Molecular Ecology (2005) 14, 457– 467
doi: 10.1111/j.1365-294X.2004.02423.x
Microsatellites reveal regional population differentiation and isolation in Lobaria pulmonaria, an epiphytic lichen
Blackwell Publishing, Ltd.
JEAN-CLAUDE WALSER, ROLF HOLDEREGGER, FELIX GUGERLI, SUSAN EVA HOEBEE and C H R I S T O P H S C H E I D E G G E R Division of Ecological Genetics, Swiss Federal Research Institute WSL, Züercherstrasse 111, CH-8903 Birmensdorf, Switzerland
Abstract Many lichen species produce both sexual and asexual propagules, but, aside from being minute, these diaspores lack special adaptations for long-distance dispersal. So far, molecular studies have not directly addressed isolation and genetic differentiation of lichen populations, both being affected by gene flow, at a regional scale. We used six mycobiontspecific microsatellite loci to investigate the population genetic structure of the epiphytic lichen Lobaria pulmonaria in two regions that strongly differed with respect to anthropogenic impact. In British Columbia, L. pulmonaria grows in continuous old-growth forests, while its populations in the old cultural landscape of Switzerland are comparably small and fragmented. Populations from both British Columbia and Switzerland were genetically diverse at the loci. Geographically restricted alleles, low historical gene flow, and analyses of genetic distance (UPGMA tree) and of differentiation (AMOVA) indicated that populations from Vancouver Island and from the Canadian mainland were separated from each other, except for one, geographically intermediate population. This differentiation was attributed to different glacial and postglacial histories of coastal and inland populations in British Columbia. In contrast to expectations, the three investigated Swiss populations were genetically neither isolated nor differentiated from each other despite the long-lasting negative human impact on the lichen’s range size in Central Europe. We propose that detailed studies integrating local landscape and regional scales are now needed to understand the processes of dispersal and gene flow in lichens. Keywords: dispersal, glaciation, genetic diversity, isolation by distance, lichen-forming fungi, population history Received 13 July 2004; revision received 27 October 2004; accepted 2 November 2004
Introduction Approximately 50% of all fungi obtain nutrients by living in close association with other organisms (Honegger 1996; Tunlid & Talbot 2002). One conspicuous example is lichenization, defined as a mutualistic symbiosis between a fungus (mycobiont) and at least one algal and /or cyanobacterial species (photobiont; Hawksworth & Honegger 1994). Lichens dominate approximately 8% of the world’s terrestrial ecosystems, and more than 20% of all fungal species are lichenized (Hawksworth et al. 1995). A recent study of evolutionary relationships in fungi revealed that several nonlichenized fungi, including plant or human Correspondence: J.-C. Walser, Tel.: (773) 834 0467; Fax: (773) 702 0037; E-mail:
[email protected] © 2005 Blackwell Publishing Ltd
pathogens, have lichen-forming ancestors (Lutzoni et al. 2001). This suggests that mutualistic, coevolutionary systems are not necessarily evolutionary dead ends, and that lichenization is probably an ancient nutritional strategy. Many regional lichen floras include endemic taxa, but most lichen species have much broader, though often scattered, geographical distributional ranges than vascular plants (Galloway 1996). It is not clear, however, whether such biogeographical patterns reflect long-distance dispersal or historic fragmentation. Except for being minute, sexually and asexually produced lichen propagules lack special morphological adaptations for long-distance dispersal (Heinken 1999; Dettki et al. 2000; Sillett et al. 2000). Therefore, it has been assumed that lichens are often dispersal-limited (e.g. Bailey 1976; Armstrong 1990). Kärnefelt (1990), suggested that a slow evolutionary or
458 J . - C . W A L S E R E T A L . speciation rate, rather than a vast dispersal capacity, could explain why lichens exhibit morphological uniformity over large geographical areas. Recent molecular studies (Walser et al. 2001; Printzen et al. 2003; Walser 2004) imply that the dispersal of at least vegetative propagules could well be limited at a landscape level, but little is known about the effectiveness of lichen dispersal and, thus, gene flow by means of either sexual or asexual propagules over larger geographical distances within regions. Molecular studies could help to understand population history, genetic differentiation or isolation, and gene exchange among lichen populations. The paucity of this type of information in the literature can be attributed to a previous lack of suitable genetic markers, but Walser et al. (2003) recently illustrated the use of mycobiont-specific microsatellites in population genetic analyses of lichens at different spatial scales. They also showed that the within-population diversity was much higher than had been suggested by earlier studies mainly based on nuclear ribosomal DNA (Bridge & Hawksworth 1998; Zoller et al. 1999; Kroken & Taylor 2001). We investigated the genetic variation at six microsatellite loci in the mycobiont of the epiphytic lichen Lobaria pulmonaria from two regions, namely, British Columbia (Canada) and Switzerland. These regions are drastically different with respect to the availability of suitable habitats for L. pulmonaria, owing to anthropogenic modification of the landscape. Populations of L. pulmonaria in British Columbia grow in undisturbed, continuous old-growth forests, whereas the species’ habitats in Switzerland are fragmented, and populations are small due to long-term forest management and dramatic environmental changes during the 20th century (Wirth et al. 1996; Zoller et al. 1999). As a consequence, patterns of population differentiation and of gene exchange among populations in British Columbia should differ from those in Switzerland. We hypothesize that populations of L. pulmonaria from British Columbia are characterized by high genetic variation and low differentiation indicating abundant historical gene flow among populations. However, because human impact has likely led to substantially higher genetic differentiation of populations in Switzerland, owing to smaller population sizes and random sampling effects such as genetic drift (Hartl & Clark 1997), we also hypothesize that Swiss populations are genetically less diverse than those from British Columbia.
