Tracking global change using lichen diversity: towards a global‐scale ecological indicator

Summary

1. Lichens have been used to efficiently track major drivers of global change from the local to regional scale since the beginning of the industrial revolution (sulphur dioxide) to the present (nitrogen deposition and climate change). Currently, the challenge is to universalize monitoring methodologies to compare global change drivers’ simultaneous and independent effects on ecosystems and to assess the efficacy of mitigation measures.
2. Because two protocols are now used at a continental scale North America (US) and Europe (EU), it is timely to investigate the compatibility of the interpretation of their outcomes. For the first time, we present an analytical framework to compare the interpretation of data sets coming from these methods utilizing broadly accepted biodiversity metrics, featuring a paired data set from the US Pacific Northwest.
3. The methodologies yielded highly similar interpretation trends between response metrics: taxonomic diversity, functional diversity and community composition shifts in response to two major drivers of global change (nitrogen deposition and climate). A framework was designed to incorporate surrogates of species richness (the most commonly used empirical trend in taxonomic diversity), shifts in species composition (compositional turnover) and metrics of functional diversity (link between community shifts to effects and ecosystem structure and functioning). These metrics are essential to more thoroughly comprehend biodiversity response to global change. Its inclusion in this framework enables future cross‐continental analysis of lichen biodiversity change from North America and Europe in response to global change. Future works should focus on developing independent metrics for response to global change drivers, namely climate and pollution, taking us one step closer to a lichen‐based global ecological indicator.

Introduction

Currently, the scientific community is challenged to develop tools to compare and assess the impacts of global change drivers and the effects of the political measures adopted in its response at the world‐wide scale (Branquinho, Matos & Pinho 2015). Because the same amount of change in a driver will not have the same effect on different ecosystems (e.g. a 2 °C increase will have different effects on desert or alpine ecosystems), ecological surrogates are an efficient tool to assess and to complement indicators based on driver's change. These are components of the ecosystems that can be used as indicators of an attribute, trait, characteristic or quality of that ecosystem (Mellin et al . 2011), providing a more cost‐effective way to assess ecosystems’ response. In this regard, the United Nations Conventions on Biological Diversity (UNCBD), Climate Change (UNCCC) and Combatting Desertification (UNCCD), have long demanded a set of globally applicable indicators, to track the impacts of global change drivers on ecosystems. These are intended to measure progress towards their targets, and to improve and guide new strategies for biodiversity conservation (Pereira & Cooper 2006). Although there are some population indicators based on vertebrates at the global scale (Pereira, Navarro & Martins 2012), data from less known taxonomic groups are needed. This global monitoring network should contemplate several indicators to provide a more robust and integrated picture of the effects of global change drivers on ecosystems’ functioning and structure. Because multiple global change drivers often act simultaneously, it is as well essential that the response of these indicators to independent environmental drivers can be disentangled (Branquinho, Matos & Pinho 2015). In order to achieve it, data collection must be based on compatible standardized methodologies performed at regional and global scales (Mace et al . 2005; MEA, 2005, Pereira & Cooper 2006).

Epiphytic lichen diversity has been used to efficiently monitor the effects of the major drivers of global change since the beginning of the industrial revolution (sulphur dioxide; Gilbert 1973; Hawksworth & Rose 1970). Modern lichen‐based environmental analyses provide low cost, high‐resolution spatial tools for modelling and mapping the effects of environmental change, when compared to the traditional networks of pollution or climate monitoring stations, allowing its detection, assessment and monitoring at the ecosystem level (Branquinho et al . 2008; Pinho et al . 2008b). They are powerful ecological indicators of pollution (Giordani, Brunialti & Alleteo 2002; McMurray, Roberts & Geiser 2015) and of macro and microclimate change (Aptroot & Van Herk 2007; Pinho, Máguas & Branquinho 2010; Matos et al . 2015; Root et al . 2015). Lichen‐based thresholds were used to establish the lowest critical levels for nitrogen concentration and critical loads for nitrogen deposition in Europe and the USA, contributing to the protection of ecosystem services and functions in both natural and semi‐natural ecosystems (Cape et al . 2009; Pardo et al . 2011; Pinho et al . 2012; Root et al . 2015). Furthermore, works have shown that it is possible to use them even under the influence of multiple factors, like multiple land uses or pollution and climate (Geiser & Neitlich 2007; Pinho et al . 2008a; Root et al . 2015).

