Ecological Indicators 115 (2020) 106355

Contents lists available at ScienceDirect

Ecological Indicators
journal homepage: www.elsevier.com/locate/ecolind

Evaluation of the Index of Atmospheric Purity in an American tropical valley T

through the sampling of corticulous lichens in different phorophyte species
Mauricio Andres Correa-Ochoaa, , Leidy Catalina Vélez-Monsalvea,
⁎

Julio César Saldarriaga-Molinaa, Margarita Maria Jaramillo-Cirob

a
Engineering and Environmental Management Research Group (Grupo de Ingeniería y Gestión Ambiental – GIGA), School of Engineering, Universidad de Antioquia
(University of Antioquia), Cl. 67 No. 53-108 Of. 20-250, Medellín, Colombia
b
Modeling and Computational Simulation Research Group (Grupo de Investigación en Modelamiento y Simulación Computacional – GIMSC), School of Engineering, San
Buenaventura University, Carrera 56C No. 51-110, Medellín, Colombia

ARTICLE INFO ABSTRACT

Keywords: Lichens have been widely used worldwide as bioindicators of air quality and also integrated into national
Bioindicator programs and protocols in several countries. The Index of Atmospheric Purity (IAP) has been a commonly used
Air quality tool for the diagnosis of air quality using corticulous lichens. However, its implementation is recommended
Phorophyte diversity when phorophytes with similar characteristics of the bark are used. Therefore, many studies are restricted to the
Corticulous lichens
use of a single supporting phorophyte. This criterion makes it difficult to implement this index in areas with
limited number of individuals and species of phorophytes. This research evaluates the functionality of this index
without differentiating among carrier phorophytes for the classification of areas with differences in air quality.
For this, the lichens were mapped and collected, along with bark samples to determine the pH bark of phor-
ophytes close to air quality monitoring stations located in Medellín City.
A total of 148 lichen species were identified, belonging to 26 families and 52 genera. In addition, indicators
such as thallus heterogeneity, richness (with correlations coefficients r = 0.52 with no phorophyte differ-
entiation and for the tree species F. uhdei and T. rosea, r = 0.578 and r = 0.777; respectively) and total coverage
of corticulous lichens (r = 0.76) vary in response to air pollution levels, no matter if their evaluation is carried
out on a single phorophyte or in a diversity of phorophytes.
This research also revealed that the evaluation of IAP in the biomonitoring zones is functional, resulting in
positive correlations between the IAP and the air quality classification when this index is evaluated without
differentiation of phorophytes (r = 0.78) and for T. rosea (r = 0.94) and F. uhdei (r = 0.99). In this sense, the
uniformity criterion in the selection of the phorophyte use for the application of this index can be complemented
taking into account the possibility of using a heterogeneous sample of phorophytes.

1. Introduction
Lichens are considered air pollution indicators, due to their anato-
mical, morphological, and physiological characteristics (Cleavitt et al.,
2015; McCune, 2000; Will-Wolf et al., 2015). These organisms obtain
nutrients directly from the atmosphere (Conti and Cecchetti, 2001),
since the absence of cuticle facilitates the absorption of elements pre-
sent in the air (lichens have no selectivity). In addition, given their
bioaccumulative nature (no elimination mechanisms), these organisms
incorporate elements into their intracellular sites (Falla et al., 2000).
These characteristics make lichens susceptible to the presence of toxic
or harmful substances (air pollutants).
Lichens are increasingly used as biomonitors, since they have ad-
vantages over conventional monitoring systems, which are expensive,
require maintenance, and their transportation can prevent their im-
plementation in difficult access areas (Agarwal, 2005). Additionally,
conventional systems are limited to certain chemical compounds or
pollutants and have no intrinsic relationship with the biological effect
of such substances (Agarwal, 2005; Rodrigo et al., 1999). In contrast,
lichens are widely distributed worldwide and may reflect the influence
of environmental factors on living beings (Barreno and Pérez-Ortega,
2003). Moreover, the implementation techniques for using lichens as
bioindicators are considered simple and not very expensive (Monge-
Nájera et al., 2002) and it is stated that lichens stablish limits difficult to

⁎
Corresponding author.
E-mail address: mandres.correa@udea.edu.co (M.A. Correa-Ochoa).

