AIR HYGIENE REPORT no. 10
Contents

II HEAVY METALS

2 Lichens

  2.1 Introduction
  2.2 Monitoring design

2.2.1 Species selection
2.2.2 Sample collection and preparation
2.2.3 Chemical analysis

  2.3 National surveys
  2.4 Regional surveys
  2.5 National park surveys
  2.6 Urban and industrial surveys

2.6.1 Urban sources
2.6.2 Line point sources
2.6.3 Industrial point sources

  2.7 Transplants

2.1 Introduction

Lichens are effective biomonitors of metal deposition in that they possess many similar characteristics to bryophytes (Section 1.1). Lichens are slow growing and assimilate metals at a rapid rate but release them at a low rate. Metal concentrations in lichen thalli have been shown to correlate with atmospheric levels (Burton, 1986). Lichens were first used as bioaccumulative indicators in relation to point emission sources, where decreasing metal concentrations in species correlated with increasing distance from the source (Burton, 1986). Lichens have also been used to assess deposition patterns and heavy metal burdens for larger scale monitoring purposes. Their application as bioaccumulators on the multi-national and national scale is not as established as it is for bryophytes. The role of lichens in multi-national monitoring programmes is primarily in relation to gaseous air pollutants. References

Some important aspects in the design of programmes utilising lichens for the biomonitoring of airborne heavy metal pollution are considered briefly at the beginning of this section. Their application on national, regional and urban/industrial scales are presented. Transplantation techniques are also discussed. The uptake, retention and toxicity of heavy metals in lichens are not treated in detail here. A comprehensive review of these aspects is provided in Tyler (1989) and Tyler et al. (1989).

2.2 Monitoring design

Most of the comments in Section 1.2 are applicable to lichen monitoring for metal deposition. Some additional remarks are presented below.

In lichens, metals can accumulate to high levels by trapping insoluble particles (Tyler, 1989), extracellular ion exchange processes (Richardson, 1988), adsorption and active uptake (Král et al., 1989). Sloof (1995a) also indicated the importance of passive processes in the uptake and release of Co and Zn. Burton (1986) reviewed some of these mechanisms in different species. The author stressed the need to consider the differences in morphology and ion exchange properties between different lichen species when selecting a species for atmospheric heavy metal monitoring.

The nature and form of the metals under study is important in the selection of species in that this often determines whether the lichen will die, show symptoms or accumulate without apparent harm (Richardson, 1991). The chemical properties of an element and its associated particles affect its accumulation by a biomonitor. The sensitivity of lichens to elevated tissue concentrations of heavy metals varies greatly between species, populations and elements (Tyler, 1989). References

Obviously species with the ability to bioaccumulate high metal concentrations without apparent damage are more beneficial in biomonitoring studies.

In metal deposition studies, fructose (shrub-like) lichens have been recommended since these forms are easier to separate from the substrate in comparison to foliose (leaf-like) and crustose (crust-forming) lichens (Puckett, 1988). For example, foliose Parmelia was chosen over crustose Lecanora species in the Netherlands (Wolterbeek and Bode, 1995).

As mentioned previously (Section 1.2.2), availability of species within study areas often determines the selection of a species. Sloof and Wolterbeek (1993) discussed the possibility of interchangeability of lichen species within polluted areas by interspecies calibration.

Several lichen species are appropriate for the bioindication of heavy metal exposure (Box 2.1 from Kovács, 1992a).

Various sampling strategies are employed during collection. Some sampling strategies are more specific than others. Standardisation of collecting methodologies would allow more appropriate comparisons to be made between surveys.

In epiphytic lichen studies, some investigators collect only from one tree species, others from numerous tree species within a specified range. Some investigators specify collection height of lichens from trees e.g. 1.5 to 2 metres above ground. Others believe that sampling from all around the tree is important while others do not. Some samplers specify composite lichen volumes while some do not. References

Wolterbeek and Bode (1995) discussed the possible substrate contributions to lichen elemental concentrations. This may vary depending on species, substrate type and element. In the Netherlands lichens are sampled from protruding rims of rough bark in an attempt to minimise effects of excessive stem flow and build up of soil and bark particles at lichen attachments.