Materials and methods
ascomycete of the order Lecanorales, suborder Peltigerineae (Tehler 1996). Its primary photosymbiotic partner is the eukaryotic green alga Dictyochloropsis reticulata (Geitler 1966), and its second partner is the nitrogen-fixing cyanobacterium Nostoc sp. ( Jordan 1970). The species is known to reproduce both sexually and asexually ( Jordan 1973; Denison 2003). However, the sexual cycle of the photobiont is suppressed, and only the mycobiont goes through sexual reproduction forming ascospores (Malachowski et al. 1980). A recent study found evidence for recombination, which indicates that L. pulmonaria is an outcrossing lichen (Walser et al. 2004). In addition to sexually derived propagules, L. pulmonaria also forms different types of asexual (symbiotic) dispersal units such as soredia, isidioid soredia, and thallus fragments (Scheidegger 1995; Büdel & Scheidegger 1996). While L. pulmonaria has a large distribution and is still widespread and locally common in boreal North America (Brodo et al. 2001), it is considered endangered in many parts of Central Europe (Wirth et al. 1996). For example, L. pulmonaria has almost completely vanished from the Swiss Plateau, and the remaining populations in the Swiss Pre-Alps and Jura Mountains have become increasingly small and fragmented (Scheidegger et al. 2002).
Sampling We sampled populations of L. pulmonaria in two regions: in British Columbia, Canada, where the species is common, and in Switzerland, where it is rare. A population was considered as a spatially distinct patch of trees colonized by L. pulmonaria. Beyond the patch perimeter, L. pulmonaria did locally not occur. In total, 565 thalli were sampled from nine populations in British Columbia and from three populations in Switzerland (Table 1). British Columbia is an ecologically diverse region with several biotic zones. Goward (1999) divided the province into four ‘life zones’. Assuming that genetic differentiation might be more pronounced between populations from different geoclimatic regions (i.e. life zones), we collected the samples from British Columbia along a transect from the south coast of Vancouver Island to the interior of the province within three of these life zones: hypermaritime, maritime, and intermontane (Table 1, Fig. 1a). The samples from Switzerland were collected from relatively large populations in the Pre-Alps and the Jura Mountains harbouring at least 30 colonized trees (Table 1). One thallus per tree was randomly taken across each population.
Species The foliose macrolichen Lobaria pulmonaria (L.) Hoffm. predominately grows epiphytically (Brodo et al. 2001) and has an estimated generation time of more than 30 years (Scheidegger et al. 1998). It is known to be a tripartite lichen species, with the fungal component (mycobiont) being an
Microsatellite analysis Owing to its tripartite nature, as well as the difficulty in obtaining algal-free thallus material, anonymous DNA fingerprinting, such as RAPDs, is not applicable to population genetic questions in L. pulmonaria. Instead, we used © 2005 Blackwell Publishing Ltd, Molecular Ecology, 14, 457– 467
I S O L A T I O N A N D D I F F E R E N T I A T I O N O F L I C H E N P O P U L A T I O N S 459 Table 1 Locations and codes of 12 investigated populations of Lobaria pulmonaria from British Columbia (Canada) and Switzerland, and their grouping in a set of five hierarchical amovas (Table 4). Populations sharing the same letter were grouped together Grouping of populations† Location
Code
Biogeoclimatic zones
British Columbia Ayum Creek, Vancouver Island Chesterman Beach, Tofino Cape Scott Provincial Park Lakelse Lake Provincial Park Date Creek, Kispiox Clayton Falls, Bella Coola Carp Lake Provincial Park Bowron Lake Provincial Park Oregana Creek, Tumtum Lake
AY TO CS PR DC BC CL BL OC
hypermaritime hypermaritime hypermaritime maritime maritime maritime intermontane intermontane intermontane
Switzerland Taaren Wald, Toggenburg Murgtal, Walensee Marchairuz, Jura Mountains
TW MT UZ
Pre-Alps Pre-Alps Jura Mountains
Altitude a.s.l.* (m)
Latitude
Longitude
(a)
(b)
(c)
(d)
(e)
15 5 5 105 550 5 860 910 723
48°23’29’’ N 49°06’47’’ N 50°40’27’’ N 54°22’54’’ N 55°24’52’’ N 52°22’12’’ N 54°52’08’’ N 53°15’19’’ N 51°59’08’’ N
123°39’37’‘ W 125°53’31’‘ W 128°16’32’‘ W 128°31’52’‘ W 127°48’52’‘ W 126°48’49’‘ W 123°15’39’‘ W 121°21’03’‘ W 119°05’21’‘ W
A A A B B B C C C
D D D E E D E E E
F F F G G G G G G
H H H I I K I I I
— — — — — — — — —
1350 1280 1200
47°10’50’’ N 47°03’52’’ N 46°29’57’’ N
9°18’15’‘ E 9°11’51’‘ E 6°10’21’‘ E
— — —
— — —
— — —
— — —
L L L
*a.s.l.: above sea level; †grouping of populations: (a) among three life zones in British Columbia; (b) British Columbia coast vs. interior; (c) Vancouver Island vs. mainland; (d) Vancouver Island vs. population BC against inland; (e) within Switzerland.
Fig. 1 Estimated number of migrants (Nm) and genetic relationship among populations of the epiphytic lichen species Lobaria pulmonaria from British Columbia (Canada) and Switzerland. (a) Location of the investigated populations in British Columbia. Different symbols refer to different life zones (circles: hypermaritime; squares: maritime; triangles intermontane). Populations with altitude < 50 m above sea level (a.s.l.) are given in white (coast) and populations > 50 m a.s.l. in black (interior). The number of migrants among populations based on private alleles is symbolized by the style of the lines (Nm-values < 1 are not given). For population codes see Table 1. (b) upgma cluster of the Canadian and Swiss populations, based on genetic distance DCE of six microsatellite loci. The values beside the nodes represent bootstrap support for genetic distance DCE and DB, respectively, over 1000 permutations. The vertical bars refer to genetic lineages from Switzerland (grey), British Columbia coast (see above; white) and British Columbia interior ( black).
© 2005 Blackwell Publishing Ltd, Molecular Ecology, 14, 457–467
460 J . - C . W A L S E R E T A L . six microsatellite markers specific to the mycobiont of L. pulmonaria (Walser et al. 2003). DNA was isolated according to Walser et al. (2003) from approximately 20 mg of apical thallus material. The multilocus genotypes were based on data obtained from five dinucleotide (LPu03, LPu15, LPu16, LPu20, LPu27) and one nine-base-pair repeat (LPu09) microsatellite loci, which were amplified in multiplex-PCR following the protocols of Walser et al. (2004). The fungal partner of the lichen is haploid and thus has only one allele per locus. In this and in an earlier study, no evidence for intrathalline variation was found (Walser et al. 2003).