https://doi.org/10.1016/j.ecolind.2020.106355
Received 8 February 2019; Received in revised form 21 March 2020; Accepted 25 March 2020
Available online 10 April 2020
1470-160X/ © 2020 Elsevier Ltd. All rights reserved.
M.A. Correa-Ochoa, et al.
detect through conventional systems (Rubiano, 1987).
Lichens have been used with great success in biomonitoring pro-
grams throughout the world (Anze et al., 2007; Van Herk et al., 2002;
Monge-Nájera et al., 2002; USDA, 2008) and also, have been integrated
into national protocols in countries such as England, Germany, the
United States, and the Netherlands (Käffer et al., 2011). In this sense,
different methodologies have been used for the diagnosis of air quality
using lichens as bioindicators, such as Index of Atmospheric Purity
(IAP) proposed by LeBlanc and De Sloover (1970) which has been
widely used to make local assessments of air quality in countries such as
Germany, Italy, Spain, France, and Slovenia, among others (Bartholmeß
et al., 1994; Cepeda and García, 1998; Gombert et al., 2004, 2006;
Loppi et al., 1998; Monaci et al., 1997; Nimis et al., 2000; Poličnik
et al., 2008; Ulshöfer and Rosner, 2001).
In temperate countries, it has been well demonstrated that en-
vironmental factors and air pollution directly influence the diversity
and distribution of lichens (Hawksworth et al., 2005). However, in
tropical countries there is limited information regarding the diversity of
lichens and air quality (Attanayaka and Wijeyaratne, 2013). Despite
this, some studies in the tropics have applied methodologies to bioin-
dicate air quality through the use of lichens. In this regard, Vareschi and
Moreno (1973) were the first to use lichens as bioindicators in Vene-
zuela. The authors mapped the areas where lichens were present and
categorized these locations based on the presence or absence of the
organisms. Furthermore, Cohn-Berger and Quezada (2016) categorized
pollution levels in 32 locations within the metropolitan corridor of
Guatemala City, using the IAP, the coverage found for each species, and
the Environmental Classification Factor. In Colombia, Jaramillo and
Botero (2010) assessed the use of lichens as air pollution bioindicators
in two areas in the Aburrá Valley in Antioquia, with two different levels
of atmospheric pollution. Researchers proposed one tree species
(Fraxinus chinensis Roxb.) and two lichen species (Canoparmelia sp. and
Parmotrema austrosinensis (Zahlbr) Hale.) as appropriate species to carry
out studies on air quality, by using lichens as bioindicators in the
Aburra Valley area.
In some studies, the IAP is suggested as a good tool for the cate-
gorization of air pollution zones (Cohn-Berger and Quezada, 2016;
Jaramillo and Botero, 2010; Lijteroff et al., 2009). However, the use of
this index requires as a fundamental premise that the phorophyte spe-
cies have similar physical and chemical bark properties (LeBlanc and De
Sloover, 1970). Therefore, in many studies it has been chosen a single
supporting phorophyte common to all areas, where sampling is carried
out, in order to avoid substrate variables such as pH and roughness of
the phorophyte have an influence on the lichen growth (Lijteroff et al.,
2009).
In some scenarios, the use of a single carrier phorophyte may
hamper the implementation of lichens as bioindicators with IAP, since
cities are currently undergoing rapid spatial and demographic growth,
which entails logging, timber transportation, and tree planting (Alvarez
del Castillo and Agredo Cardona, 2013), causing tree species to be
different, both in taxa and in longevity. Additionally, in the tropics
there is a large diversity of tree species (FAO, 2018), so guaranteeing
uniformity of bearing phorophytes represents a difficult challenge. The
purpose of this research was to assess the behavior of the Index of At-
mospheric Purity (IAP) in five areas with pollution differences in Me-
dellín City, without differentiating among carrier phorophytes, to de-
termine the functionality of this index under different conditions of
phorophyte diversity.
2. Materials and methods
2.1. Study area
Medellín is located in the central mountain range of the Colombian
Andes at 1450 m above sea level, in a deep and narrow valley (1000 m
and 10 km on average), known as Aburra Valley (AV). Medellín city has
2
Ecological Indicators 115 (2020) 106355
Table 1
Location of air quality stations in the biomonitoring zones.
Biomonitoring Air Quality Location of air quality station Altitude
Zones Monitoring Coordinate System (WGS_1984) MASL
Station
N W
UNNV MED_UNNV 6°15′48.39″ 75°34′38.11″ 1450
MANT MED_MANT 6°15′9.75″ 75°34′10.71″ 1479
EXSA MED_EXSA 6°14′48.04″ 75°34′3.40″ 1480
PJIC MED_PJIC 6°12′32.24″ 75°34′39.84″ 1506
LAYE MED_LAYE 6°10′57.32″ 75°33′2.52″ 1950
a mild-dry climate, with annual average temperature values of 22.5 °C,
precipitation 1685 mm and relative humidity of 67% (IDEAM, 2018a).
The prevailing direction of the winds in the axis of the valley is given in
the north-south direction (Correa et al., 2009; IDEAM, 2018b). Medellín
has an urban woodland of around 500,000 trees in green public spaces,
distributed among about 550 species (del Medio and de Medellín,
2011). Most tree species are located in linear transects consisting of
platforms, road separators and corridors.
In recent years, pollution levels in the AV area have raised great
concern and have gained importance for different actors (AMVA, 2007).
The particulate matter classified as PM2.5 is the pollutant that most
deteriorates air quality in the metropolitan area of the Aburrá Valley
(AMVA, 2016). It has been confirmed that increased concentrations of
such pollutant mainly stem from mobile sources, contributing nearly
79% of this agent and its precursors (AMVA, 2016, AMVA, 2017).
2.2. Biomonitoring areas
For this study, five areas were strategically selected near air quality
monitoring stations in Medellín city (see Table 1 and Fig. 1), which are
currently operated by the Early Warning System (Sistema de Alerta
Temprana – SIATA). These stations register significant gradients in
terms of pollutant concentrations and physical and environmental
conditions. In addition, these locations represent the topographic and
altitudinal variability of the study area. The location of biomonitoring
stations is presented in Table 1.
MANT zone is located in the downtown area of the city. This area is
characterized by being a commercial and institutional sector with
emissions mainly related to mobile sources. It has little vegetation, with
trees located in isolation and that are part of the urban planning of the
area. For its part, EXSA is characterized by having a high vehicular flow
with roads where collective public, intermunicipal public service routes
and vehicles belonging to urban centers converge. It has scarce vege-
tation located mainly on the sides of the roads and road separators. In
general, this area presents similar conditions to those in MANT zone.
UNNV zone presents heterogeneous conditions by combining ex-
posure to high vehicular traffic and large vegetation cover. MED_UNNV
station is located inside the National University of Colombia that is
characterized by having great diversity and coverage of tree species. On
the other hand, PJIC zone is located in a highly urbanized sector with
mainly institutional and residential land uses. It has trees located in
isolation in the most residential areas and located on roads and road
separators. In contrast, LAYE area is located in a sector where vegeta-
tion predominates, as well as low emissions of industrial origin. Despite
being directly exposed to vehicular emissions from automobile traffic in
its access roads, it is an important contrast area since it is characterized
by mostly forest land use.
2.3. Air quality
Worldwide, particulate matter is considered an object of study be-
cause it has negative effects on human health (WHO, 2016). In this
sense, this pollutant has been widely used to establish zonal
M.A. Correa-Ochoa, et al.
Fig. 1. Location of air quality stations and the biomonitoring zones object of study. A. EXSA (R = 300 m), B. MANT (R = 300 m), C. UNNV (R = 500 m), D. PJIC
