SEARCH

SEARCH BY CITATION

Keywords:

  • fungi;
  • gene diversity;
  • genotyping;
  • Glomus;
  • multilocus population structure;
  • nested multiplex PCR

Summary

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References
  • • 
    The impact of fallowing on the genetic structure of arbuscular mycorrhizal fungi (AMF) was studied by hierarchical sampling of spores from four plots in a fallow and a cultivated field.
  • • 
    A nested multiplex PCR approach was used to assign the spores to genotypes. Variable introns of the two protein-coding genes GmFOX2 and GmTOR2 were used as co-dominant genetic markers together with the large subunit (LSU) rDNA. The gene diversity and genetic structure of Glomus mosseae, Glomus geosporum and Glomus caledonium were compared within and between the fields.
  • • 
    Spores of G. caledonium and G. geosporum were more abundant in the cultivated field, whereas G. mosseae was more frequent in the fallow field. The number of genotypes was not different between the two fields.
  • • 
    Analysis of gene diversity of G. caledonium in the fallow field indicated that a larger part of the heterogeneity could be attributed to variation between plots rather than subplots, suggesting that the lack of soil cultivation resulted in more heterogeneous population genetic structures. Analyses of haplotype networks of the fungi suggested a subdivision of G. mosseae haplotypes between the two fields, whereas no such division was seen in G. geosporum and G. caledonium. The results show that agricultural practices differently affect both the abundance and the population structure of different AMF species.

Introduction

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Arbuscular mycorrhizal fungi (AMF) within the Glomeromycota (Schüβler et al., 2001) are obligate symbionts associated with most terrestrial plants (Smith & Read, 1996). AMF play an important role in plant nutrient acquisition, including that of crops in fertilized agricultural systems (Jakobsen, 1986). There may be differences among species of AMF in their effects on plant growth, but important differences have also been observed between isolates of the same species (Gamper et al., 2005). For example, Munkvold et al. (2004) found intraspecific differences in the production of external hyphae and in hyphal phosphorus uptake, and Koch et al. (2006) recorded different growth responses to different isolates of Glomus intraradices. Intraspecific genetic differences between isolates within fields have also been demonstrated within G. intraradices (Koch et al., 2004) and within Glomus mosseae and Glomus caledonium (Stukenbrock & Rosendahl, 2005b). However, it is unclear how this intraspecific diversity is maintained. Two main hypotheses have been considered; the apparently homogeneous agricultural soil environment encompasses several microhabitats with different niches, or there is isolation by distance through limited dispersal of propagules (Croll et al., 2008).

In agriculture, the soil disturbance may affect the diversity and community structure of soil organisms. Several studies have compared microbial communities between fields, and have shown that tillage significantly affects AMF communities (Jansa et al., 2003). However, soil cultivation may also impact the distribution of genetic variation within one field. The distribution of genetic variation can be referred to as the genetic structure of the fungal population and can be examined by hierarchical sampling and analysis of gene diversity (Nei, 1973). Gordon et al. (1992) used this method to compare the distribution of genotypes of Fusarium oxysporum in a cultivated field to that in a native grassland. They found that the genotypes in the two areas originated from the same population, but the genetic structure of the species was spatially more heterogeneous in the native grassland compared with the cultivated field. Based on these results they concluded that tillage caused a more homogenous genotype distribution.

Only a few studies of the genetic structure of AMF populations have been published. Koch et al. (2004) found a high genetic variability of G. intraradices within a single field, and Stukenbrock & Rosendahl (2005b) showed that Glomus spp. populations were spatially structured in agricultural fields and interpreted the populations as being composed of individual mycelial networks with a high degree of genetic integrity. It is not known how such mycelial networks are formed, but anastomoses between hyphae originating from germinating spores of the same individual could lead to genetically homogeneous networks (Giovannetti et al., 1999). This would explain how the spatial structure can be maintained in spite of intensive soil cultivation. If this hypothesis is correct, such a network should grow bigger if the soil is left undisturbed, which would result in a spatial structure in which a larger proportion of the heterogeneity is attributable to variation between plots rather than between subplots.

Fungal population genetic structures can be revealed by analyses of multilocus genotypes (McDonald, 1997). AMF are obligate biotrophs, and it is difficult to obtain multiple markers from single individuals. The genetic structure of AMF populations has been studied by Koch et al. (2004), who used single-spore root organ cultures and genotyping with amplified fragment length polymorphism (AFLP) to examine differences between tillage and no-tillage fields. Stukenbrock & Rosendahl (2005b) used a multiplex nested PCR (Stukenbrock & Rosendahl, 2005a) to examine and compare the genetic structures of three Glomus species from an ecologically and a conventionally cultivated field. In this study we use multiplex PCR to generate multilocus genotypes of Glomus spp. from a cultivated and a fallow field to compare the genetic diversity in the two fields and to determine whether the lack of soil cultivation in the fallow field leads to a higher spatial heterogeneity.