Data analysis Recurrent multilocus genotypes can either be the result of vegetative reproduction or selfing in homothallic lichen species or chance products of sexual reproduction. If asexual reproduction is abundant, shared multilocus genotypes among populations could be interpreted as long-distance dispersal of asexual propagules. On the other hand, individuals within or among populations with the same multilocus genotype could also be generated by independent sexual events. Within populations, the probability of detecting sexually produced individuals with the same multilocus genotype (Pse) was calculated as the product of the given genotype’s allele frequencies according to Parks & Werth (1993) and Wang et al. (1997). The proportion of different multilocus genotypes (M), genotypic diversity (number of observed genotypes divided by the number of samples, Go /N; Stoddart & Taylor 1988), the mean number of alleles (A), and the mean effective number of alleles (Ae) were estimated using the free statistics software r version 1.8.1 (R Development Core Team 2003; http://www.R-project.org). Two distance estimators were used to assess different evolutionary assumptions. First, genetic distances among populations (without recurring genotypes) were estimated using the chord distance DCD (Cavalli-Sforza & Edwards 1967). Additional to this measure based on the infinite allele model (IAM), we also applied an estimate based on the stepwise mutation model (SMM), namely, DB (Bowcock et al. 1994), i.e. the proportion of shared allele distances. Genetic distance estimations were carried out using microsatellite analyser 2.65 (Dieringer & Schlötterer 2003) and msatbootstrap 1.1 (Landry et al. 2002). The resulting genetic distance matrices were compared by creating an unweighted pair– group method with arithmetic mean (upgma) tree with confidence estimates assigned to its topology based on 1000 bootstrap replicates. The upgma dendrograms were constructed using the neighbour and consense components of phylip 3.57c (Felsenstein 1989). Differentiation among populations (without recurring genotypes) was estimated with hierarchical analysis of molecular variance (amova) using both RST (SMM) and FST
(IAM) values in arlequin version 2.000 (Schneider et al. 2000). The populations were assigned to different groups: (a) three life zones in British Columbia, (b) British Columbia coast (altitude < 50 m above sea level) vs. interior (all others), (c) Vancouver Island against mainland, (d) Vancouver Island vs. population B C vs. interior, and (e) within Switzerland (Table 1). Mantel tests (Mantel 1967), with 1000 permutations, were conducted using the statistics software r 1.8.1 (R Development Core Team 2003) to determine the correlation between geographical distance and genetic differentiation for populations from British Columbia. According to Nybom et al. (2004), we used both RST and FST estimates. As Raybould et al. (1998) suggested, separate Mantel tests for populations less than 500 km apart were additionally performed. Pairwise multilocus estimates of the effective number of migrants (Nm) for populations from British Columbia and from Switzerland based on private alleles (Slatkin 1985; Barton & Slatkin 1986) were computed using genepop 3.1c (http://wbiomed.curtin.edu.au/genepop/), and are, thus, independent from FST estimates. The results were adjusted since there are only half the numbers of migrant genes for haploid data. We conducted a second Mantel test with 1000 permutations to determine the relationship between geographical distances and number of migrants (Nm) in British Columbia populations using r (R Development Core Team 2003).
Results Genetic and genotypic diversity Three thalli were excluded from the data set because of incomplete genotype assessment. The number of alleles per locus within populations ranged from one to a maximum of 22, and, on average, between 2.3 and 12.4 alleles were found per locus across the 12 populations (Table 2). The total number of different alleles per locus ranged from three to 31 in populations from British Columbia and from four to 20 in Swiss populations (Table 2). In populations from British Columbia, mean allele sizes at loci LPu20 and LPu27 were generally shorter than in Swiss populations, but longer at locus LPu09 and LPu15 (Fig. 2). Furthermore, and in contrast to all the other loci, the allele size range at locus LPu27 did not overlap between samples from the two continents (Fig. 2). Some alleles were geographically confined to a single or to only few populations within British Columbia (Fig. 2). At locus LPu03, which showed little variation (Table 2), the shortest allele of 187 bp was exclusively found in the coastal populations AY (Ayum Creek, Vancouver Island), CS (Cape Scott Provincial Park), and BC (Clayton Falls, Bella Coola). Populations PR (Lakelse Lake Provincial Park), DC (Date Creek, Kispiox), CL (Carp Lake Provincial Park), BL (Bowron Lake © 2005 Blackwell Publishing Ltd, Molecular Ecology, 14, 457– 467
I S O L A T I O N A N D D I F F E R E N T I A T I O N O F L I C H E N P O P U L A T I O N S 461
British Columbia
Switzerland
Loci
Min
Max
Mean ± SE
Total
Min
Max
Mean ± SE
Total
LPu03 LPu09 LPu15 LPu16 LPu20 LPu27
2 7 6 9 5 1
3 17 9 15 22 4
2.3 ± 0.1 11.3 ± 1.1 7.2 ± 0.4 11.6 ± 0.7 12.4 ± 1.7 2.7 ± 0.3
3 28 14 23 31 7
3 5 5 4 11 2
4 9 9 9 14 6
3.3 ± 0.3 7.0 ± 1.2 6.7 ± 1.2 6.7 ± 1.5 12.0 ± 1.0 4.3 ± 1.2
4 13 10 10 20 7
Table 2 Allele numbers at six microsatellite loci in Lobaria pulmonaria from British Columbia (Canada) and Switzerland. Minimum (Min) and maximum number (Max), mean (Mean), standard error (SE) of number of alleles per population and the total number of alleles (Total) over all populations from British Columbia and Switzerland, respectively
Fig. 2 Box-plots of the allele size distributions of five microsatellite loci (LPu09, LPu15, LPu16, LPu20, LPu27) in Lobaria pulmonaria populations from British Columbia (Canada) and Switzerland. Locus LPu03 had a maximum of only four different alleles and is not included in the figure. Circles denote outliers. For population codes see Table 1.