Bargagli (1990a) demonstrated that the younger, more external part of the thallus (corresponding to biomass produced the previous year) contained lower concentrations than the older, more internal part. In a study of abandoned mines in Italy, Bargagli (1990b) used the outermost three to five mm part of Parmelia sulcata thalli for the following reasons:

• It is easily distinguished by its colour and lack of rhizinae.
• It has known age (about one year).
• The outermost zone is almost completely devoid of rhizinae and as a result has little connection with the bark.
• Physiologically, it is the most active portion of the thallus.

Similarly in another lichen biomonitoring study in Italy, Loppi et al. (1994) used only the external part of Parmelia caperata. References

Variations in cleaning procedures can yield different results. For example, Bennett et al. (1996) confirmed that failure to remove excess matter from lichen samples affected elemental determinations.

The need for method standardisation has been expressed above (Section 1.2.3). For the provision of accurate results and error minimisation, the Standards, Measurements and Testing Programme of the European Commission has proposed the use of lichen certified reference materials. This is the first stage of a certification procedure as part of overall quality control in the analysis of lichen material throughout Europe (Quevauviller et al., 1996).

Other aspects of quality control such as experience of investigator, sample replication and awareness of local variations should also be applied. Wolterbeek and Bode (1995) treat these aspects in detail. References

2.3 National surveys

The gradient method, based on decreasing metal concentrations in species with increasing distance to source must be be applied with caution in large-scale surveys as unknown sources may contribute to metal content in lichens.

In the Netherlands, elemental concentrations in Parmelia sulcata obtained during two national lichen surveys undertaken over a five year period were presented as maps of geographical concentration patterns (Sloof and Wolterbeek, 1991). A standard sampling technique was utilised throughout. Healthy samples were taken from pH neutral bark substrates such as ash, alder, elm, oak, poplar and willow. Lichens were collected from all around the tree at heights between 1 to 2.5 m above ground. Samples were collected from open locations removed from farms and motorways. Application of target transformation factor analysis enabled the determination of elemental source profiles and source contribution. For example factor group Al, Cr, Fe, Mn, Sc and Th were associated with crustal material, Ni, V and Co with oil combustion processes and As, Cd, Sb, W and Zn with zinc smelters and/or electronic industry. Changes in the aerial heavy metal levels and their distribution were evident over the five-year period. Sloof (1995a) recommended the use of Monte-Carlo-assisted factor analysis (Kuik et al., 1993) as a refinement to data analysis of the 1986/87 lichen data in terms of source apportionment. References

2.4 Regional surveys

Herzig et al. (1989) evaluated passive monitoring using Hypogymnia physodes in the Swiss Midlands. This method was compared to the Calibrated Lichen Indication Method (which uses an index called IAP18, Chapter III, Section 2.3) as a qualitative and quantitative measure of individual air pollutants. Pb, Fe, Cu, Cr, total sulphur (S), Zn and phosphorus (P) concentrations in lichens showed good correlations with IAP18 values. Good correlations between Pb and Cu deposition gradients and Pb and Cu accumulation in Hypogymnia physodes were observed. The investigators recommended the combined use of the Calibrated Lichen Indication Method and passive biomonitoring method as an integrated biological air pollution monitoring system in Switzerland. By concentrating on human-toxic trace-elements measurements of risk to human health could be assessed. Both methods complement each other and can be applied to spatial and temporal monitoring on larger scales.