Materials and Methods

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Field sites and sampling

Spores were collected in May 2005, from two adjacent fields with a moraine clay loam at the University of Copenhagen experimental farm in Taastrup, Denmark. One of the fields (field #25.2) had been plowed annually and sown with spring barley (2001–2003 and 2005) and winter wheat (2004), while the other (field #25.1) had been fallow since 1993. The plants in the fallow field plots were identified and the dominant species were noted based on visual observation (Table 1).

Table 1.  List of plant species found in the fallow field
SpeciesPlot 1Plot 2Plot 3Plot 4
  1. ‘x’ indicates that the species was recorded in the plot, and ‘D’ indicates the dominant species.

Arrhenatherum elatiusDD  
Elytrigia repensDDD 
Epilobium augustifoliumx   
Taraxacum sp.xx  
Carduus arvensexxxD
Urtica dioica ssp. dioicax   
Symphytum xuplandicumx xx
Epilobium parviflorumxxxx
Festuca rubra ssp. rubraxx D
Vicia hirsutax  x
Senecio jacobaea x x
Trifolium repens x  
Vicia sativa ssp. segetalis x  
Trifolium repens   x
Epilobium augustifolium   x
Dactylis glomerata ssp. glomerata   x
Galium aparine   x
Tragopogon pratensis spp. pratensis   x
Viola tricolors spp. tricolour   x
Cerastium fontanum spp. vulgare   x

A hierarchical sampling was used to examine the heterogeneity at three levels (field, plot and subplot). Four plots, 10 m apart, within each field were placed along parallel transects in the two fields. The two transects were placed 15 m apart. Within each plot, 2-kg samples were collected from four subplots 40 cm apart, resulting in 16 samples from each field (Fig. 1).

image

Figure 1. Hierarchical sampling of spores from the fallow and cultivated fields. The fields were separated by a narrow track. The transects were placed 15 m apart, with 10 m between the plots and 40 cm between the subplots.

Download figure to PowerPoint

Spore extraction

In the laboratory, 100-g soil samples were washed under tap water through a 100-µm sieve. Samples were centrifuged with tap water to remove debris and dead spores, and the resulting pellet was thereafter centrifuged with 50% sucrose to recover spores in the supernatant (Daniels & Skipper, 1982). All healthy-looking spores of Glomus mosseae (T. H. Nicholson & Gerd.) Gerd. & Trappe, Glomus caledonium (T. H. Nicholson & Gerd.) Trappe & Gerd. and Glomus geosporum (T. H. Nicholson & Gerd.) C. Walker were collected under dissection microscope, and transferred to sterile filter paper with flamed forceps. Because of the low spore abundance in the fallow field, spores were extracted from 200 g of soil from all four fallow plots.

Nested multiplex PCR

Single spores were transferred to Eppendorph tubes containing 1.4 µl of extraction buffer (167.5 mM Tris/HCl, pH 8.5, 5 mM (NH4)SO4 and 25 mM β mercaptoethanol) and 8.6 µl of sterile H2O. The spores were crushed with flamed forceps and the samples were denatured for 1 min at 94 °C. Seven µl of the crushed spore solution was transferred to a solution consisting of 8 µl of GATC mix, 1 µl of TQ buffer, 2 µl of sterile H2O, 0.1 µl of Taq polymerase (Amersham Biosciences, Piscataway, NJ, USA) and 0.2 µl of 100 mM of each of the three primer pairs LSU0061f, NDL22r; FOX603f, FOX1376r, and TOR1071f, TOR1638r (Stukenbrock & Rosendahl, 2005a). PCR conditions for primary multiplex PCR were: 94°C for 2 min followed by 30 cycles of 94°C for 1 min, 53°C for 1 min and 72°C for 1 min (Kjøller & Rosendahl, 2000).

The primary PCR products were diluted 50 times and used as templates in the three separate nested PCRs. The nested PCR was performed in a total volume of 20 µl, consisting of 2 µl of template, 8 µl of GATC mix, 1 µl of TQ buffer, 5 µl of sterile H2O, 0.1 µl of Taq polymerase and 2 µl of 10 mM of each of the nested primer pairs: Rk4f and Rk7 mr for the large subunit rDNA (LSU), FOX603f and FOX868r for FOX, and TOR1071f and TOR1444r for TOR (Stukenbrock & Rosendahl, 2005a). Amplification cycles for the nested PCR were identical for all primer combinations: initial denaturation at 94°C for 2 min followed by 25 cycles of 94°C for 1 min, 60°C for 1 min and 72 °C for 1 min. The amplicons were sequenced at Macrogen (Seoul, Korea; http://www.macrogen.com) using the sequencing primers from Stukenbrock & Rosendahl (2005a).

