxic
metal ions (Purvis and Halls, 1996).
Another conclusion is that species sensitive to one
element are not automatically sensitive to the others. A
comparable conclusion was drawn in an earlier study on
the e€ect of atmospheric trace gases (Van Dobben and
Ter Braak, 1999). Although in that study many species
were found to signi®cantly respond to any of the gaseous pollutants SO2, NO2 and NH3, only a few signi®cantly responded to all of them, and there were no
two species that showed the same type of response
(increase or decrease) to all three pollutants. Apparently, lichens in general are not suitable indicators for
the general level of pollution as has sometimes been
hypothesized (Section 1). Most obvious is their sensitivity to SO2, and in some cases this compound may in
itself be indicators for the general level of pollution
(Cislaghi and Nimis, 1997).
Our conclusions are at variance with those of Herzig
et al. (1989), who claim a species-inspeci®c sensitivity of
lichens to both trace metals, gaseous pollutants and
dust. However, among the trace metals studied by this
group (Pb, Cu, Zn, Cd), only Cd substantially contributed to the explained variance in a regression model
of lichen species number against pollutant concentrations. Unfortunately, standard deviations of individual
regression coecients are not given in this study, and
therefore it is not possible to determine whether these
correlations represent a real e€ect or are caused by
a correlation with SO2.
There seems to be one exception to the general rule of
species-speci®c sensitivity, namely L. conizaeoides. This
species has apparently adapted itself completely to the
industrial environment by developing resistance to both
toxic trace gases and trace metals. The resistance of this
species to SO2 has also been shown experimentally
(Bates et al., 1996). L. conizaeoides has dramatically
expanded all over Europe and northern America during
the last century (Wirth, 1993). However, presently it is
in a process of decline, concomitant with the general
improvement of air quality in these regions (Wirth,
1993; Kirschbaum et al., 1996; Van Dobben and De
Bakker, 1996). In areas where L. conizaeoides is common air pollution may therefore be suciently monitored by mapping this single species. It will then act as
a `reverse' monitor, indicating pollution by its presence
H.F. van Dobben et al. / Environmental Pollution 112 (2001) 163±169
rather then by its absence as with most pollution monitors. However, even when using this species there may
be interference from NH3 which causes it to decrease.
References
Anonymous, 1990. Luchtkwaliteit - jaarverslag (1989). Report of the
National Institute of Public Health and Environmental Hygiene
222101006, Bilthoven, The Netherlands.
Anonymous, 1995. Measurement of Immission E€ects: Measurement
and Evaluation of Phytotoxic E€ects of Ambient Air Pollutants
(Immissions) with Lichens; Mapping of Lichens for Assessment of
Air Quality. VDI-Richtlinien 3799.
Aptroot, A., Van Herk, K., Sparrius, L., Van den Boom, P., 1999.
Checklist van de Nederlandse lichenen en lichenicole fungi. Buxbaumiella 50, 4±64.
Asman, W.A.H., Van Jaarsveld, H.A., 1990. A Variable-Resolution
Statistical Transport Model Applied for Ammonia and Ammonium.
Report of the National Institute of Public Health and Environmental Hygiene 228471007, Bilthoven, The Netherlands.
Barkman, J.J., 1958. Phytosociology and Ecology of Cryptogamic
Epiphytes. Van Gorcum, Assen.
Bates, J.W., Mcnee, P.J., Mcleod, A.R., 1996. E€ects of sulphur
dioxide and ozone on lichen colonization of conifers in the Liphook
forest fumigation project. New Phytologist 132, 653±660.
Bode, P., De Bruin, M., 1990. Routine neutron activation analysis of
environmental samples. In: Lieth, H., Markert, B. (Eds.), Element
Concentration Kadasters in Ecosystems. VCH Verlagsgesellschaft,
Weinheim, pp. 171±178.
Brown, D.H., Beckett, R.P., 1984. Uptake and e€ect of cations on
lichen metabolism. Lichenologist 16, 173±188.
Cislaghi, C., Nimis, P.L., 1997. Lichens, air pollution and lung cancer.
Nature 387, 463±464.
De Wit, A., 1976. Epiphytic lichens and air pollution in The Netherlands. Bibliotheca Lichenologica 5, Cramer, Vaduz.
Folkesson, L., Andersson-Bringmark, E., 1988. Impoverishment of
vegetation in a coniferous forest polluted by copper and zinc.
Canadian Journal of Botany 66, 417±428.
Garty, J., Karary, Y., Harel, J., 1993. The impact of air pollution on
the integrity of cell membranes and chlorophyll in the lichen
Ramalina duriaei (Denot) Bagl. transplanted to industrial sites in
Israel. Archives of Environmental Contamination and Toxicology
24, 455±460.
Gebel, T., 1997. Arsenic and antimony: comparative approach on
mechanistic toxicology. Chemico-biological Interactions 107, 131±
144.
Herzig, R., Liebendorfer, L., Urech, M., Amman, K., Cuecheva, M.,
Landolt, W., 1989. Passive biomonitoring with lichens as a part of
an integrated biological measuring system for monitoring air pollution in Switzerland. International Journal of Environmental Analytical Chemistry 35, 43±57.