Since the 1970s, many different methods were used to assess lichen diversity. Among these, two standardized methodologies for sampling lichen epiphytic diversity predominate presently, one in Europe (EU) and another in the United States of America (US). The European standard methodology was developed by Asta et al . (2002) and was recently adopted under the Comité Européen de Normalization (CEN) framework (Ambient air – Biomonitoring with lichens – Assessing epiphytic lichen diversity. European Standard EN 16413:2014). The US method was developed by the US Forest Service (USDA, 2011) to monitor air quality and climate change, and it is also used in Canada and Mexico. The methods differ largely in two main aspects: (i) the EU method records all lichen species (macro and microlichens) occurring inside a size‐standardized grid placed on a tree trunk in a fixed number of trees, resulting in metrics of frequency; (ii) the US method surveys all macrolichen species detected on any tree or shrub inside a large circular sampling plot (0·4 ha), visually rating species abundance. Although desirable, it would likely be a Herculean task to convince scientists on both continents to adopt only one of the methodologies, or even to develop a new common one. If achieved, we would still have problems using retrospective data. It is essential to investigate these methods compatibility and develop a way to use their outcomes jointly. This would enable cross‐continental analysis of lichen monitoring data under a global change perspective, and possibly allow its inclusion in the global monitoring network.

Assessing outcomes of methods under a global change perspective requires a proper selection of biodiversity metrics. Species richness metrics are necessary because they measure the irreversible component of biodiversity (species extinctions, biodiversity loss), at local, regional or global scales (MEA 2005; Pereira et al . 2010; Pereira, Navarro & Martins 2012). However, biodiversity loss metrics cannot be so immediately linked to ecosystem services and are not so responsive to global change drivers (Balmford, Green & Jenkins 2003; Dornelas et al . 2014). Hence, shifts in community structure, integrating measures of species abundances, should be considered as they are more responsive (Balmford, Green & Jenkins 2003; Dornelas et al . 2014). Because functional diversity is a potentially more universal indicator of community changes (independent of species identity) and is linked to ecosystem functionality in response to anthropogenic drivers across broad spatial scales and environmental gradients (Lavorel et al . 2011), measures of functional diversity should likewise be considered.

Here, we present a framework to jointly analyse the interpretation of data generated by the two most widely used lichen survey methods using broadly accepted biodiversity metrics to track the effects of global change. We aim to answer these questions: (i) can lichen data acquired by two different survey methodologies, US and EU, give comparable interpretation in terms of surrogates of biodiversity metrics, namely in terms of change in taxonomic diversity (species richness, Shannon, Simpson and evenness diversity indices), shifts in species composition and functional diversity (community weighted mean, CWM) along climate and pollution gradients; (ii) despite the expected different absolute values resulting from the different methodologies, are methods outcomes similar in terms of interpretation (quality and level of change); (iii) if so, how should they be compared under a global change perspective? To answer these questions lichen epiphyte diversity was determined using both methods at 28 sites spanning regional nitrogen deposition and climatic gradients in the northwestern US. Climate and nitrogen pollution gradients were included because these are currently some of the emergent and most pressing drivers of global change (Steffen et al . 2015).

Materials and methods

Study area and sampling sites

The study area was located in the northwestern United States (Fig. 1). Eleven low elevation sites were surveyed in the Columbia River Gorge National Scenic Area following a pollution gradient from Portland, Oregon. The remaining sites were on‐frame USFS Forest Inventory Analysis P2 plots (http://www.fia.fs.fed.us/library/field-guides-methods-proc/index.php) in the temperate rain forests of the western Oregon and Washington Cascades [Rogue River‐Siskiyou (1), Mount Hood (1), and Gifford‐Pinchot (2) national forests] and dry coniferous national forests of the eastern Cascades and Blue Mountains of Oregon [Fremont‐Winema (6), Ochoco (1), Umatilla (5) and Wallowa‐Whitman (1)]. Sites were visited during June and July 2013. Elevational, climatic and pollution profiles are given for each of these areas in Table 1.

Figure 1

Map of the study area showing sampling plots location, in Northwestern US . Different symbols represent different geographic and climatic areas, as grouped by cluster analysis. N  = 28. Squares stand for plots in high elevation wilderness areas (14), circles for plots of in eastern Columbia River Gorge (8), and triangles for plots in western Columbia River Gorge (6).