Pronvincial Park), and OC (Oregana Creek, Tumtum Lake) from the interior of British Columbia exhibited high frequencies of the 167 bp allele at locus LPu20. This allele was rare in the coastal population BC and absent in the populations from Vancouver Island AY, TO, and CS (Fig. 2). At the same locus, alleles equal to or shorter than 165 bp were only found in the populations from Vancouver © 2005 Blackwell Publishing Ltd, Molecular Ecology, 14, 457–467
Island and in the coastal population BC. Similarly, populations originating from the coast (AY, TO, CS, BC) displayed an allele of 180 bp at locus LPu27, which was absent from the interior populations (Fig. 2). In addition, there were no alleles specific to any of the three life zones investigated. A high number of those alleles that were only found once in single individuals were detected at the nine-
462 J . - C . W A L S E R E T A L .
Population
N
M
A
Ae
Go / N
Pse
British Columbia AY TO CS PR DC BC CL BL OC
39 47 50 50 52 50 51 49 52
0.79 0.62 0.52 0.90 0.92 0.90 0.53 0.90 0.92
7.8 7.2 7.7 8.7 7.3 10.5 6.3 7.2 8.7
5.1 4.3 4.8 4.4 3.4 5.7 3.1 3.3 4.2
0.62 0.26 0.26 0.83 0.87 0.83 0.28 0.78 0.84
96.8% 96.6% 100.0% 80.0% 60.4% 97.8% 48.1% 47.7% 75.0%
Switzerland TW MT UZ
52 38 32
0.52 0.47 0.53
7.8 5.3 6.8
4.4 3.6 3.8
0.26 0.26 0.30
96.3% 77.8% 88.2%
British Columbia mean Switzerland mean Total mean
49 ± 1.3 41 ± 5.9 47 ± 1.9
0.78 ± 0.06 0.51 ± 0.02 0.71 ± 0.06
7.9 ± 0.4 6.6 ± 0.7 7.6 ± 0.4
4.3 ± 0.3 3.9 ± 0.2 4.2 ± 0.2
0.62 ± 0.09 0.27 ± 0.01 0.53 ± 0.08
base-pair repeat locus LPu09 (Fig. 2), while locus LPu03 had no such unique alleles. The proportion of different multilocus genotypes sampled per population (M) varied between 0.52 and 0.92 in British Columbia with a mean (± SE) of 0.78 ± 0.06 (Table 3). The three Swiss populations revealed lower values between 0.47 and 0.53, similar to populations CL and CS from British Columbia, and a mean of 0.51 ± 0.02. The Canadian population BC harboured the highest mean and mean effective number of alleles ( Table 3). It also shared characteristic alleles with populations from Vancouver Island and the interior of British Columbia. In general, populations from Switzerland and British Columbia exhibited similar values for the mean and the mean effective number of alleles ( Table 3), while size-corrected genotypic diversity (Go /N) was comparatively low in the Swiss populations. Only three populations from British Columbia exhibited similarly low values (Go /N < 0.3; Table 3). Most populations from the maritime zone and the intermontane zone of British Columbia, as well as population AY from Vancouver Island, had substantially higher values of genotypic diversity (Go /N > 0.6; Table 3). The 440 samples from British Columbia revealed 343 different multilocus genotypes (78%) and 296 singleoccurrence genotypes. For most genotypes, the probability of a second encounter as a result of sexual reproduction within a population was smaller than 5% (Table 3). This indicates that recurrent genotypes at a given location were most likely the product of asexual propagation. The 122 samples from the three Swiss populations harboured 62 distinct multilocus genotypes (51%) of which 41 occurred only once. In accordance to the results presented for British Columbia, the probability of a second encounter of a given genotype resulting from sexual reproduction within
Table 3 Number of samples (N), proportion of different multilocus genotypes (M), mean number of alleles (A), mean effective number of alleles (Ae), sample size-corrected genotypic diversity (Go / N), and the proportion of multilocus genotypes which had a lower probability than Pse < 0.05 to be found twice as the result of sexual reproduction within a given population of Lobaria pulmonaria from British Columbia (Canada) and Switzerland. Means and standard errors (SE) are given for British Columbia, Switzerland and the total data set. For population codes see Table 1
— — —
Table 4 Correlation between geographical distances and genetic differentiation (RST, FST) among populations of Lobaria pulmonaria from British Columbia (Canada) using Mantel tests. ns = not significant Mantel test Geographic distance
Genetic differentiation*
rm
P-value
All All < 500 km < 500 km
RST FST RST FST
0.572 − 0.400 0.197 0.392
< 0.001 ns 0.003 < 0.001
*RST: Slatkin (1995), FST: Wright (1951).
the Swiss population was mostly less than 5% (Table 3). The Swiss populations did not share a common genotype.
Genetic differentiation There was substantial genetic differentiation among the populations based on RST and on FST estimates for both British Columbia and Switzerland. Most of the pairwise comparisons of populations were significantly different from zero; only nine of the RST estimates (14%) and two of the FST estimates (3%) were not significant. RST estimates were almost twice as high as FST values (data not shown). Mantel tests between pairwise geographical distance and genetic differentiation for populations from British Columbia (Table 4) only showed a significantly positive relationship when based on RST values (P < 0.001). For FST values, such a positive relationship was only found at shorter distances of less than 500 km (Table 4). Hence, the two measurements © 2005 Blackwell Publishing Ltd, Molecular Ecology, 14, 457– 467
I S O L A T I O N A N D D I F F E R E N T I A T I O N O F L I C H E N P O P U L A T I O N S 463 Table 5 Results of analyses of molecular variance (amovas) with six microsatellite loci of Lobaria pulmonaria from nine populations from British Columbia (Canada) and three populations from Switzerland based on RST- and FST-values. Recurring genotypes were excluded from the analysis. For different groupings of populations see Table 1. Significance levels were based on 1000 permutations. Variance components values: among groups = RCT/FCT, among populations within groups = RSC/FSC, among populations = RST/FST RST
Grouping of populations
Source of variation
d.f.