(R = 300 m), and E. LAYE (R = 500 m). R (radius around air quality station.
classifications of air quality and is considered as an objective for the
evaluation of air quality (WHO, 2013). In the same way this pollutant is
used in the AV for the classification of the contamination degree
(AMVA, 2016).
Based on studies in which concentrations of particulate matter have
been related to the composition of corticulous lichen communities in
different zones (Massimi et al., 2019; Varela et al., 2018), the study
areas were categorized using time series analysis of PM2.5 and PM10
recorded by air quality monitoring stations for the period 2013–2016
(AMVA, 2018). In this regard, the behavior of PM2.5 average monthly
concentrations reported for MED_MANT, MED_UNNV, MED_PJIC, and
MED_LAYE stations were analyzed. Since MED_EXSA station does not
measure this pollutant, then a representativeness coefficient was de-
termined (for each month), considering that MED_MANT and ME-
D_EXSA stations are located in zones that present similar conditions in
3
Ecological Indicators 115 (2020) 106355
terms of location, land use, emission sources, and vegetation presence.
This similarity was also verified by comparing the reports on the PM10
concentrations for the two stations during the period October
2015–December 2016, using the Wilcoxon test and the two-sample
Kolmogorov-Smirnov test.
The representativeness coefficient was calculated with Eq. (1):
C1 = PM 2. 51/ PM 101 2 = PM 2. 52 / PM 102 (1)
where: C1 symbolizes the representativeness coefficient of PM2.5 in
relation to PM10 concentrations for MED_MANT station and C2 the
representativeness coefficient of PM2.5 in relation to PM10 concentra-
tions for MED_EXSA station. Once monthly coefficients were obtained,
the values were averaged.
After establishing that MED_MANT and MED_EXSA were compar-
able, it was possible to affirm that C1 ≃ C2; therefore, the following
M.A. Correa-Ochoa, et al.
equation could be empirically established:
PM 2. 52 = PM 102 × C1
The result of Eq. (2) represents an estimation of PM2.5 concentra-
tions on the basis of PM10 concentrations reported in MED_EXSA station
by the representativeness coefficient of PM2.5 calculated for MED_-
MANT station. The behavior of the estimated PM2.5 concentrations for
MED_EXSA station was analyzed together with PM2.5 average monthly
concentrations reported for the other stations.
2.4. Sampling
In each of the five biomonitoring areas, 28 to 30 phorophytes were
sampled (without species differentiation) in a radius (R) no lager than
600 m around the station (See Fig. 1). For the selection of the phor-
ophytes, Barreno and Pérez-Ortega (2003) methodology was im-
plemented, in which individuals must have a diameter at breast level
above 20 cm, and a trunk inclination angle less than 20° with respect to
the vertical, they cannot belong to closed forest formations, and should
have tree morphological characteristics that correspond to healthy in-
dividuals with no resprouting at the trunk base.
Each selected phorophyte was mapped using a 100 cm × 50 cm
acetate sheet with the lower edge placed 50 cm from the ground, where
the contour of each lichen species was drawn using markers of different
colors depending on morphotype (Fig. 2).
To identify lichen morphotypes, samples of the mapped species
were collected following the methodology proposed by Chaparro and
Aguirre (2002). During collection, the healthiest specimens were se-
lected, with evidence of propagules or reproductive structures. These
samples were stored separately in paper bags to prevent rotting and
labeled with their corresponding station code, phorophyte number, and
morphotype mapping color. The obtained samples were identified with
different taxonomic keys. Lichen collections were preserved at the
Herbarium of Universidad de Antioquia (HUA).
For measuring bark pH, pieces as thin are possible were removed
from the phorophyte surface using a knife (Kricke, 2002). These sample
were stored in paper bags.
Determination of lichens cover and richness
The acetate films obtained were digitized on a 1:1 scale in the
presence of a metric pattern, and stored in a TIF format without com-
pression. Lichens coverage was quantified using a code developed for
digital image analysis programmed with MATLAB 2016a. This code
individually isolates each closed perimeter curve that represents the
outline of a lichen, calculates the area inside, and grouping the areas
according to the color code Fig. 2. Also, the number of lichens (rich-
ness) was calculated as the number of different color codes.
Fig. 2. Mapping of lichen species in each phorophyte, identification of color
patterns and digital quantification of lichens coverage.
Ecological Indicators 115 (2020) 106355
2.5. Measurement of bark pH
(2) The quantification of bark pH was carried out following the meth-
odology proposed by Kricke (2002). The collected samples were ma-
cerated until a fine powder was obtained. One gram of the sample was
weighed and placed in test tubes with 10 mL of distilled water. Each
test tube was shaken to obtain a homogeneous mixture. After four
hours, the mixture was filter, and the pH was measured with a multi-
parameter meter (ref. HACH-HQ40d).
2.6. Index of Atmospheric Purity
To associate the cover and richness values of lichen species found in
each station with pollution levels, the Index of Atmospheric Purity was
determined, as proposed by LeBlanc and De Sloover (1970) and mod-
ified by Rubiano (2002), according to the following Eq. (3)
IAPj = ( i = 1: n Qi × fi × Ci )/n (3)
where IAPj is the Index of Atmospheric Purity for j zone; n is the
number of sampling units in j zone; and Ci is the relative cover of the
species in j zone, expressed as the ratio between the value for total
species cover in j zone and the maximum cover value for the i species
reached in any zone. Additionally, fi is the frequency of i species
(number of phorophytes in j zone where i is present), and Q is the re-
sistance factor of the species expressed by means of Eq. (4), where Aj is
the number of species present in each zone (richness) where i is found,
and Ej is the number of zones where i is found.
Qi = i = 1: n (Aj 1)/ Ej (4)
In order to determine the behavior of IAP, when this is calculated
using different phorophyte species, ten random combinations of eight
phorophytes were generated for each zone and the IAP was calculated
for different combinations, obtaining ten different IAP calculations in
each area. Since the same species of phorophytes were not found in all
areas, the IAP was calculated for two of the common species that were
present in at least three of the five zones under study.
2.7. Data analysis
Correlations between the zone classification, richness, cover lichens
and the IAP calculated for the combinations with different phorophyte
species and for two common phorophyte species were tested with
Spearman’s rank correlation coefficient. On the other hand, Kruskal
Wallis test was performed in order to determine differences between pH
bark (of the different phorophyte species) and Q factor, in the zone
classification.
A cluster analysis was applied in order to group the IAP calculated
for the different combinations of phorophyte species and for two
common species and the zone classification.
3. Results and discussion
3.1. Classification of biomonitoring areas based on air quality
Fig. 3 shows the behavior of the monthly average concentrations of
PM10 and PM2.5 for MED_MANT station and PM10 for MED_EXSA sta-
tion during the period October 2015–December 2016. The Wilcoxon
test (p = 0.222 > 0.05) determined that no statistically significant
differences exist between the reports of the two stations. Furthermore,
the Kolmogorov-Smirnov test comparing sample distributions de-
termined that concentrations in both locations have the same dis-
tribution (p = 0.925 > 0.05). Thus, the particulate matter con-
centrations in both stations are established as comparable. Therefore,
Eq. (2) can be used to establish an average representativeness coeffi-
cient of PM2.5 in relation to PM10 concentrations for the MED_EXSA
station, which in this case corresponds to 0.61.
4
M.A. Correa-Ochoa, et al. Ecological Indicators 115 (2020) 106355