Table 1. Range of elevational, climatic and pollution conditions of the study sites, as grouped by cluster analysis: E CRG, Eastern Coumbia River Gorge; W CRG, western Columbia River Gorge; HWA, high elevation wilderness areas (N  = 28)

	E (m)	T max (°C)	T min (°C)	T (°C)	P (cm)	RH (%)	N (lichen N kghayr)
E CRG	58–462	28·3–29·8	−3·4 to −1·8	9·6–11·1	46·1–63·8	45–52	1·61–6·42
W CRG	66–617	26·0–27·8	−2·7 to −0·1	9·6–10·9	70·7–191·8	52–55	1·01–2·11
HWA	1298–2267	20·4–27·6	−10·3 to −4·0	3·9–8·0	42·7–143·1	38–51	0·58–2·47

• Climate variables: E , elevation; T _max, maximum temperature in August; T _min, minimum temperature in December; T , annual mean temperature; P , annual precipitation; RH, relative humidity. Pollution variable: N, nitrogen deposition given in terms of dry weight N concentrations in lichens.

US Lichen sampling method

The US method to survey epiphyte macrolichens followed the Forest Inventory and Analysis (FIA) lichen community indicator protocol (USDA FS 2011). Briefly, a trained surveyor circumambulates a circular plot of 0·38 ha for up to 2 h, collecting a voucher of each epiphytic macrolichen species detected on woody plants above 0·5 m from the ground: including trees boles, branches and twigs, branch litter, saplings, shrubs and standing dead trees. For each species detected, an abundance rating is assigned based on the number of individual thalli observed during the survey: 1 – rare (1–3 individuals); 2 – uncommon (4–10 individuals); 3 – common (>10 individuals, occurring on less than half of available substrates); 4 – abundant (species present on more than half of all substrates). So, a species that was detected on more than half of the trees received an abundance rating of 4. All vouchers were identified in the laboratory to species following Mccune & Geiser (2009) and taxonomy followed Esslinger (2012).

EU Lichen sampling method

To apply the EU method to the US plots, a minimum of four and maximum of eight trees near the centre of the plot were selected for sampling, determined by the time spent on each tree and the total time (2 h) available to survey the site. At 130 cm above ground level, each tree selected: (i) trunk circumference between 50 and 250 cm; (ii) not leaning more than 20°; and (iii) with a clear area on the trunk without damage, decortication, branching, knots, or other epiphytes preventing lichen growth. A 10 × 50 cm frame divided into five 10 × 10 cm grid cells was placed on the north face of the tree trunk. The uppermost edge of the frame was positioned at 150 cm from the ground level, adjusted up to a maximum of 2 m height if the trunk was unsuitable at the desired height (e.g. to avoid snow lines or branches). Each lichen species occurring inside each grid cell was identified and recorded or was collected for later laboratory identification. Sampling was repeated on the S, E and W‐facing sides of the trunk. Lichen abundance (frequency) was recorded as the number of grid cells (out of 20 possible) in which each species was detected. Lichen identification followed Mccune & Geiser (2009) and B. Mccune (unpublished data); taxonomy followed Esslinger (2012). The LDV (Lichen Diversity Value) index was calculated for all species (EU) and for macrolichens only (EUm) following Asta et al . (2002). The frequency for each species (species LDV) is calculated as the mean frequency on all the trees sampled at the site.

Lichen functional diversity

Lichens are sensitive to a wide range of pollutants, and their sensitivity to nitrogen pollution is particularly well studied (Pinho et al . 2008b; Geiser et al . 2010; Munzi et al . 2014; Root et al . 2015). As all organisms, lichens need nitrogen for their nutrition, but different species have different nitrogen requirements. While some can live in nutrient‐rich environments, others more adapted to nutrient‐poor environments, disappear if nitrogen load in the atmosphere is high. Based on this differential sensitivity, each macrolichen species was assigned to a N‐sensitivity functional group (Table S1, Supporting Information) related to the depositional loading (kg N⁻¹ ha⁻¹ per year) above which the probability of being detected in the US Northwest declines: oligophilic (≤2·6); mesophilic (2·7–8); nitrophilic (>8). Ratings follow Geiser et al . (2010) and Root et al . (2015) for species west and east of the Cascades Range crest respectively. When species scored differently east and west of the Cascades Range crest, east score was given (due to the highest number of sites in this part of the range). Because previous studies did not contemplate microlichens, only macrolichens were assigned to functional groups.

Species functional group assignments were combined with species abundance data from each sampling method (US, EU and EUm) to obtain the community‐weighted mean (CWM; Lavorel et al . 2008) for each functional group at each plot. This index represents the mean functional group values in the community, weighted by the abundance of species belonging to that functional group (Lavorel et al . 2008). This was calculated using the ‘dbFD’ function of the CRAN software r (R Core Team 2013), fd package (Laliberté & Legendre 2010).