Analyses of indigenous lichen species, Parmelia praesorediosum and Ramalina stenospora, in south-west Louisiana demonstrated spatial and temporal changes in airborne heavy metal levels and distribution throughout the area (Thompson et al., 1987 and Walther et al., 1990a). The first survey took place in 1983/84 (Thompson et al., 1987) and the second in 1987/88 (Walther et al., 1990a). Contour maps, graphs and three-dimensional plots representing mean concentrations over the eighteen stations indicated higher element concentrations near the metropolitan and industrial areas than in the background sites. The results also demonstrated dramatic decreases in metal levels and distribution over time, which were attributed to reductions in industrial activity from 1982 to 1988. Discriminant analysis obtained statistical differences in element concentrations in lichens between stations up to 10.4 km from source (group 1) and stations beyond this distance (group 2). The authors showed a straight-line relationship for Fe and Zn concentrations with the distance from industrial corridor and concluded that their similar slopes imply a common origin and similar rate of washout for these elements (Thompson et al., 1987). Enrichment factors for each element were investigated by dividing the mean concentrations of group 1 stations by the mean concentration of the background stations.References

Residence times for Al, Fe, Hg and Zn were calculated according to the equations:

Residence times were postulated to be independent of background values and dependent on original concentrations. The mean life residence times for Al, Fe, Hg and Zn were 1.7, 2.0, 2.3 and 3.5, respectively, which correspond to mean residence times of two to five years proposed by Nieboer and Richardson (1981). This suggests that lichens are expected to show a decrease in elemental content a year or two after exposure to an emission source ceases. References

2.5 National park surveys

Increasing research has been conducted into the effects of air pollution on national parks in the US. This probably relates to concern regarding the threat of air pollution from nearby cities. Furthermore, such relatively unpolluted areas represent background sites and provide good baseline data. The application of lichens as bioaccumulators of heavy metals in this type of monitoring is illustrated below.

Observations of Hypogymnia physodes and Evernia mesomorpha at eighteen localities in the Isle Royale National Park, Michigan between 1983 and 1992 demonstrated the temporal and geographical pattern of elements in the area (Bennett, 1995). Results of sixteen elemental analyses were compared to the sequence of element concentrations in the earth's crust. Zn, Pb, Se and sulphur (common air pollutants) for example ranked higher in lichen tissue than they did in the earth's crust. Heavy metal elements associated with atmospheric deposition increased over the nine-year study period to a greater extent than non-metallic elements. Geographic distributions of anthropogenic elements were elevated around known local pollution sources and some elements (e.g. sulphur) reached toxic levels in lichens. The author recommended this method, i.e. employing more than one species and analysing elemental data in aggregate rather than individually, as an appropriate indication of early stages of air pollution within less polluted areas.

Frenzel et al. (1990) observed higher burdens of As, Cd, Cu and Zn in Alectoria sarmentosa in Mount Rainier National Park, Washington, than in a control site (Olympic National Park). Industrial sources near to Mount Rainier National Park were thought to be the cause. In contrast, higher levels of Pb were obtained in Olympic National Park, which were attributed to dispersed regional and global sources. References

Comparisons of lichen tissue metal levels in the aforementioned national park studies are presented in Table 2.5. Despite slight differences in methodologies and different lichen species, Isle Royale National Park showed consistently higher tissue element levels than the other two parks.

2.6 Urban and industrial surveys

Urban and industrial studies generally entail spatial biomonitoring such as distribution/fall-out patterns and correlations with distance from source.

The application of lichen biomonitoring in three urban surveys in different continents is compared in Table 2.6. Species utilised, number of sampling sites, type of analyses and thalli metal concentrations are expressed. Details of each study are provided in the text where the increasing importance of statistical analyses in such studies is apparent. References

A distinct difference in metal concentrations in lichen tissue between industrial and urban zones was evident in the Baton Rouge area, Louisiana (Walther et al., 1990b). Levels decreased with increasing distance from the city. Concentration ranges (µg g^-1) are shown in Table 2.6. Discriminant analysis indicated insignificant differences in metal levels between the two lichen species, and average values of the two species were used to construct three dimensional and contour plots of metal concentrations in the area. Industrial activity and urban traffic were responsible for the observed heavy metal trends.