(a) Among life zones in British Columbia
Among groups Among populations within groups Within populations Among populations
2 6 334
Among groups Among populations within groups Within populations Among populations
1 7 334
Among groups Among populations within groups Within populations Among populations
1 7 334
Among groups Among populations within groups Within populations Among populations
2 6 334
(b) British Columbia coast vs. British Columbia interior
(c) Vancouver Island vs. mainland
(d) Vancouver Island vs. population BC vs. interior
(e) Within Switzerland
Among populations Within populations
of genetic differentiation led to conflicting results when geographical distances were greater than 500 km. The estimates of the effective number of migrants per generation (Nm) based on private alleles ranged between 0.42 and 2.28. Within British Columbia, Nm values among populations from the maritime and intermontane zone were higher compared with those among populations from Vancouver Island and the mainland (Fig. 1a). Particularly high numbers of migrants were estimated for population BC and all other populations (Fig. 1a). The Mantel test showed a significant negative correlation (rm = −0.64, P < 0.001) between geographical distance and Nm values indicating increased population isolation with increasing distance in British Columbia. The mean Nm value for the three Swiss populations was below 0.9 (data not shown). The upgma dendrogram based on DCE clearly separated the Lobaria pulmonaria populations from British Columbia from those of Switzerland (Fig. 1b). The upgma tree based on DB had an identical topology (Fig. 1b). In British Columbia, there was also a well-supported genetic divergence between populations from the coast and the interior (Fig. 1b). When the populations from British Columbia were assigned to three different life zones ( Table 1, Fig. 1a), most of the total variance resided within populations, and a comparably small proportion of the genetic variance was found among the life zones (RCT: 7%, P = 0.081; FCT: 4%, P = © 2005 Blackwell Publishing Ltd, Molecular Ecology, 14, 457–467
2 59
Variance component 6.74% 8.81% 84.45% 15.55% 15.49% 4.76% 79.75% 20.25% 11.84% 8.06% 80.10% 19.90% 13.34% 5.19% 81.46% 18.53% −3.94% 103.94%
FST Variance component
P 0.081 < 0.001 < 0.001 0.007 < 0.001 < 0.001 0.023 < 0.001 < 0.001 0.015 < 0.001 < 0.001 0.943
4.17% 3.80% 92.03% 7.97% 6.94% 3.03% 90.03% 9.97% 7.29% 3.57% 89.14% 10.86% 7.06% 2.47% 90.47% 9.53% 1.24% 98.76%
P < 0.021 < 0.001 < 0.001 0.007 < 0.001 < 0.001 0.016 < 0.001 < 0.001 0.008 < 0.001 < 0.001 0.121
0.021). When the coastal populations were grouped separately from the interior populations, 15% (RCT, P = 0.007) and 7% (FCT, P = 0.007) of the total genetic variance resided between the two groups (Table 5). The genetic variance between groups was again smaller when populations from Vancouver Island (hypermaritime) were compared with mainland populations (maritime and intermontane; RCT = 12%, P = 0.023; FCT = 7%, P = 0.016) or when they were tested against population BC and against all other populations from British Columbia (RCT = 13%, P = 0.015; FCT = 7%, P = 0.008), respectively. The amova for populations from Switzerland revealed that the total genetic variation residing among the three populations was very small and not significantly different from zero (both for RST and FST; Table 5).
Discussion The use of microsatellites is now routine in plant and animal population genetics (Goldstein & Schlötterer 1999). In lichenforming fungi, microsatellite loci have only recently been introduced, and, consequently, few studies are yet available (Walser et al. 2003). We used six microsatellite loci to investigate population differentiation and isolation by distance in the epiphytic lichen Lobaria pulmonaria, and to draw conclusions on its history at regional scales. Genetic analysis
464 J . - C . W A L S E R E T A L . of nine populations of L. pulmonaria from British Columbia and three populations from Switzerland revealed high levels of genetic variation within populations. The microsatellite loci used were sufficient to detect 405 different multilocus genotypes among the 562 analysed lichen thalli. This confirms the high resolving power of microsatellites for determining genotypic diversity in lichen species and their subsequent use in analyses of genetic structure. In this study, populations of L. pulmonaria from British Columbia were found to harbour substantially greater genotypic diversity than Swiss populations.
Genetic patterns in British Columbia Our first premise was that populations of L. pulmonaria in the continuous old-growth forests of British Columbia should be genetically diverse and show low differentiation because of abundant gene flow among them. The Canadian populations indeed exhibited high genetic variation but, in contrast to our expectations, low historical gene flow and the results of the analyses of genetic distance (upgma tree) and differentiation (amova) indicated a distinct genetic geographical structure of L. pulmonaria populations. Specifically, populations from Vancouver Island were clearly separated from the mainland populations, and the mainland population BC, close to the coast, showed an intermediate position. This geographical structure is possibly not caused by anthropogenic factors or geoclimatic zonation, but may be due to postglacial population history (Fig. 1a). In this context, molecular data suggest that the ice ages profoundly influenced the genetic architecture of the flora and fauna of the Pacific Northwest (Soltis et al. 1997). Most plants and animals of British Columbia are descendants of immigrants that colonized the province after the retreat of the Pleistocene ice sheet 10 000 –13 000 bp (Cannings & Cannings 1996). Glaciation, however, did not thoroughly deplete the diversity of the British Columbia biota. Many of the extant taxa survived the ice ages either in one or more ice-free but isolated refugia to the north or south of the province or alternatively on a few peninsulas and offshore islands including the Brooks Peninsula on the northwestern coast of Vancouver Island (Soltis et al. 1997). After glaciation, migration and the mixing of once isolated and possibly genetically differentiated glacial populations (owing to restricted gene flow) resulted in the formation of continuous geographical distributions (Soltis et al. 1997). It has been suggested that the different refugial source areas and the different migration corridors east and west of the Coastal Mountains have led to present-day genetic differentiation of populations from the coast and from the mainland of British Columbia (Cannings & Cannings 1996). Because the sea level was lower (Josenhans et al. 1997) and the coastline elevated by upwarping during glaciation (Benson et al. 1999; Clague & James 2002), the coastal refugia