Climate and pollution data

Estimates of 30 years normal (1970–2000) annual mean precipitation, temperature and relative humidity were extracted from the Parameter‐elevation Regressions on Individual Slopes Model (PRISM Climate Group, Oregon State University, http://prism.oregonstate.edu, created 15 August 2013) at an 800 m cell resolution using a GIS overlay on the plot coordinates. Twenty grams of Letharia vulpina or Platismatia glauca were collected and analysed for total elemental N following Geiser (2004). Dry weight N concentrations (%) in lichen thalli were used to estimate canopy through‐fall deposition of total N from nitrate and ammonium ions following Root et al . (2013), as no good model of N deposition was available for the whole range of climatic conditions we had in this gradient. Because values between species were similar, when both species were available for collection, the final value corresponded to an average of all the samples from that site.

Data analysis

Four taxonomic diversity indices (species richness, Shannon–Wiener index, Simpson's index and local Pielou's evenness) were computed per plot using the species abundance matrices for all species (EU) and macrolichens only (US and EUm). We calculated Pearson correlations (r ) between indices derived from the different sampling methods (correlations were considered significant for P  <  0·05).

Plots were aggregated into distinct geographical and/or climatic areas using hierarchical, agglomerative cluster analysis with Euclidean distances and Ward's linkage method, based on the three species matrices derived from the US and EU methods using the PC‐Ord Software Version 6.08. The choice of optimum number of groups to prune the dendrogram was done using Indicator Species Analysis (ISA; Dufrêne & Legendre 1997). Plots were clustered into up to 17 groups and an ISA analysis was performed on the climate matrix for each group memberships. Resulting P ‐values for each climate variable were averaged for each level of grouping and the number of significant (P  <  0·05) indicator climate variables was registered. The optimal number of groups was chosen pondering the lowest P ‐value with the highest number of indicator climate variables (Mccune, Grace & Urban 2002). Significance of the groups formed was assed using multi‐response permutation procedure, with groups considered significantly different if P  <  0·05.

Non‐metric multidimensional scaling (NMS) was used to ordinate the EU, US and EUm species matrices to extract prominent gradients in lichen community composition. When sampling sites include a large gradient with very distinct local site characteristics unrelated to the environmental gradient of interest, values from the EU method are usually relativized (by total species frequencies at plot level) to prevent biasing results and impairment of comparisons (Matos et al . 2015). Because the US species abundance scale is log‐like, values from EU species matrices were log transformed after relativizing. This transformation improves comparison outcomes (similarity between distance matrices and ordination scores improves around 3%), and although not essential, we recommend log‐transforming as a step in this framework. The best NMS solution was run with Bray‐Curtis distance (Mccune, Grace & Urban 2002), chosen from 500 runs, each starting randomly (500 iterations per run), and evaluated with a Monte Carlo test (250 runs with randomized data). The coefficients of determination (r ²) between the original plot distances and distances in the final ordination solution were calculated to assess how much of the lichen community variability was represented by the NMDS axes (Mccune, Grace & Urban 2002). Climate, pollution and functional variables were overlaid on the NMS ordination as correlation vectors (Mccune, Grace & Urban 2002). Individual correlations between these variables and NMS site scores were determined using Pearson correlations (correlations were considered significant for P  < 0·05). The correlation (r ) and redundancy (r ², squared correlation) between ordinations were assessed using a Mantel test (comparison of matrices sites × axis scores; matrices were considered related for P  < 0·05). Associations between distance matrices obtained with both methods were assessed using the Mantel test (comparison of distance matrices built based on sites × species matrices; matrices were considered associated when P  < 0·05).

Results

Species richness and Shannon‐Wiener indices correlated strongly between US and EU (total and only macro) data sets (Fig. 2). This was also observed for Simpson index (r  = 0·78, r  = 0·80 and P  < 0·001 respectively). Species richness was higher in the EU data set, and lower in the EU macrolichen only data set. Conversely, the US method yielded higher values of Shannon‐Wiener and Simpson's indices than both EUm and EU data sets, which were in turn very similar. Pielou's evenness index values for US vs . EU data sets were weakly correlated (r  = 0·38, P  =  0·044); and we found no significant correlation between macrolichen only data sets (US vs. EUm; r  = 0·25, P  = 0·199). Species richness, Shannon‐Wiener and Simpson indices based on both methods were all significantly correlated with some climate variables: negatively with elevation, and positively with mean annual temperature and relative humidity (Table 2). Annual precipitation and lichen nitrogen content weren't significantly correlated with any of these diversity indices. Pielou's evenness index derived from both methods showed no correlation with the environmental variables.