Fig. 4 shows the behavior of PM2.5 monthly concentrations reported

Fig. 3. PM10 and PM2.5 concentrations for MED_MANT and MED_EXSA sta-
tions during the period October 2015 to December 2016.
by the Air Quality Monitoring Network in MED_MANT, MED_UNNV,
MED_PJIC, and MED_LAYE stations and the estimates for MED_EXSA. A
significant increase is noted in March 2016, when an environmental
contingency was declared in the city of Medellín. In 2015, some con-
centration values were not reported for MED_MANT station. This
omission occurred because it was not possible to collect 75% of the
valid data necessary for the report. In relation to MED_PJIC station,
PM2.5 began to be monitored with an automated equipment on May 5,
2015. As such, it was not possible to validate the automated equipment
series with the series obtained with a manual equipment for January
and February. The summary reported for 2015 corresponds to the re-
sults obtained with the automated equipment. In addition, a report for
MED_LAYE station is only available beginning in October 2015, since
this station was added to the network on September 16 of that year.
MED_EXSA Station reported the highest concentrations, exceeding
more than 51% of the time the concentrations reported for MED_MANT
station, which exceeded 100% of the times the records for MED_UNNV
station. Finally, MED_UNNV concentrations were higher than
MED_PJIC concentrations 97% of the time.
Comparisons between reports for MED_PJIC and MED_LAYE stations
could not be carried out because they did not have any reporting per-
iods in common. However, considering the environmental conditions,
land uses, and zone types it is possible to infer that PJIC zone has higher
polluting emissions than LAYE.

Fig. 4. PM2.5 average monthly concentrations in the air quality stations located in the biomonitoring zones. Monitoring period January 2013–December 2016 A.
2013, B. 2014, C. 2015, D. 2016.

5
M.A. Correa-Ochoa, et al. Ecological Indicators 115 (2020) 106355

Table 2
Air quality classification of biomonitoring zones.
Zone Air quality classification

EXSA 1
MANT 2
UNNV 3
PJIC 4
LAYE 5

Taking into account the previous analysis, the biomonitoring zones

can be classified from 1 to 5, 1 being the most polluted station and 5 the
least polluted station Table 2 presents the classification corresponding
to each zone:

3.2. Tree species

In this study, 21 tree species randomly distributed in the biomoni-

toring areas were sampled, totaling 147 phorophytes; the different
phorophyte species found in the study areas are presented in Appendix
A.

3.3. Lichen cover, richness, and diversity

In this research, a total of 148 lichen species were identified

(Appendix B), belonging to 26 families, of which Parmeliaceae was the
most represented family with 37 species, followed by Physciaceae with
30. The great richness of these families lends support to the results
obtained by authors such as Käffer et al. (2011), who found dominance
of the same families in similar studies. According to Divakar et al.
(2006) and Nimis et al. (1990), the largest lichen family is Parmeliaceae,
which includes taxa that are frequently used as bioindicators of atmo-
spheric pollution, additionally showing a wide habitat range and geo-
graphical distribution. In addition, the studies conducted by Rindita
et al. (2015) show that Physciaceae family is characterized by abun-
dance in urban areas. Finally, Chaparro and Aguirre (2002) indicate
that both families are the most representative families in terms of genus
and species in Colombia. Fig. 5. Lichen cover and richness in relation to air quality classification given to
It was possible to identify 52 genera, of which Parmotrema was the each biomonitoring zone.
most representative genus with 18 species, followed by Lecanora and
Physcia with 12 each one. Regarding species, Phaeophyscia chloantha
general, richness and cover values increase as the air quality classifi-
(Ach.) Moberg was the most frequent, present on 50 of the 147 phor-
cation also increases (Fig. 5); in other words, when atmospheric pol-
ophytes assessed at four zones, followed by Candelaria concolor (Dicks.)
lution diminishes. In agreement with these finding Conti and Cecchetti
Arnold, present on 45 phorophytes at all areas.
(2001) affirm that changes in the composition of lichens communities
Of the total species found, 82 were exclusive LAYE. This finding can
are related with air pollution levels, with a negative correlation be-
be explained as a result of the species adaptation to a better air quality.
tween lichens diversity and air pollutants concentration.
Eight lichen species were exclusive to UNNV, three to PJIC, and four to
The cover, richness, and diversity values for UNNV do not agree
EXSA. Regarding morphological groups, 37.16% belonged to crustose
with how the zone was classified in terms of air quality. This difference
species, 51.35% to foliose, 4.05% to gelatinous, 2.03% to fruticose, and
may be because this zone is located within an area of big vegetation
1.35% to squamulose.
cover, as described in Table 1, which may contribute to the develop-
LAYE zone had the largest number of species (118) followed by
ment of suitable conditions for establishing lichen communities.
UNNV with 44. In contrast, MANT zone had the lowest number of
Barreno and Pérez-Ortega (2003) argue that stand structure is also a
species, 13. In terms of lichen cover per mapped unit, LAYE had the
factor that should be considered, in relation to the different types of
highest cover with 0.253 cm2/cm2 of mapped phorophyte, followed by
microenvironments that can be created and immediately detected by
UNNV with 0.252 cm2/cm2 of mapped phorophyte. Additionally, EXSA
lichens.
presented the smallest cover with 0.054 cm2/cm2 of mapped phor-
In general, the highest number of thallus types in the biomonitoring
ophyte (see Fig. 5). These findings are in agreement with similar ob-
zones (except for LAYE) corresponded to foliose species; according to
servations to those made by Conti and Cecchetti (2001), who reported
Lücking (1997), these types of lichens are usually sensitive indicators of
correlations in lichen community compositions and atmospheric pol-
pollution levels, since they do not attach closely to substrate and are
lution levels. This result is further confirmed by Käffer et al. (2011),
therefore highly dependent on the air. However, Nimis et al. (1990)
who found that environments with limited human intervention present
affirm that, in some species, sensitivity does not depend on growth
a higher lichen diversity than highly urbanized areas.
forms but on the buffering capacity of the thallus, the anatomy, the
The correlation coefficients (r) between zone classification and total
water retention capacity, and their detoxification mechanisms and that
richness (r = 0.83) and cover values per mapped area (r = 0.76) in-
this sensitivity can be altered by different environmental conditions and
dicate that a strong correlation exists among these variables.
distribution areas. On the other hand, the number of crustose species
Additionally, positive correlations were found, indicating that, in