were probably larger than would be apparent today. In accordance with this Vancouver Island refugium hypothesis, our data revealed that populations from Vancouver Island were genetically differentiated from inland populations, suggesting that populations of L. pulmonaria from the two areas may well have had different glacial and postglacial histories. It seems possible that L. pulmonaria populations from Vancouver Island originating from in-situ surviving populations or from genetic lineages migrating to Vancouver Island from southern coastal refugia, have left their genetic footprints on the current genetic structure. The affiliation of population BC either to populations from Vancouver Island or to populations from the mainland was unclear. The geographically intermediate position of this population could also explain its apparent genetic position midway between island and mainland populations. For example, the relative values of estimated historical gene flow for population BC were particularly high when compared with those of most other populations (Fig. 1a). This result has to be interpreted with caution because of the possible effects of homoplasy. Furthermore, population BC shared geographically restricted alleles with both the coastal and the inland populations and, consequently, showed the highest mean number of alleles per locus of all investigated populations (Table 3). The results suggest that populations from Vancouver Island and from the interior of the continent have influenced the genetic structure of population BC and that this population is situated in the area of a contact zone of postglacial migration routes of L. pulmonaria (cf. Petit et al. 2003). The comparison of genetic and geographical distances in isolation-by-distance tests resulted in conflicting results when populations more than 500 km apart were taken into consideration (Table 4). Whereas the overall Mantel test was only significant for RST values, it was not so for FST values. On the other hand, there was significant isolation by distance for FST values at distances of less than 500 km. This is to be expected in microsatellite analyses (Raybould et al. 1998), reflecting the sensitivity to large genetic distances of the IAM-based FST values where the effect of isolation by distance and that of mutation overlap (Raybould et al. 1998; Balloux & Lugon-Moulin 2002). Distances greater than 500 km mostly occurred between populations from Vancouver Island and the inland regions, where historical gene flow rates were comparatively low (Fig. 1a) so perhaps one can have more confidence, at least in this instance, in the RST estimates of overall genetic differentiation. Population differentiation values are often lower when based on the assumption of the IAM as compared to the SMM, particularly if polymorphism is high and rates of migration are low (Bossart & Prowell 1998; Raybould et al. 1998; Reusch et al. 2000; Collevatti et al. 2001; Balloux & Lugon-Moulin 2002). The IAM-based estimates indicate lower differentiation because they do not distinguish among © 2005 Blackwell Publishing Ltd, Molecular Ecology, 14, 457– 467
I S O L A T I O N A N D D I F F E R E N T I A T I O N O F L I C H E N P O P U L A T I O N S 465 shared alleles in different populations that are not identical by descent (i.e. they arose by independent mutations and are not shared by gene exchange). Similar results are only to be expected when mutation rates are negligible in comparison to migration and drift. Conversely, when stepwise-like mutations contribute to population differentiation, RST values should be larger than FST values (Hardy et al. 2003), except for the case of parallel evolution.
Genetic patterns in Switzerland Our second hypothesis stated that, owing to long-lasting human impact in Central Europe, Swiss populations of L. pulmonaria should be clearly differentiated because of the effects of smaller population sizes and enhanced random sampling (cf. populations from British Columbia). Given that L. pulmonaria suffered a significant decline in Switzerland during the last century and is now considered endangered in lowland regions (Scheidegger et al. 2002), the remaining populations in the Swiss Pre-Alps and Jura Mountains are geographically strongly isolated from each other (Walser 2004). Apart from the fact that the geographical distances among the studied Swiss populations were relatively small (15–244 km) it is likely that high mountains represent substantial natural barriers to dispersal and that current genetic exchange therefore appears improbable. Despite a high degree of clonal reproduction within the Swiss L. pulmonaria populations (Walser et al. 2004), and contrary to our expectations, population diversity was still high and the estimates of pairwise genetic differentiation among populations were insignificant (Table 5). While this might point to a common postglacial history of Swiss populations of L. pulmonaria, it could also indicate substantial among-population gene exchange at least in the past. Concordantly, the low differentiation found in this study suggests that, historically, the populations of L. pulmonaria were connected. That the recent dramatic demographic changes that the species has suffered have not yet result in detectable alterations of its genetic structure in Switzerland might be explained by the longevity of L. pulmonaria individuals.
Gene flow between continents Among the many evolutionary processes acting at the species level, it is still not clear to what extent modern disjunct distribution patterns represent long-distance colonization events (Högberg et al. 2002) and/or contiguous population expansion followed by range contraction and fragmentation (i.e. vicariance; Kärnefelt 1990; Galloway 1996). In lichenized fungi, Kärnefelt (1990) hypothesized that the broad, but often scattered, global distributions characteristic of many lichen species are caused not by long-distance dispersal, but by historical fragmentation of formerly more © 2005 Blackwell Publishing Ltd, Molecular Ecology, 14, 457–467
continuous distribution areas. The nonoverlapping allele size distribution at locus LPu27 (Fig. 2) might be taken as evidence against recent gene flow between the Swiss and the Canadian populations and could point towards past fragmentation. However, further data from the east of North America, eastern Eurasia and western Europe may provide more information on historical gene flow between continents. Furthermore, this is only one level (i.e. the fungal) of investigating diversity and differentiation in lichen populations. Subsequent research on either the algal or cyanobacterial photobiont might then be overlaid to give a complete representation of lichen population structure(s).
Conclusions Earlier studies on genetic differentiation and isolation of lichen populations at regional scales are sparse and limited to investigations of nuclear ribosomal DNA, which permit the detection of only rather low levels of genetic diversity. Our research has extended these studies by using more variable microsatellite markers in the epiphytic lichen species Lobaria pulmonaria. The results of our study showed that populations both from British Columbia and Switzerland were genetically diverse, but that the Swiss populations showed a higher degree of vegetative propagation; and that populations from mainland British Columbia were substantially differentiated from coastal populations, but not genetically isolated. This latter result may be attributable to different glacial and postglacial histories of the populations studied. Finally Swiss populations, in contrast to our expectations, were not genetically isolated from each other despite the strong negative human impact on the lichen’s range size in this Central European region. We suggest that regional population genetic investigations should now be extended by in-depth studies of the dispersal or gene flow processes of lichens integrating local, landscape, regional and intercontinental scales.