Figure 2

Taxonomic diversity indexes computed using data collected with US and EU methods to assess epiphytic lichen diversity. Gray symbols represent EU macrolichens data and black symbols indicate EU data for all lichens. Numbers indicate Pearson correlation coefficients and significance. Solid lines are US vs. EU regression lines, whereas dashed lines indicate 1 : 1 relationships were US and EU values are equivalent.

Table 2. Summary of Pearson correlations between taxonomic diversity indices and climate, pollution variables (N  = 28)

Type	Variable	US				EUm				EU
		S	H	D	E	S	H	D	E	S	H	D	E
Climate	E	−0·61*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·63*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·63*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·29	−0·73*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·73*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·68*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·19	−0·69*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·67*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·63*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·23
	T	0·52** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·56** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·57** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·36	0·70*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·73*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·69*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·15	0·64*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·65*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·65** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·19
	P	−0·02	−0·04	−0·05	−0·21	−0·08	−0·18	−0·23	−0·22	−0·01	−0·15	−0·18	0·01
	RH	0·66*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·70*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·70*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·13	0·76*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·78*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·74*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·3	0·78*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·79*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·74*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·36
Pollution	N	−0·01	0·01	0·03	0·23	0·23	0·27	0·27	0·08	0·14	0·17	0·18	−0·11

• Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Taxonomic diversity variables: S, species richness; H, Shannon‐Wiener; S, Simpson's diversity index; E, evenness. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.

A cluster analysis of the US and EU species matrices assigned the same sampling sites to the same three groups when the dendrogram cut‐off was set to three groups. The NMS extracted nearly similar environmental gradients from data generated by both methodologies. The ordination (Fig. 3) shows the NMS solution for the location of the sampling sites in species space, that is, based on lichen community composition. The analysis suggested a two‐dimensional solution for the three data sets; the addition of a third axis yielded only a slight reduction in minimum stress. The final stabilities of the US, EUm and EU ordinations were 7·04%, 7·86% and 8·23% respectively. Minimum stresses were lower than would be expected by chance for the three solutions (P  =  0·004). Total variation explained was 95·8%, 81·4% and 83·5% respectively.

Figure 3

Non‐metric multidimensional scaling (NMDS ) ordinations of lichen community data from US and EU sampling methods. Vector overlays indicate the magnitude and direction of correlations between environmental variables and the ordination (see also Table 1). Climate variables: T, annual mean temperature; P, annual precipitation and E, elevation. Pollution variable: N, dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species and Oligo, CWM of oligophilic species. Symbols represent the model climatic and geographic areas defined by cluster analysis: circles, eastern Columbia River Gorge, triangles, western Columbia River Gorge and squares, high elevation wilderness areas.

Climatic and geographic groups detected by cluster analysis were also apparent in the NMS ordinations. US ordination scores were very similar to the EUm (Mantel r  = 0·81, P  <  0·001) and EU (Mantel r  = 0·80, P  < 0·001), with information redundancies of 64% and 65% respectively. The US distance matrix was also highly associated with both EU matrices (both EUm and EU with Mantel r  = 0·81, P  <  0·001). As expected from the overlap between data, EU ordination scores (macro and total) are highly similar (Mantel r  = 0·98, P  =  0·001), and with an information redundancy of 96%. Ordinations of the US, EUm, and EU lichen community composition also extracted the same two major macroclimatic gradients. The vector overlays in Fig. 3 indicate correlating climate, pollution and functional group variables. The first axis represents a temperature and elevation gradient, evidenced by the high correlation coefficients of temperature and elevation with this axis (Table 3). The linear regressions between axis one scores and temperature almost completely overlap, emphasizing the similarity between methods. To a less extent, this axis also represents a relative humidity gradient, slightly stronger on the US method ordination (Table 3). In all the ordinations, the second axis represents a moisture gradient of precipitation and relative humidity (Table 3). Nonetheless, with the US method, this gradient is more strongly related to precipitation, while with the EU method (both EU and EUm) relative humidity is slightly stronger. Similar pollution gradients were predicted by ordinations of all three data sets. Lichen nitrogen content was significantly correlated with both axes, except for axis 2 of EUm ordination (Table 3). Nonetheless, while for the US method this pollution gradient was almost equally distributed by both axes, in the EU method the gradient was more strongly reflected on the first axis of the ordination. A rotation of the ordination was attempted, but as it did not generate independent axes for climate and pollution, the original configuration was kept.