Kirschbaum, U., Marx, A., Schiek, J.E., 1996. Beurteilung der lufthygienischen Situation Gieszens und Wetzlars mittels epiphytischer
Flechten (1995). Angewandte Botanik 70, 78±96.
Lambinon, J., Maquinay, A., Ramaut, R.L., 1964. La teneur en zinc
de quelques lichens des terrains calcaires Belges. Bulletin du Jardin
Botanique de l'Etat 34, 273±282.
LeBlanc, F., De Sloover, J., 1970. Relation between industrialization
and the distribution and growth of epiphytic lichens and mosses in
Montreal. Canadian Journal of Botany 48, 1485±1496.
Lippo, H., Poikolainen, J., Kubin, E., 1995. The use of moss, lichen and
pine bark in the nationwide monitoring of atmospheric heavy metal
deposition in Finland. Water, Air and Soil Pollution 85, 2241±2246.
169
Loppi, S., Bargagli, R., 1996. Lichen biomonitoring of trace elements
in a geothermal area (Central Italy). Water, Air and Soil Pollution
86, 177±187.
Nash, T.H., 1971. Lichen sensitivity to hydrogen ¯uoride. Bulletin of
the Torrey Botanical Club 98, 103±106.
Nash, T.H., 1976. Sensitivity of lichens to nitrogen dioxide fumigations. Bryologist 79, 103±106.
Nylander, W., 1866. Les lichens du Jardin du Luxemboxurg. Bulletin
de la Societe Botanique FrancËaise 13, 364±372.
Payne, R.W., Lane, P.W., Todd, A.D., Digby, P.G.N., Thompson, R.,
Harding, S.A., Tuncli€e Wilson, G., Leech, P.K., Welham, S.J.,
Morgan, G.W., White, R.P., 1993. GENSTAT 5 release 3 Reference
Manual. Clarendon Press, Oxford.
Purvis, O.W., 1984. The occurrence of copper oxalate in lichens
growing on copper sulphide-bearing rocks in Scandinavia. Lichenologist 16, 197±204.
Purvis, O.W., 1996. Interactions of lichens with metals. Science Progess 79, 283±309.
Purvis, O.W., Halls, C., 1996. A review of lichens in metal-enriched
environments. Lichenologist 28, 571±601.
Seaward, M.R.D., 1992. Large-scale air pollution monitoring using
lichens. GeoJournal 28, 403±411.
Shimwell, D.W., Laurie, A.E., 1972. Lead and zinc contamination of
vegetation in the Southern Pennines. Environmental Pollution 3,
291±301.
Sigal, L.L., Nash, T.H., 1983. Lichen communities on conifers in
southern California mountains: an ecological survey relative to oxidant air pollution. Ecology 64, 1343±1354.
Skye, E., 1968. Lichens and air pollution: a study of cryptogamic epiphytes and enviroment in the Stockholm region. Acta Phytogeographica Suecica 52, 1±123.
Sloof, J.E., 1995. Lichens as quantitative biomonitors for atmospheric
trace-element deposition, using transplants. Atmospheric Environment 29, 11±20.
Ter Braak, C.J.F., 1994. Canonical community ordination Part I:
basic theory and linear methods. Ecoscience 1, 127±140.
Ter Braak, C.J.F., 1995. Ordination. In: Jongman, R.H.G., ter Braak,
C.J.F., van Tongeren, O.F.R. (Eds.), Data Analysis in Community
and Landscape Ecology. Cambridge University Press, Cambridge,
pp. 91±173.
Ter Braak, C.J.F., Smilauer, P., 1998. CANOCO Reference Manual
and User's Guide to Canoco for Windows: Software for Canonical
Community Ordination (version 4). Microcomputer Power, Ithaca
NY.
Van Dobben, H.F., 1993. Vegetation as a monitor for deposition
of nitrogen and acidity. PhD thesis, University of Utrecht, The
Netherlands.
Van Dobben, H.F., De Bakker, A.J., 1996. Re-mapping epiphytic
lichen biodiversity in the Netherlands: e€ects of decreasing SO2 and
increasing NH3. Acta Botanica Neerlandica 45, 55±71.
Van Dobben, H.F., Ter Braak, C.J.F., 1999. Ranking of epiphytic
lichen sensitivity to air pollution using survey data: a comparison of
indicator scales. Lichenologist 31, 27±39.
Van Egmond, N.D., Tissing, O., Onderdelinden, D., Bartels, C., 1978.
Quantitative evaluation of mesoscale air pollution transport.
Atmospheric Environment 12, 2279±2287.
Weast, R.C., Astle, M.J., Beyer, W.H., 1987. CRC Handbook of
Chemistry and Physics. CRC Press, Boca Raton, FL, USA.
Wirth, V., 1993. Trendwende bei der Ausbreitung der anthropogen
gefoÈrterten Flechte Lecanora conizaeoides? Phytocoenologia 23,
625±636.
Wolterbeek, H.Th., Kuik, P., Verburg, T.G., Wamelink, W.W., Van
Dobben, H., 1996. Relations between sulphate, ammonia, nitrate,
acidity and trace element concentrations in tree bark in The Netherlands. Environmental Monitoring and Assessment 40, 185±201.