Acknowledgements We would like to thank Trevor Goward for accommodation, lively discussions, and field assistance and Andy MacKinnon for accommodation and field assistance. We are also grateful to Phil LePage, Davide Cuzner, and Traci Leys-Schirok from the British Columbia Forest Service for their support. We are indebted to Trevor Goward, Gary Walker and the anonymous reviewers for constructive comments on the manuscript. This research was funded by the Swiss National Science Foundation (SNF 31–59241.99) and is associated with the research program NCCR Plant Survival PS6.
References Armstrong RA (1990) Dispersal, establishment and survival of soredia and fragments of the lichen Hypogymnia physodes (L.) Nyl. New Phytologist, 114, 239–245.
466 J . - C . W A L S E R E T A L . Bailey RH (1976) Ecological aspects of dispersal and establishment in lichens. In: Lichenology: Progress and Problems (eds Brown DH, Hawksworth DL, Bailey RH), pp. 215 –247. Academic Press, London. Balloux F, Lugon-Moulin N (2002) The estimation of population differentiation with microsatellite markers. Molecular Ecology, 11, 155–165. Barton NH, Slatkin M (1986) A quasi-equilibrium theory of the distribution of rare alleles in a subdivided population. Heredity, 56, 409–415. Benson BE, Clague JJ, Grimm KA (1999) Relative sea-level change inferred from intertidal sediments beneath marshes on Vancouver Island, British Columbia. Quaternary International, 60, 49–54. Bossart JL, Prowell DP (1998) Genetic estimates of population structure and gene flow: limitations, lessons and new directions. Trends in Ecology and Evolution, 13, 202 – 206. Bowcock AM, Ruizlinares A, Tomfohrde J et al. (1994) High resolution of human evolutionary trees with polymorphic microsatellites. Nature, 368, 455 – 457. Bridge PD, Hawksworth DL (1998) What molecular biology has to tell us at the species level in lichenized fungi. Lichenologist, 30, 307–320. Brodo IM, Duran Sharnoff S, Sharnoff S (2001) Lichens of North America. Yale University Press, London. Cannings R, Cannings S (1996) British Columbia — A Natural History. Greystone Books, Vancouver. Cavalli-Sforza LL, Edwards AWF (1967) Phylogenetic analysis: models and estimation procedures. American Journal of Human Genetics, 19, 233–257. Clague JJ, James TS (2002) History and isostatic effects of the last ice sheet in southern British Columbia. Quaternary Science Reviews, 21, 71–87. Collevatti RG, Grattapaglia D, Hay JD (2001) Population genetic structure of the endangered tropical tree species Caryocar brasiliense, based on variability at microsatellite loci. Molecular Ecology, 10, 349–356. Büdel B, Scheidegger C (1996) Thallus morphology and anatomy. In: Lichen Biology (ed. Nash TH III), pp. 37– 64. University Press, Cambridge. Denison WC (2003) Apothecia and ascospores of Lobaria oregana and Lobaria pulmonaria. Mycologia, 95, 513 – 518. Dettki H, Klintberg P, Esseen PA (2000) Are epiphytic lichens in young forests limited by local dispersal? Ecoscience, 7, 317–325. Dieringer D, Schlötterer C (2003) microsatellite analyser (msa): a platform-independent analysis tool for large microsatellite data sets. Molecular Ecology Notes, 3, 167–169. Felsenstein J (1989) phylip — phylogeny inference package. Cladistics, 5, 164–166. Galloway DJ (1996) Lichen biogeography. In: Lichen Biology (ed. Nash TH III), pp. 199 –216. University Press, Cambridge. Geitler L (1966) Die Chlorococcalen Dictyochloris und Dictyochloropsis, nov. Gen. Österreichische Botanische Zeitschrift, 113, 155– 164. Goldstein DB, Schlötterer C (1999) Microsatellites: Evolution and Applications. Oxford University Press, London. Goward T (1999) The Lichens of British Columbia. British Colombia Ministry of Forests, Victoria. Hardy OJ, Charbonnel N, Fréville H, Heuertz M (2003) Microsatellite allele size: a simple test to assess their significance on genetic differentiation. Genetics, 163, 1467–1482. Hartl DL, Clark AG (1997) Principles of Population Genetics. Sinauer, Sunderland.
Hawksworth DL, Honegger R (1994) The lichen thallus: a symbiotic phenotype of nutritionally specialized fungi and its response to gall producers. In: Plant Galls (ed. Williams MAJ), pp. 77– 98. Clarendon Press, Oxford. Hawksworth DL, Kirk PM, Sutton BC, Pegler DN (1995) Ainsworth and Bisby’s Dictionary of the Fungi. Cambridge University Press, Cambridge. Heinken T (1999) Dispersal patterns of terricolous lichens by thallus fragments. Lichenologist, 31, 603–612. Högberg N, Kroken S, Thor G, Taylor JW (2002) Reproductive mode and genetic variation suggest a North American origin of European Letharia vulpina. Molecular Ecology, 11, 1191–1196. Honegger R (1996) Mycobionts. In: Lichen Biology (ed. Nash TH III), pp. 24–36. Cambridge University Press, Cambridge. Jordan W (1970) The internal cephalodia of the genus Lobaria. Bryologist, 73, 669–681. Jordan W (1973) The genus Lobaria in North America north of Mexico. Bryologist, 76, 225–251. Josenhans H, Fedje D, Pienitz R, Southon J (1997) Early humans and rapidly changing Holocene sea levels in the Queen Charlotte Islands Hecate Strait, British Columbia, Canada. Science, 277, 71–74. Kärnefelt I (1990) Isidiate taxa in the Teloschistales and their ecological and evolutionary significance. Lichenologist, 22, 307– 320. Kroken S, Taylor JW (2001) Outcrossing and recombination in the lichenized fungus Letharia. Fungal Genetics and Biology, 34, 83 – 92. Landry PA, Koskinen MT, Primmer CR (2002) Deriving evolutionary relationships among populations using microsatellites and (δµ)2: all loci are equal, but some are more equal than others. Genetics, 161, 1339–1347. Lutzoni F, Pagel M, Reeb V (2001) Major fungal lineages are derived from lichen symbiotic ancestors. Nature, 411, 937– 940. Malachowski J, Baker