Table 3. Summary of Pearson correlations between NMDS ordination axes and climate, pollution and functional variables. Percentage of variation explained by each axis is shown right under the axis number (N  = 28)

Type	Variable	US		EU (macro)		EU
		Axis 1	Axis 2	Axis 1	Axis 2	Axis 1	Axis 2
		77·1%	18·7%	61·4%	20%	65·8%	17·7%
Climate	E	0·98*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·01	0·93*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·18	0·93*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·20
	T	−0·93*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·01	−0·92*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·11	−0·93*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·12
	P	0·178	−0·64*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·38* Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·43* Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·37	−0·41* Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.
	RH	−0·70*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·39* Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·61** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·54** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·63*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·59** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.
Pollution	N	−0·55** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·58** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·65*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·39* Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·64*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·34
Functional diversity	Nitro	−0·84*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·51** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·80*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·50** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·82*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·42* Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.
	Meso	−0·69*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·58** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·42*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·66*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·43* Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·69*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.
	Oligo	0·96*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·17	0·88*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·06	0·97*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·03

• Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.

Lichen functional groups indicating tolerance to nitrogen pollution were highly correlated with the ordinations in both survey methodologies (Table 3). Individual correlations of CWM of lichen functional and climate and pollution variables show that US and EU methodologies give the same response trend, although with slight strength differences (Table 4). The sole exception was related with the relationship between CWM of nitrophilic species and relative humidity, which was only significant with the US method. In both methodologies, functional groups seem to reflect a stronger response to climate than to pollution.

Table 4. Summary of Pearson correlations between climate, pollution and functional variables (N  = 28)

Type	Variable	US			EU (macro)			EU
		Nitro	Meso	Oligo	Nitro	Meso	Oligo	Nitro	Meso	Oligo
Climate	E	−0·80*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·71*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·95*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·72*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·54** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·95*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·69*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·58** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·82*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.
	T	0·79*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·62*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·90*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·73*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·47* Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·93*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·71*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·51** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·82*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.
	P	−0·43* Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·15	0·25	−0·58** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·30	0·32	−0·57** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·28	0·30
	RH	0·41* Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·79*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·64*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·20	0·76*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·61** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·14	0·79*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·55** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.
Pollution	N	0·74*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·04	−0·62*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·74*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·10	−0·60** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	0·79*** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.	−0·08	−0·56** Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.

• Climate variables: T , annual mean temperature; P , annual precipitation; E , elevation; RH , relative humidity. Pollution variable: N , dry weight N concentrations in lichens. Functional diversity variables: Nitro, CWM of nitrophilic species; Meso, CWM of mesophilic species; Oligo, CWM of oligophilic species. Significant correlations are marked: *P  < 0·05, **P  < 0·01, ***P  < 0·001.

Discussion

For the first time, we show that the interpretation of lichen community data collected with the two most widely used survey methodologies in North America and Europe are highly comparable, using broadly accepted metrics of biodiversity to measure the effects of global change drivers on ecosystems. Results showed that despite the differences in the absolute values obtained by each method, their interpretation in terms of quality and level of change is the same, meaning that both methodologies depicted the same impact across the environmental gradients studied. These findings are very promising, and ‘open the door’ to common analysis of the outcomes of data sets from both continents to assess temporal and spatial global trends in lichen diversity.

Comparability of different surrogates of biodiversity metrics using different methodologies

Taxonomic diversity measures, species richness, Shannon‐Wiener and Simpson's indices at individual survey sites were highly correlated, depicting the same trends across the study area and allowing a potential future inter‐calibration. As we expected, the different absolute values obtained with both methods were due to differences in the sampling methodologies resulting from the detail paid to inspect lichens and to the different scales used to rate their abundances (point‐based or ocular‐based). A previous work with vegetation identified similar differences in terms of species richness, which, like in our work, still significantly correlated between methods (Godínez‐Alvarez et al . 2009). In works addressing the effects of global change, data intend to represent a surrogate of the biodiversity metrics, rather than exactly the same absolute value. Relativizing values by the maximum of each data set may be one way to overcome such differences. Despite this, both methods consistently described the same trends in response to climate, and were both unresponsive to this pollution gradient. The unresponsiveness of taxonomic diversity (species richness) to nitrogen pollution has been previously recognized (Pinho et al . 2014). Pielou's evenness index was the only index weakly related between US and EU data sets and with no relation to the environmental variables. This index measure can be problematic as it is based on the ratio of the sample‐size independent Shannon‐Wiener index and species richness, a more variable measure that is strongly dependent on sample size (Hurlbert 1971; Mccune, Grace & Urban 2002). We do not recommend relying on Pielou's evenness measures for joint analyses of data sets derived from both protocols. However, we note that the Shannon–Wiener index is in part a measure of evenness and can be used to partly overcome this limitation (Mccune, Grace & Urban 2002).