6
M.A. Correa-Ochoa, et al.
Fig. 6. Number of species by thallus type sampled in the five Biomonitoring
zones.
increased as air quality improved, in this sense, the highest percentage
of crustose lichen was found at LAYE; the analysis of the crustose
species found at this particular zone revealed that they had developed
reproductive structures. According to Jaramillo and Botero (2010),
reproductive structures do not develop in environments where lichens
are exposed to pollution and the presence of these structures has only
been reported for relatively undisturbed environments, reason for
which the presence of reproductive structures is considered as an in-
dicator of good air quality. Additionally, the presence of fruticose li-
chens (Usnea and Ramalina) at LAYE (see Fig. 6 and Appendix B) agrees
with the findings of Perlmutter et al. (2017), who noted that the pre-
sence of fruticose lichens was infrequent in areas exposed to high
concentrations of air pollutants exposed to high concentrations of air
pollutants, due to the fact that fruticose lichens are considered “sensi-
tive” to disturbances and atmospheric pollution (Cleavitt et al., 2009;
Perlmutter et al., 2017).
Along the same lines, the largest diversity of thallus types were
found in LAYE (6 types), followed by UNNV (5 types); while MANT
exhibited the lowest diversity (1 type) (Fig. 6). These results show
differences in lichen morphological composition for each area and
confirms the fact that the lichen morphology observed in different areas
changes according to environmental characteristics (pollution levels,
microenvironment, macroenvironment, etc.); in this regard, areas with
Fig. 7. Lichen richness with no phorophyte differentiation, for two phorophyte species. A. No phorophyte differentiation, B. F. uhdei, and C. T. rosea.
7
Ecological Indicators 115 (2020) 106355
higher thallus heterogeneity were associated with lower pollution le-
vels.
Richness in terms of air quality and phorophyte type
The correlation coefficient between the zone classifications and
richness values with no phorophyte differentiation (r = 0.52) and for
the tree species F. uhdei (r = 0.578) and T. rosea (r = 0.777) common
in at least three zones, demonstrated a positive correlation higher than
0.5, thus indicating that, regardless of phorophyte species, richness
increases as air quality improves Fig. 7. This finding suggests that there
exists a wide range of adaptive responses of lichens to different sub-
strate types and that air quality is a constraining factor for the host tree
species selected for studying lichens (Jaramillo and Botero, 2010) and
are in agreement with Cáceres et al. (2007) concluded that lichen
community formation is weakly related to factors, such as substrate
texture, phorophyte bark pH, and microclimate but is strongly related
to stochastic effects of dispersal species. In this regard, similar results
were also obtained by Soto et al. (2012).
3.4. Bark pH
Bark pH for phorophytes ranged between 4.0 and 8.3 pH units at the
different zones. This parameter varies considerably for a single species,
even more than one unit (Fig. 8). At the zone level, the pH values ob-
tained for the different phorophytes varied in the following intervals:
4.4 to 7.4 at EXSA, 5.3 to 7.3 at MANT, 4.0 to 6.7 at UNNV, 4.6 to 8.3 at
MED_PJIC, and 4.9 to 6.9 at LAYE.
In order to establish if there are differences in bark pH of the same
phorophytes species in the biomonitoring areas, a Kruskall Wallis test
was carried out for the most common tree species in the biomonitoring
areas (present at least in three of the five zones), yielding the following
values: F. udhei (p = 0.4610), T. rosea (p = 0.0004), S. campanulata
(p = 0.2061), and M. indica (p = 0.3503). The p values assessed for the
different tree species determined that, in most species, no statistically
significant differences were noted in terms of the pH values among
areas, with the exception of the species T. rosea, where the p value was
less than 0.05. Fig. 9 shows that pH values vary widely for each species.
Previous research studies, such as those conducted by Bates (1992)
and Pereira et al. (2014) note that environmental factors and anthro-
pogenic disturbance may drastically change bark pH in phorophytes.
Furthermore, they say that pH fluctuations do not remain constant but
instead changes over time with phorophytes age.
3.5. IAP
The IAP calculated from the results obtained for the combinations
with no differentiation of phorophyte species shows a similar behavior
with the IAP calculated for the two most common species (F. uhdei and
T. rosea) among the zones (Fig. 10). Positive correlations were obtained
M.A. Correa-Ochoa, et al. Ecological Indicators 115 (2020) 106355

Fig. 8. Variation of bark pH in different phorophytes species sampled in the five
biomonitoring zones.
between the IAP and the air quality classification in each area, either
data set with no phorophyte differentiation in the different cycles
(r = 0.94) or the data obtained for tree species F. uhdei (r = 0.99) and
T. rosea (r = 0.98). In all cases, correlation coefficients higher than 0.9
were obtained. This finding indicates a strong relationship between the
variables.
The strong relationship between the air quality classification for
each zone and the IAP values calculated from the result obtained for the
random combinations with no phorophyte differentiation, as well as
differentiating between F. uhdei and T. rosea (Fig. 10), and their similar
behavior indicates that the IAP for Medellín may be assessed using
phorophytes of different taxa.
Cluster analysis (Fig. 11) indicates that there are 3 kind of groups
according to the calculated IAP for each zone. Clusters C1 and C2 are
similar when the distance increases, that is to say, these two groups
form the same cluster. On the other hand, the group C3 is clearly dif-
ferent from the other groups, as well as, that its conglomerate does not
group with the others.
Following with the analysis, the IAP calculated from different
combinations (see Fig. 12), in all areas, it was found that no matter the
combination of tree species, the IAP values show a similar behavior in
the different zone classifications. A positive and strong relationship was
found between zone classification and the IAP data set in all combi-
nations, the correlation coefficient r = 0.74, indicate that IAP increases
when the air quality improves. This finding indicates that the adjust-
ment suggested for this research related to the modification of the
methodology for the selection of the phorophytes still provides

Fig. 9. pH variation in the different tree species and for each monitoring station F. uhdei, b) T. rosea, c) S. campanulata, d) M. indica.

Fig. 10. IAP calculation for different stations (with different combination without differentiating phorophyte species and for two common phorophyte species): a) No
phorophyte differentiation, b) F. uhdei, and T. rosea.