K, Hooper G (1980) Anatomy and the algal–fungal interactions in the lichen Usnea cavernosa. Journal of Phycology, 16, 346–354. Mantel N (1967) Detection of disease clustering and a generalized regression approach. Cancer Research, 27, 209–220. Nybom H, Esselink GD, Werlemark G, Vosman B (2004) Microsatellite DNA marker inheritance indicates preferential pairing between two highly homologous genomes in polyploid and hemisexual dog-roses, Rosa L. Sect. Caninae DC. Heredity, 92, 139–150. Parks JC, Werth CR (1993) A study of spatial features of clones in a population of Bracken Fern, Pteridium aquilinum (Dennstaedtiaceae). American Journal of Botany, 80, 537–544. Petit RJ, Aquinagalde I, de Beaulieu et al. (2003) Glacial refugia: hotspots but not melting pots of genetic diversity. Science, 300, 1563–1565. Printzen C, Ekman S, Tønsberg T (2003) Phylogeography of Cavernularia hultenii: evidence of slow genetic drift in a widely disjunct lichen. Molecular Ecology, 12, 1473–1486. R Development Core Team (2003) r: a language and environment for statistical computing. R Foundation for Statistical Computing. Vienna, Austria. Raybould AF, Mogg RJ, Aldam C et al. (1998) The genetic structure of sea beet (Beta vulgaris ssp. maritima) populations. III. Detection of isolation by distance at microsatellite loci. Heredity, 80, 127–132. Reusch TBH, Stam WT, Olsen JL (2000) A microsatellite-based estimation of clonal diversity and population subdivision in © 2005 Blackwell Publishing Ltd, Molecular Ecology, 14, 457– 467
I S O L A T I O N A N D D I F F E R E N T I A T I O N O F L I C H E N P O P U L A T I O N S 467 Zostera marina, a marine flowering plant. Molecular Ecology, 9, 127–140. Scheidegger C (1995) Early development of transplanted isidioid soredia of Lobaria pulmonaria in an endangered population. Lichenologist, 27, 361– 374. Scheidegger C, Clerc P, Dietrich M et al. (2002) Rote Liste der Gefährdeten Arten der Schweiz: Baum- und Erdbewohnende Flechten. BUWAL, Bern. Scheidegger C, Frey B, Walser J-C (1998) Reintroduction and augmentation of populations of the endangered Lobaria pulmonaria: methods and concepts. In: Lobarion Lichens as Indicators of the Primeval Forests of the Eastern Carpathians (eds Kondratyuk SY Coppins BJ), pp. 33 – 52. Ukrainian Phytosociological Centre, Kiev. Schneider S, Roessli D, Excoffier L (2000) ARLEQUIN Version 2.000: A Software for Population Genetic Analysis. Genetics and Biometry Laboratory University of Geneva, Geneva. Sillett SC, McCune B, Peck JE, Rambo TR, Ruchty A (2000) Dispersal limitations of epiphytic lichens result in species dependent on old-growth forests. Ecological Applications, 10, 789 –799. Slatkin M (1985) Rare alleles as indicators of gene flow. Evolution, 39, 53–65. Slatkin M (1995) A measure of population subdivision based on microsatellite allele frequencies. Genetics, 159, 457– 462. Soltis DE, Gitzendanner MA, Strenge DD, Soltis PS (1997) Chloroplast DNA intraspecific phylogeography of plants from the Pacific Northwest of North America. Plant Systematics and Evolution, 206, 353–373. Stoddart JA, Taylor JF (1988) Genotypic diversity — estimation and prediction in samples. Genetics, 118, 705 –711. Tehler A (1996) Systematics, phylogeny and classification. In: Lichen Biology (ed. Nash TH III), pp. 217– 239. University Press, Cambridge. Tunlid A, Talbot NJ (2002) Genomics of parasitic and symbiotic fungi. Current Opinion in Microbiology, 5, 513 – 519. Walser J-C (2004) Molecular evidence for limited dispersal of vegetative propagules in the epiphytic lichen Lobaria pulmonaria. American Journal of Botany, 91, 1273 –1276. Walser J-C, Gugerli F, Holderegger R, Kuonen D, Scheidegger C
© 2005 Blackwell Publishing Ltd, Molecular Ecology, 14, 457–467
(2004) Recombination and clonal propagation in different populations of the lichen Lobaria pulmonaria. Heredity, 93, 322–329. Walser J-C, Sperisen C, Soliva M, Scheidegger C (2003) Fungusspecific microsatellite primers of lichens: application for the assessment of genetic variation on different spatial scales in Lobaria pulmonaria. Fungal Genetics and Biology, 40, 72–82. Walser J-C, Zoller S, Büchler U, Scheidegger C (2001) Speciesspecific detection of Lobaria pulmonaria (lichenized ascomycetes) diaspores in litter samples trapped in snow cover. Molecular Ecology, 10, 2139–2138. Wang X, Ennos R, Szmidt A, Hansson P (1997) Genetic variability in the canker pathogen fungus, Gremmeniella abietina. 2. Fine scale investigation of the population genetic structure. Canadian Journal of Botany, 75, 1460–1469. Wirth V, Scholler H, Scholz P et al. (1996) Rote Liste der Flechten (Lichenes) der Bundesrepublik Deutschland. Schriftenreihe für Vegetationskunde, 28, 307–368. Wright S (1951) The genetical structure of populations. Annals of Eugenics, 15, 323–354. Zoller S, Lutzoni F, Scheidegger C (1999) Genetic variation within and among populations of the threatened lichen Lobaria pulmonaria in Switzerland and implications for its conservation. Molecular Ecology, 8, 2049–2059.
This study was part of the PhD thesis of Jean-Claude Walser on population genetic processes and ecological adaptation in lichenized fungi. Rolf Holderegger studies the population and landscape genetics of plants and their application in conservation biology. Felix Gugerli’s research focuses on the application of molecular markers for elucidating population processes in space and time. Susan Hoebee is a postdoctoral fellow whose primary research interest includes population genetics, ecology, and conservation of threatened flora. Christoph Scheidegger is head of the Ecological Genetics Division at the Swiss Federal Research Institute WSL. His research interests are the morphology, physiology, and conservation biology of lichenized fungi.