The assessment of biodiversity shifts using community composition provided by both methods yielded even more striking similarities. The cluster analysis assigned the same sites to the same three groups independent of survey methodology, which correspond spatially to major biogeographical and climate/pollution groups (Fig. 1), showing that methods give the same quality of interpretation. The community composition from both data sets also described the same trends along pollutions and climate gradients. In fact, the nearly similar scores of site classification along these environmental gradients obtained with both methods showed that they depict the same level of change. A work comparing different methodologies to sample dragonflies as indicators of environmental quality found contrasting results (Giugliano, Hardersen & Santini 2012). These were attributed to the fact that they sample different life stages with different ecological requirements, leading the authors to advice against its interchangeable use. This highlights the importance of our results under a global change perspective, and the potential of lichens as ecological indicators. Shifts in species composition at local, regional and global scales in response to global change drivers have been observed (Dornelas et al . 2014; Savage & Vellend, 2015), and these results stress the benefits that can be derived from collectively analyzing the data from both lichen methodologies henceforth.

The community weighted mean of functional groups related to nitrogen deposition at each site was similar across data sets and trended similarly along the nitrogen gradient. These results are very important since these functional groups have been a key concept in the assessment and interpretation of community response to environmental change (Lavorel et al . 2011), with regard to both air pollution and climate in both continents (Cape et al . 2009; Pinho et al . 2009; Pardo et al . 2011; Matos et al . 2015; Root et al . 2015). As it is not linked to species identity, functional diversity can be potentially more universal when applied across broad geographical scales (Branquinho, Matos & Pinho 2015), illustrating at the same time the community structure. Our work reinforces this and shows that it can be a new way to explore and jointly analyse cross‐continental data from distinct survey methods. Due to the low nitrogen pollution rate, nitrophilic species responded more strongly to climate than to nitrogen pollution. Previous works under mixed pollution and climate gradients were successful in extracting independent indicators (Geiser & Neitlich 2007; Root et al . 2015). Given that in our work no independent axes could be extracted for climate and pollution, our results suggest that under a low pollution gradient like ours, functional indicators independent response to climate or pollution may be less clear. This emphasizes the need to have future works that focus on developing strategies to overcome this limitation, as this is likely to happen frequently under a global change context. While species indicator lists could be a way to disentangle environmental factors, its regional context dependency may prevent its use within this comparative framework at a global scale. The solution may pass by developing new ways to disentangle the independent responses of functional indicators, or by developing new metrics based on these species indicator lists in a comparable way, using data from both methodologies.

Specificities and caveats of each methodology

The detail paid in inspecting the microlichen community appears to compensate for the lower number of trees measured using the EU method, thus resulting in higher species counts. Further, the different measures of species abundance (US‐abundance ratings follow a log‐like scale with only four choices; EU‐allows a finer, more precise quantification of frequency but fewer trees sampled) account for the differences in Shannon and Simpson's indices. In general, macrolichens alone provided similar interpretation about trends along climate and pollution gradients as macrolichens plus microlichens. This was previously shown in Europe for land use and climate gradients, where macrolichen diversity alone, when compared to total diversity assessed using the EU method, was able to give the same response (Bergamini et al . 2005, 2007). Our results show that this is also true for climate and pollution. In this work, macrolichens accounted for around 50% of the total number of species found using the EU method, or 31% more species than the US macrolichens only method. Incorporating microlichens might be an advantage of the EU method in climates where macrolichens are less abundant than microlichens, like for example, closed tropical forests (Koch et al . 2013) or dry Mediterranean woodlands (Matos et al . 2015). These advantages can be weighted against extra costs incurred for field observer training and laboratory identification of the microlichens (Bergamini et al . 2007). These two different sets of the community sampled with each method, also related with different strengths to precipitation and relative humidity, despite the almost overlapping sites scores. The US method responded stronger to precipitation, while the EU method responds stronger to relative humidity. The differential sensitivity of different growth forms to water sources is already known (Gauslaa 2014) and these slight differences may be also explained by this.