8
M.A. Correa-Ochoa, et al.
Fig. 11. IAP Cluster for different stations (with different combination without
differentiating phorophyte species and for two common phorophyte species)
(ZC: Zone Clasification, Fud: F. uhdei; Tro: T. rosea).
functionality to indicate the pollution levels among areas.
Fig. 13 shows resistance factor values (Q) for the different lichen
species. The corresponding factor related to the species were grouped,
considering the presence of each species in the areas. Through the
Kruskal-Wallis test, it was determined that no statistically significant
differences exist (p = 0.00 < 0.05) between the distribution medians
of factor Q in the biomonitoring zones. As shown in Fig. 13, the paired
comparison shows that stations MANT and EXSA do not show
Fig. 12. IAP calculation for the different combinations without differentiating
phorophyte species in the five study areas.
9
Ecological Indicators 115 (2020) 106355
Fig. 13. Resistance factors for lichen species in relation to the biomonitoring
areas.
significant differences. The same applies for PJIC and UNNV.
The IAP value calculated for UNNV, again, did not match with the
air quality classification assigned to this zone (Fig. 12). However, the
resistance factors estimated for lichen species found in that area are
similar to those found for PJIC. This finding indicates that the species
found in UNNV may be considered resistant. Additionally, the re-
sistance factors values found in EXSA and MANT zones suggest that
species (mostly foliaceous thallus) found in this zone are resistance as
well. The opposite result occurred with the species found in LAYE,
which showed the highest resistance factors, thus indicating that the
species in this area are more sensitive in comparison with those found
in the other studied areas (see Fig. 13).
4. Conclusions
In cities like Medellín, with marked gradients (> 10 µg/m3) or
differences in the concentrations of air pollutants among the different
zones and great diversity of tree species, the IAP implementation re-
presents a difficult challenge. Therefore, the adjustment suggested by
this study related to the modification of the methodology for the se-
lection of the supporting phorophytes, is an important contribution for
the application of this index in highly urbanized cities in which there is
a limited number of individuals and species of phorophytes.
In this sense, the uniformity criterion in the selection of the phro-
phyte used for the application of the IAP can be complemented taking
into account that the possibility of using a heterogeneous sample of the
Phorophytes is included, since the results with both criteria are com-
parable or without significant differences. The results showed correla-
tion values between the IAP and atmospheric pollution levels above 0.9,
both when making the differentiation of phorophytes in the different
cycles (r = 0.,94) and considering the species F. uhdei (r = 0.98) and T.
rosea (r = 0.99) in the study. That indicates that IAP for Medellín may
be assessed using phorophytes of different taxa.
Indicators such as thallus heterogeneity, richness (with correlations
coefficients r = 0.52 with no phorophyte differentiation and for the
tree species F. uhdei and T. rosea, r = 0.578 and r = 0.777; respec-
tively) and total coverage of corticulous lichens (r = 0.76) vary in re-
sponse to air pollution levels, no matter if their evaluation is carried out
on a single phorophyte or in a diversity of phorophytes. These results
coincide with the scientific literature on this subject, given that a po-
sitive relationship was found among areas with better air quality and
increased richness and lichens coverage; while an opposite condition is
M.A. Correa-Ochoa, et al.
observed in areas with high levels of contamination. This could be
evidenced indifferently by using a single phorophyte or diversity of
these in the areas of biomonitoring.
In this study, it was found that the pH of the supporting substrate
presents significant variations (> at a unit of pH) for the same phor-
ophyte species among the biomonitoring areas. Therefore, maintaining
the criterion of a single tree species, as an object of study, does not
guarantee homogeneity in the properties of the substrate.
Special conditions found in some biomonitoring areas (UNNV)
highlight the importance of taking into account the interaction lichen
has with other factors different from pollution, such as the structure of
the forest mass, since this contributes to the establishment and adap-
tation of lichen species even with high presence of air pollutants.
This research allowed the identification of dominant species under
conditions of environmental stress that can be classified as resistant
species (Phaeophyscia chloantha (Ach.) Moberg and Candelaria concolor
(Dicks.) Arnold). Likewise, species sensitive to environmental dis-
turbances (Candelaria fibrosa (Fr.) Müll. Arg) were identified and are
recognized as bioindicators of good air quality under conditions of the
American tropics.
It would be important to validate or use the methodology for the
selection of phorophytes described in this study in other latitudes than
the tropics.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://
doi.org/10.1016/j.ecolind.2020.106355.
References
Agarwal, S.K., 2005. Environmental Monitoring. Kul Bhushan Nangia. Retrieved from
https://books.google.es/books?hl=es&lr=&id=zz-InaYHh2cC&oi=fnd&pg=PR5&
dq=Environmental+monitoring+SK+Agarwal&ots=C6CGaVcacc&sig=
KmGdNd0IK4GrXuWn8GuLDIZey3k#v=onepage&q=Environmental monitoring SK
Agarwal&f=false.
Alvarez del Castillo, J., Agredo Cardona, G.A., 2013. Pérdida de la cobertura vegetal y de
oxígeno en la media Montaña Del Trópico Andino, Caso Cuenca Urbana San Luis
(Manizales). Luna Azul 37, 30–48 Retrieved from http://www.scielo.org.co/pdf/
luaz/n37/n37a04.pdf.
AMVA. (2007). CONTAMINACIÓN ATMOSFÉRICA Y EFECTOS SOBRE LA SALUD DE LA