Levels of atmospheric Cd, Cr, Cu, Hg, Ni, Pb and Zn around Pistoia in central northern Italy were assessed using the widely distributed indigenous lichen Parmelia caperata (Loppi et al., 1994). Concentration ranges (µg g^-1) are presented in Table 2.6.

Distribution maps and cluster analysis indicated similar distribution patterns and fall-out patterns for Cd, Pb and Zn, with maximum values observed in the centre of Pistoia. Cd was a by-product of lead and zinc smelting industry in the area. Fertiliser and pesticide use in plant nurseries explained correlation between Zn and Cd south east of town. Correlations, cluster analysis and distribution maps demonstrated an association between Cr and Ni, which was attributed to metal-plating industries in the southern part of the study area. The authors used coefficients of variation for each element (Garty and Ammann, 1985) as a relative measure of dispersion of particles. Cu showed a low coefficient of variation implying it was dispersed relatively consistently as small particles over the study area. Overall, the analysis indicated that most metal levels were modest in comparison to heavily polluted areas. Fertilisers and pesticides, which produced the relatively high zinc concentrations, posed the biggest pollution threat in the area. References

The degree of heavy metal contamination by heavy industry and motor vehicles in Kampala, Uganda was indicated by metal bioaccumulation in the lichen Calyrneferes usambaricum (Nyangababo, 1987). Heavy metal concentrations in lichen tissue were higher in the rural area than in the urban area (Table 2.6). Pb levels were higher than those obtained two metres from a highway in northern Nigeria (Kapu et al., 1991), suggesting a heavy metal burden from traffic within the study area. The author applied Grodzinska's (1978) synthetic pollution indices for mosses to the lichen data. This used standardised metal values divided by the number of elements to calculate a pollution index at each site. The author used these values to classify the sites into clean and heavily contaminated, which corresponded to the metal concentration gradient. However, it is debatable whether these methods would identify intermediately polluted sites as distinctly.

The species mentioned in Table 2.6 appear to be adequate biomonitors of heavy metal pollution patterns. The extent of data analysis is important. Multivariate analyses are useful in 'grouping' sites and metals, making conclusions easier. The surveys varied in their choice of metals for analysis, which ultimately depends on knowledge of the local area and pollution sources. References

Kapu et al. (1991) used the bioaccumulative properties of Parmelia sp. to assess the aerial fallout of heavy metals from traffic in Zaria, northern Nigeria. Metal concentrations in epiphytic lichens, with the exception of Fe, decreased significantly with distance from Zaria/Samaru-Sokoto Highway but showed no significant difference along the residential road in the Samara Campus, Ahmadu Bello University, Zaria (Table 2.7). The former area was therefore suffering from aerial heavy metal dispersion and contamination. Such studies are vital in developing countries where heavy metals pose a serious health risk.

2.6.3 Industrial point sources

(i) Species distribution

Lichen distribution studies in relation to metals are rare in comparison to their use in gaseous air monitoring. St. Clair et al. (1995) compared lichen communities in two sections of Anaconda-Pintler wilderness area, Montana. Only 21 species, of which two were regarded as sensitive, were observed in the sites located in the area between a defunct copper smelter and the wilderness boundary. In comparison, 161 species and 17 indicator species were found in the sites across the rest of the wilderness area. Wind blown dust and contaminated soil containing Ni, As, Cu and Pb from the smelter were thought to be responsible. References

(ii) Bioaccumulation

Nash and Gries (1995) summarised various studies utilising lichens as bioaccumulators around point sources in arctic/boreal localities. All examples showed significantly lower lichen concentrations in sites removed from the industrial point source.