The US method does not expect field observers to examine every tree, only to survey the range of substrates and microhabitats on the plot with a two‐hour time constraint. The EU survey protocol calls for a complete examination of a standard number of grid cells, using a limited number of phorophytes species (in this case a maximum of two per site), on a variable number of trees (usually between 4 and 10, determined by the environmental problem) and is not time constrained. Given the practical time restraint to keep up with the US field crews during this study, the average number of trees tallied by the EU method observer was six. Therefore, if anything, our work may underestimate the comparability of the data generated by the US and EU survey methods.

Towards a global scale ecological indicator

This work shows that, despite very different protocols and different absolute values, lichen survey data from US and EU methodologies can yield highly similar interpretation and depict the same level of change regarding taxonomic and functional biodiversity metrics and community shifts along two major drivers of environmental change (nitrogen deposition and climate) and can be analysed jointly. A global biodiversity monitoring network is being developed with the purpose of assessing the impact of global change drivers on ecosystems (Pereira & Cooper 2006; Scholes et al . 2012). To enable retrospective and future analysis, the inclusion in this monitoring network demands that data collection is based on compatible standardized methodologies performed at regional and global scales, and that responsive metrics are used to analyse the resulting data (Mace et al . 2005; MEA, 2005, Pereira & Cooper 2006). However, this is only possible if standardized methodologies are applied at world‐wide spatial and long‐term temporal scales. We recommend that future cross‐continental comparisons of US and EU data sets to track global change effects should be done using spatial and temporal trends of these three metrics, as suggested in our conceptual framework in Fig. 4. Prior to these comparisons, EU lichen frequency values should be relativized. Although not essential, log transforming these relativized frequencies may improve data comparability. After this, data from community sifts and functional diversity (namely community weighted mean) may be readily compared under spatial and temporal trend analysis, as illustrated in the conceptual framework (Fig. 4). Regarding taxonomic diversity, the differences found in absolute values stress the need for a calibration to check the equivalence between methods. For that, works similar to this under different environmental gradients and contemplating different ecosystems are necessary and should be planned. Meanwhile, we suggest that absolute values from taxonomic diversity obtained at each site should be relativized by the maximum value found along the studied environmental gradient for each method data set, as recommended in the framework (Fig. 4). Afterwards, results can be compared under a spatial or temporal perspective along these environmental gradients ensuring that the focus of the analysis is on the trend, quality and intensity of change along those gradients. Our results indicate that Pielou's evenness index should be excluded from these analyses.

Figure 4

Conceptual framework for joint analysis of US and EU lichen diversity data sets for tracking global change.

The US and EU methods are used at a continental scale (North America and Europe) and could be used for global trend analysis at these continental scales. This framework confirms these methods compatibility and takes us one step closer towards lichens inclusion in this monitoring network. This framework integrates measures of taxonomic diversity, community shifts and functional diversity and these are essential to more thoroughly comprehend ecosystems response to global change (Thuiller et al . 2006; Mcgill et al . 2015). The spatial patterns and trends obtained with these metrics across an environmental gradient supported the design of this framework and efficiently depicted the same quality and intensity of change. Henceforth, this framework will enable to detect analogous trends and patterns of response to track global change effects over time, emphasizing and potentiating lichens as large‐scale integrated ecological indicators of global change.

Our study covered a limited environmental range in a global change perspective, and a limited number of ecosystems. Nonetheless, because this gradient also encompassed sites with different historical environmental change (more pristine sites in the wilderness plots, and sites under more anthropogenic influence), and because this framework was developed to address the quality, intensity and trend of change, we expect these results to hold true under environmental conditions different than these. Nonetheless, future work should explore additional biomes and larger environmental gradients in terms of climate, pollution and other drivers of global change, to strengthen the robustness of the comparison and calibration between US and EU methodologies.

Authors’ contributions

P.M., L.G., A.M.V.M.S. and C.B. conceived the ideas and designed methodology; P.M., L.G., A.H. and D.G. collected the data; P.M., L.G., A.N., P.P. and C.B. analysed the data; P.M., L.G. and C.B. led the writing of the manuscript. All authors contributed critically to the drafts and gave final approval for publication.

Acknowledgements

We thank the US Forest Service International Volunteer Program for facilitating this collaboration, and Bruce McCune for advice on identification and analysis. This work was supported by the USFS Pacific Northwest Region Air Resource Management Program and FCT‐MEC through: project PTDC/AAG‐GLO/0045/2014; P.M. by contract SFRH/BD/51419/2011; P.P. by contract SFRH/BPD/75425/2010; A.N. by contract SFRH/BD/51407/2011; C.B. by contract Investigador FCT.

Data accessibility

Data deposited in the Dryad repository: https://doi.org/10.5061/dryad.86h2k (Matos et al . 2016).