POBLACIÓN: Medellín y su área metropolitana (Estado del arte). Retrieved from
https://www.medellin.gov.co/irj/go/km/docs/wpccontent/Sites/Subportal del
Ciudadano/Salud/Secciones/Publicaciones/Documentos/2012/Investigaciones/
Contaminación atmosférica y efectos sobre la salud de la población.pdf.
AMVA. Acuerdo Metropolitano N 15 (2016).
AMVA. (2017). Inventario de atmosféricas del Valle de Aburrá, actualización 2015.
AMVA. (2018). Herramientas de Gestión/Resúmenes de Calidad del Aire. Retrieved from
http://www.metropol.gov.co/CalidadAire/Paginas/wz_resumen_aire.aspx.
Anze, R., Franken, M., Zaballa, M., Pinto, M.R., Zeballos, G., Cuadros, M.D.L.Á., Estellano,
V.H., 2007. Bioindicadores en la detección de la contaminación atmosférica en
Bolivia. Revista Virtual Redesma 53–74.
Attanayaka, A.N.P.M., Wijeyaratne, S.C., 2013. Corticolous lichen diversity, a potential
indicator for monitoring air pollution in tropics. J. Natl. Sci. Found. Sri Lanka 41 (2),
131–140. https://doi.org/10.4038/jnsfsr.v41i2.5707.
Barreno, E., Pérez-Ortega, S., 2003. Líquenes de la Reserva Natural Integral de Muniellos,
Asturias. Principado de Asturias. KRK ediciones https://doi.org/10.15713/ins.mmj.3.
Bartholmeß, H., Boemer, R., Roessler, G., Schober, G., Nobel, W., Arndt, U., 1994.
Flechten als bioindikatoren zur erfassung von immissionswirkungen. Environ. Sci.
Eur. 6–81. https://doi.org/10.1007/BF02937739.
Bates, J.W., 1992. Influence of chemical and physical factors on quercus and fraxinus
epiphytes at loch sunart, western scotland – a multivariate-analysis. J. Ecol. 80 (1),
163–179. https://doi.org/10.2307/2261073.
Cáceres, M.E.S., Lücking, R., Rambold, G., 2007. Phorophyte specificity and environ-
mental parameters versus stochasticity as determinants for species composition of
corticolous crustose lichen communities in the Atlantic rain forest of northeastern
Brazil. Mycol. Progress 6 (3), 117–136. https://doi.org/10.1007/s11557-007-0532-2.
Cepeda, J.M., García, J., 1998. The effect of air pollution from nitrogen dioxide (NO2) on
epiphytic lichens in Seville, Spain. Aerobiologiia 14, 241–247.
Chaparro, M., Aguirre, J., 2002. Hongos Liquenizados.
Cleavitt, N.L., Dibble, A.C., Werier, D.A., 2009. Influence of tree composition upon epi-
phytic macrolichens and bryophytes in old forests of Acadia National Park, Maine.
Bryologist 112 (3), 467–487. https://doi.org/10.1639/0007-2745-112.3.467.
Cleavitt, N.L., Hinds, J.W., Poirot, R.L., Geiser, L.H., Dibble, A.C., Leon, B., Pardo, L.H.,
2015. Epiphytic macrolichen communities correspond to patterns of sulfur and ni-
trogen deposition in the northeastern United States. Bryologist 118 (3), 304–324.
10
Ecological Indicators 115 (2020) 106355
https://doi.org/10.1639/0007-2745-118.3.304.
Cohn-Berger, G., Quezada, M., 2016. Líquenes como bioindicadores de contaminación
aérea en el corredor metropolitano de la ciudad de Guatemala. Revista Científica 26
(1), 20–39.
Conti, M.E., Cecchetti, G., 2001. Biological monitoring: lichens as bioindicators of air
pollution assessment — a review. Environ. Pollut. 114 (3), 471–492. https://doi.org/
10.1016/S0269-7491(00)00224-4.
Correa, M., Zuluaga, C., Palacio, C., Pérez, J., Jiménez, J., 2009. Surface wind coupling
from free atmosphere winds to local winds in a tropical region within complex ter-
rain. Case of study: Aburra Valley Antioquia, Colombia. Dyna 76 (158), 17–27
Retrieved from http://www.scopus.com/inward/record.url?eid=2-s2.0-
75249103588&partnerID=40&md5=e60a6899153e19cf150c93bce39d8f02.
Divakar, P.K., Crespo, A., Blanco, O., Lumbsch, H.T., 2006. Phylogenetic significance of
morphological characters in the tropical Hypotrachyna clade of parmelioid lichens
(Parmeliaceae, Ascomycota). Mol. Phylogenet. Evol. 40 (2), 448–458. https://doi.
org/10.1016/j.ympev.2006.03.024.
Falla, J., Laval-Gilly, P., Henryon, M., Morlot, D., Ferard, J.F., 2000. Biological air quality
monitoring: a review. Environ. Monit. Assess. 64 (3), 627–644. https://doi.org/10.
1023/A:1006385924945.
FAO. (2018). El estado de los bosques en el mundo. Retrieved from http://www.fao.org/
publications/es.
Gombert, S., Asta, J., Seaward, M.R.D., 2004. Assessment of lichen diversity by index of
atmospheric purity (IAP), index of human impact (IHI) and other environmental
factors in an urban area (Grenoble, southeast France). Sci. Total Environ. 324 (1–3),
183–199. https://doi.org/10.1016/j.scitotenv.2003.10.036.
Gombert, S., Asta, J., Seaward, M.R.D., 2006. Lichens and tobacco plants as com-
plementary biomonitors of air pollution in the Grenoble area (Isère, southeast
France). Ecol. Ind. 6 (2), 429–443. https://doi.org/10.1016/j.ecolind.2005.06.001.
Hawksworth, D.L., Iturriaga, T., Crespo, A., 2005. Líquenes como bioindicadores in-
mediatos de contaminación y cambios medio-ambientales en los trópicos. Rev.
Iberoam. Micol. 22 (2), 71–82. https://doi.org/10.1016/S1130-1406(05)70013-9.
Van Herk, C.M., Aptroot, A., Van Dobben, H.F., 2002. Long-term monitoring in the
Netherlands suggests that lichens respond to global warming. Lichenologist 34 (2),
141–154. https://doi.org/10.1006/lich.2002.0378.
IDEAM. (2018a). Atlas Climatológico de Colombia. Retrieved from http://atlas.ideam.
gov.co/visorAtlasClimatologico.html.
IDEAM, 2018b. Información Aeronáutica Climatología. Medellín, Régimen anual de
viento Retrieved from http://bart.ideam.gov.co/cliciu/rosas/viento.htm.
Jaramillo, M., Botero, L., 2010. Comunidades liquénicas como bioindicadores de calidad
del aire. Revista Gestión y Ambiente 13 (1), 97–110.
Käffer, M.I., Martins, S.M.D.A., Alves, C., Pereira, V.C., Fachel, J., Vargas, V.M.F., 2011.
Corticolous lichens as environmental indicators in urban areas in southern Brazil.
Ecol. Ind. 11 (5), 1319–1332. https://doi.org/10.1016/j.ecolind.2011.02.006.
Kricke, R., 2002. Monitoring with lichens–monitoring lichens. (P. Nimis, C. Scheidegger
and P. Wolseley, Eds.) (Vol 7). Wales, United Kingdom: The Netherlands: Kluwer