In a study of trace-element concentrations in Parmelia caperata and Parmelia rudecta in the vicinity of a coal-fired power plant near Washington D.C. no significant differences in concentrations were apparent in lichens at distances ranging from 1.6 to 20 km from the plant (Olmez et al., 1985). Various reasons for this apparent anomaly were proposed, e.g. tall stacks, and preference of lichens to collect large airborne particles of similar composition to their crustal material.

In their comparison of the suitability of a variety of plant species as bioindicators of steel works in coniferous forests in Raahe, Finland, Mukherjee and Nuorteva (1994) found that Grodzinska's (1978) pollution index system could be applied using the lichen Hypogymnia physodes. The pollution index of site j (Sj) was calculated according to the following equation:

Calculated indices were correlated with forest condition and lichen quality. For example, at a distance of one km from the steel works, a pollution index of 4.39 corresponded to extremely serious forest damage and lichen poor areas. In contrast, at a distance of 10 km an index of -0.63 was found, in accordance with a 'healthy forest' and lichens 'slightly affected by air pollution'. References

The pattern of air pollution in the vicinity of an aluminium smelter in Yugoslavia was investigated using epiphytic and lithophytic lichens (Hypogymnia caperata, Diploicia canescens and Lecanora expallens) (Jovanovic et al., 1995). Epiphytic lichens were absent from the immediate zone around the smelter. Correlations existed between the elemental composition in the lichens and air pollutants associated with the works. Metal concentrations in lichen tissue decreased with increasing distance from the smelter. On comparison with other biological indicators (grasses and pine needles), lichens proved more effective in terms of bioaccumulative characteristics. This work provided good baseline data for future biomonitoring programmes in this area.

Relatively low levels of trace-metals were found in Parmelia caperata samples around the Travale-Radicondoli geothermal area in Italy (Loppi and Bargagli, 1996). High correlations of many elements with Al, a common indicator of crustal material, implied soil dust was a source of these metals. B and Hg, common geothermal pollutants, were correlated to distance from geothermal sources. B concentrations ranged from 5.1 to 22.1 µg g^-1 and Hg from 0.062 to 0.555 µg g^-1. The other major contaminant associated with geothermal pollution, As, appeared to be derived from power plant sources and neighbouring thermal springs.

In Idaho, inverse relationships between Cd, Cr, Zn and P concentrations in Rhizoplaca melanophthalma and distance from phosphate refineries were obtained (Dillman, 1996). More intensive monitoring of this semi-arid area was recommended.

The epiphytic lichen Parmotrema madagascariaceum was used to assess atmospheric trace-element contamination to Venezuelan cloud forests (Gordon et al., 1995). Increasing Pb from leaded gasoline was proposed as the greatest risk to the forests due to the lack of abatement procedures such as those adopted in temperate regions.

Several studies have concentrated on the specific biomonitoring of aerial mercury. These are discussed below. References

In an abandoned mining area around Mount Amiata, Italy, mercury levels in the indigenous lichen Parmelia sulcata were used to assess mercury emissions to enable subsequent environmental reinstatement of the area (Bargagli, 1990b; Bargagli et al., 1989 and Ferrara et al., 1988). Mercury concentrations in lichen tissue within the study area ranged from 0.31 µg g^-1 to 7.80 µg g^-1. Mercury concentrations in the lichen had a highly significant exponential relationship (p < 0.001) with distance from one of the plants and a very highly significant relationship with mercury content in surface soil (p < 0.0005). The latter suggested that lichens accumulated Hg from degassing of nearby soils. Aluminium assays (indicator of crustal derived material) implied that soil particles are trapped by P. sulcata but did not show a correlation with mercury. Lichen, soil and air sampling concluded that the main sources of gaseous mercury were large piles of roasted cinnabar, geothermal power plant emissions and ventilation systems of mineshafts. P. sulcata was recommended for further monitoring in the area. Other metals did not show similar patterns to Hg. Zn levels in the lichen were possibly a result of long-range atmospheric transport. Fe and Mn levels were attributed to accumulation of soil particles.

Dispersion of Hg from volcanic eruptions in Hawaii was investigated by measuring concentrations of the element in the lichen Stereocaulon vulcani (Davies and Notcutt, 1996). Mercury concentrations ranged from <8 µg g^-1 to 59 µg g^-1 with most values below background atmospheric levels for the island. Elevated mercury levels (twice background levels) were associated with two local irregular sites. This data proved more beneficial in assessing dispersion and sources of Hg than previous physico-chemical air monitoring. References

2.7 Transplants

Transplantation exercises are rarely undertaken on a large scale and are mainly focused around point source emissions. Examples of their application in various countries are illustrated below. Vestergarrd et al. (1986) compared transplantion data from 1977 and 1982 to assess the effect of changing from oil-fired open-hearth furnaces to electric-arc furnaces in a Danish steel factory. Hypogymnia physodes samples removed from Pinus sp. 7.6 km from the works were exposed for seven months at a height of 1.5 m on wooden stands at twelve transplant sites in the vicinity of the works in 1977 and in 1982. Regression analysis was used on the data.

Metal concentration was related to distance from source by following equation:

Metal concentration in lichens and bulk precipitation were related by following equation:

Lichens showed a decrease in metal content with increasing distance from source. Comparison of regression lines showed that Cr, Cu and Pb had decreased significantly between survey years 1977 and 1982. Relationships between lichen and bulk precipitation data showed that metal uptake of lichen was not proportional to the concentrations measured in bulk precipitation although direct proportionality was observed when each station was considered individually. The difference in relative uptake in lichens between 1977 and 1982 was attributed to changes in the particle size distribution of the emission.

In Israel, Garty and Hagemeyer (1988) carried out a similar impact assessment study by comparing metal content in lichens transplanted within the vicinity of a coal-fired power plant before (1979/80) and after (1983/84) initiation of operation of the plant. Twigs containing Ramalina duriaei were suspended from a variety of local trees at ten sampling locations, ranging from four to thirty km from the plant. In comparison to the previous example, this survey covered a much greater area and spanned urban, agricultural and rural territories. Furthermore, slightly different statistical analyses were employed. Duncans' multiple range tests compared metal concentrations in lichen thalli between sampling locations in each study period. Correlations and two-way analysis of variance (ANOVA) showed that Cr levels in the region had increased, regional Cu and Zn concentrations had decreased and regional Ni remained the same. However, some local elevations in metal concentrations in lichens were apparent (e.g. Ni levels increased from 7.7 µg g^-1 to 33.6 µg g^-1 at a local nature reserve site, 9.4 km away from the plant). Reduction in Cu and Zn levels were attributed to changes in agricultural practises in the region. References

Potential improvement in air quality in Indiana Dunes National Lakeshore Park was examined by transplanting Hypogymnia physodes (Bennett et al., 1996). This species was formerly present in the park and its impoverishment has been attributed to air pollution. The lichens, attached to small branches, were positioned on artificial trees at four sampling locations across 333 km north to south. Tissue concentrations of 20 elements were analysed every year for three years. ANOVA and Tukey's Single-Degree-of-Freedom Test for Non-Additivity showed that the transplanting process itself did not effect lichens. Most elements showed elevated concentrations in lichens at Indiana Dunes compared to control sites and some elements increased significantly over the three-year study period. Increased mortality between the second and third year was attributed to the deleterious effects of significant increases in a number of elements, synergism of Zn, Cd and K and exceedance of maximum concentrations of some elements. Although transplant density was not intensive, the study was able to illustrate general trends in the area.

Biological parameters other than element bioaccumulation in thalli have been performed to assess heavy metal contamination in transplants. In Israel, inverse relationships were obtained between Pb and Cu content and ATP concentration in transplanted lichen thalli. Additionally, the degradation of chlorophyll-a to phaeophytin-a, expressed as a ratio, was inversely correlated to levels of Cu, Pb and Zn in lichen (Garty et al., 1988). References

Home