Academic Publishers. https://doi.org/10.1007/978-94-010-0423-7.
LeBlanc, S.C.F., De Sloover, J., 1970. Relation between industrialization and the dis-
tribution and growth of epiphytic lichens and mosses in Montreal. Can. J. Bot. 48 (8),
1485–1496. https://doi.org/10.1139/b70-224.
Lijteroff, R., Lima, L., Prieri, B., 2009. Uso de liquenes como bioindicadores de
contaminación atmosférica en la ciudad de San Luis Argentina. Rev. Int. Contam.
Ambiental 25 (2), 111–120.
Loppi, S., Nascimbene, J., 1998. Lichen bioindication of air quality in the Mt. Amiata
geothermal area (Tuscany, Italy). Geothermics, 27(98), 295–304.
Lücking, R., 1997. Abstracta Botanica. Akadémiai Kiadó 21 (1), 99–116.
Massimi, L., Conti, M.E., Mele, G., Ristorini, M., Astolfi, M.L., Canepari, S., 2019. Lichen
transplants as indicators of atmospheric element concentrations: a high spatial re-
solution comparison with PM 10 samples in a polluted area (Central Italy). Ecol. Ind.
101 (August 2018), 759–769. https://doi.org/10.1016/j.ecolind.2018.12.051.
McCune, B., 2000. Lichen communities as indicators of forest health. Bryologist 103 (2),
353–356. https://doi.org/10.1639/0007-2745(2000)103[0353:LCAIOF]2.0.CO;2.
Monaci, F., Bargagli, R., Gasparo, D., 1997. Air pollution monitoring by lichens in a small
medieval town of central Italy. Acta Botanica Neerlandica 46 (4), 403–412. https://
doi.org/10.1111/plb.1997.46.4.403.
Monge-Nájera, J., González, M. I., Rivas Rossi, M., Méndez, V.H., 2002. A new method to
assess air pollution using lichens as bioindicators. Revista de Biología Tropical, 50(1),
321–325. Retrieved from http://www.ncbi.nlm.nih.gov/pubmed/12298260.
Nimis, P.L., Castello, M., Perotti, M., 1990. Lichens as biomonitors of sulphur dioxide
pollution in la spezia (northern italy). Lichenologist 22 (3), 333–344. https://doi.org/
10.1017/S0024282990000378.
Nimis, P.L., Lazzarin, G., Lazzarin, A., Skert, N., 2000. Biomonitoring of trace elements
with lichens in Veneto (NE Italy). Sci. Total Environ. 255 (1–3), 97–111. https://doi.
org/10.1016/S0048-9697(00)00454-X.
Pereira, I., Müller, F., Moya, M., 2014. Influence of Nothofagus bark pH on the lichen and
bryophytes richness, Central Chile. Gayana. Botánica 71 (1), 120–130. https://doi.
org/10.4067/S0717-66432014000100012.
Perlmutter, G.B., Blank, G.B., Wentworth, T.R., Lowman, M.D., Neufeld, H.S., Plata, E.R.,
2017. Effects of highway pollution on forest lichen community structure in western
Wake County, North Carolina, U.S.A. Bryologist 120 (1), 1–10. https://doi.org/10.
1639/0007-2745-120.1.001.
Poličnik, H., Simončič, P., Batič, F., 2008. Monitoring air quality with lichens: a com-
parison between mapping in forest sites and in open areas. Environ. Pollut. 151 (2),
395–400. https://doi.org/10.1016/j.envpol.2007.06.003.
Rindita, Sudirman, L.I., Koesmaryono, 2015. Air quality bioindicator using the population
of epiphytic macrolichens in Bogor City, West Java. HAYATI J. Biosci., 22(2), 53–74.
https://doi.org/10.4308/hjb.22.2.53.
M.A. Correa-Ochoa, et al.
Rodrigo, A., Avila, A., Gómez-Bolea, A., 1999. Trace metal contents in Parmelia caperata
(L.) Ach. compared to bulk deposition, throughfall and leaf-wash fluxes in two holm
oak forests in Montseny (NE Spain). Atmos. Environ. 33 (3), 359–367. https://doi.
org/10.1016/S1352-2310(98)00167-8.
Rubiano, L.J., 1987. Delimitación de áreas de isocontaminación en Cali y Medellín uti-
lizando líquenes como bioindicadores. Perez-Arbelazia 2 (4), 7–41.
Rubiano, L.J., 2002. Monitoría de áreas de isocontaminación en la región de influencia de
la Central Termoelectrica Martin del Corral utilizando líquenes como bioindicadores.
Pérez-Arbelaiza, 13, 91–105. Retrieved from http://dx.doi.org/10.1016/j.asw.2013.
04.001%5Cnhttp://journals.cambridge.org/abstract_S0140525X00005756%5CnLib
scanned%5Cnhttp://www.br-ie.org/pub/index.php/rbie/article/view/
1293%5Cnhttp://www-psych.nmsu.edu/~pfoltz/reprints/Edmedia99.html
%5Cnhttp://urd.
Secretaría del Medio Ambiente de Medellín. (2011). Árboles nativos y ciudad. aportes de
la silvicultura urbana de medellín. Retrieved from https://www.medellin.gov.co/irj/
go/km/docs/wpccontent/Sites/Subportal del Ciudadano/Medio Ambiente/
Secciones/Publicaciones/Documentos/2012/Arboles Nativos y Ciudad – Libro.pdf.
Soto, E., Lücking, R., Bolaños, A., 2012. Especificidad de forófito y preferencias micro-
ambientales de los líquenes cortícolas en cinco forófitos del bosque premontano de
finca Zíngara, Cali, Colombia. Rev. Biol. Tropical 60 (2), 843–856. https://doi.org/
11
Ecological Indicators 115 (2020) 106355
10.15517/rbt.v60i2.4017.
Ulshöfer, J., Rosner, H.J., 2001. GIS-based analysis of lichen mappings and air pollution
in the area of Reutlingen (Baden-Württemberg, Germany). Meteorol. Z. 10 (4),
261–265. https://doi.org/10.1127/0941-2948/2001/0010-0261.
USDA, 2008. Lichen Bioindication of Biodiversity, Air Quality, and Climate: Baseline
Results From Monitoring in Washington, Oregon, and California. Retrieved from
https://www.fs.fed.us/air/documents/pnw_gtr737.pdf.
Varela, Z., López-Sánchez, G., Yáñez, M., Pérez, C., Fernández, J.A., Matos, P., Aboal, J.R.,
2018. Changes in epiphytic lichen diversity are associated with air particulate matter
levels: the case study of urban areas in Chile. Ecol. Ind. 91 (September 2017),
307–314. https://doi.org/10.1016/j.ecolind.2018.04.023.
Vareschi, V., Moreno, E., 1973. La contaminación en Caracas en los años 1953 y 1973.
Sociedad Venezolana de Ciencias Naturales, 30, 387–444.
WHO, 2013. Health Effects of Particulate Matter: Policy implications for countries in
Eastern Europe, Caucasus and central Asia (Vol. 50). https://doi.org/10.1097/
00002030-200204120-00009.
Will-Wolf, S., Makholm, M.M., Nelsen, M.P., Trest, M.T., Reis, A.H., Jovan, S., 2015.
Element analysis of two common macrolichens supports bioindication of air pollution
and lichen response in rural midwestern U.S.A. Bryologist 118 (4), 371–384. https://
doi.org/10.1639/0007-2745-118.4.371.