Data analysis

Sequences were manually aligned using the program BioEdit Sequence Alignment Editor (http://www.mbio.ncsu.edu/BioEdit/bioedit.html/bioedit.com). LSU, TOR and FOX sequences from the National Center for Biotechnology Information (NCBI) GenBank were used to assign the spores to genotypes. The sequences were blasted in GenBank (http://www.ncbi.nlm.nih.gov/Genbank), and only sequences with a 100% match were included (Table 4).

Table 4.  Variable sites in the large subunit (LSU) D2 region and the FOX and TOR genes of the haplotypes of Glomus mosseae, Glomus caledonium and Glomus geosporum
HaplotypeLSU sequenceKnown sequence/ accession numberFOX sequenceKnown sequence/ accession numberTOR sequenceKnown sequence/ accession number
  1. The ‘Known sequence/accession number’ columns give the accession number and name of sequences with a 100% match in GenBank. ND, not determined.

Glomus mosseae      
Gm1AACCCGCTTTTTGG. mos. BEG83/ AF145737.1CTAACA-GCTTTTG. mos. BEG84 AY763591.1ACAGA-ATGG. mos. Genotype 3 AY835881.1
Gm2--T----------G.mos. V150/ AJ628054.1------------- ----C---- 
Gm3--TT-A------- -----TT-----A --------C 
Gm4------TC-----G. mos. Genotype 6/ AY835851.1-----TT---C-AG. mos. Genotype 1 AY835870.1--------- 
Gm5-------C-----G. mos. Genotype 2/ AY835848.1-----TT--C--A ---A---A- 
Gm6--TT-A---C---G. mos. BEG185–03/ AY541909.1-----TTT----A -T------C 
Gm7--TCCA------- T----TT--C--A ---A-T--- 
Gm8------T------G. mos. Genotype 7/ AY835852.1-----TT---C-A -TT------ 
Gm9----TATC-C--- -----T------A ND 
Gm10CG-----G--GG- ND ND 
Glomus geosporum      
Gg1CTTTTTAGCGGAGCTTTCTCTG. geosp. Genotype 4 AY835863.1GTT-TAGCGTG. geosp. Genotype 3 AY835873.1--CTGGTCATTGG. geosp. Genotype 4 AY835884.1
Gg2----------T----------G. geosp. Genotype 3 AY835862.1-C--------G. geosp. Genotype 2 AY835872.1------------ 
Glomus caledonium      
Gc1CCTGTTAGCGGAGCTTTCTCTUncult. SN135 EF066661.1ATTTCGACAA --AGAAGCAACG 
Gc2---------------GC---- --------G- ------------ 
Gc3-------C----------C-- ---------- ------------ 
Gc4-----C-----------T--- ---------- ------------ 
Gc5----------AT---------Uncult. Mtrun_ER21 EF066707.1---------- ------------ 
Gc6T---C-------TC-------G. caled. SC_658 AF396799.1ATT-CAACGT --AGAATTTACAG. caled. Genotype 5 AY835882
Gc7T---C------T-C-----T- -----G---AAY835866.1-T---------- 
Gc8T-C-C---T---TC------- -----G----AY835868.1T----------- 

The genetic diversity in the fields was estimated using Simpson's index:

  • image(Eqn 1)

where ni is the number of spores of the ith genotype and N is the total number of spores. D ranges between zero and one and represents the probability that a new spore belongs to a new genotype. A value of zero would indicate that all spores belong to the same genotype, whereas a value of one would suggest that every spore represents a unique genotype.

The genetic heterogeneity in the populations was determined using Nei's gene diversity (Nei, 1973):

  • image(Eqn 2)

(HT, the total haplotype diversity; HC, the average haplotype diversity within the plot; DCS, the average haplotype diversity between plots; DST, the average haplotype diversity between the fallow and the cultivated field.) The proportion of diversity attributable to subdivision between the fallow and cultivated fields is expressed as GST = DST/HT and that between plots as GCS = DCS/HT. Within the two fields, the haplotype diversity within the plots (HC) was further broken down into HC = HSP + DSP, where HSP is the haplotype diversity within subplots and DSP the diversity between subplots. The average diversity attributable to division by subplots can then be expressed as GSP = DSP/HT. These analyses were performed using the program popgene version 1.32 (http://www.ualberta.ca/~fyeh/popgene.pdf).

Sample effort and the actual number of haplotypes were estimated using a first-order jackknife:

  • image(Eqn 3)

(Sj, the total number of haplotypes in the examined area; Sn, the number of observed haplotypes; Q1, the number of haplotypes only found in one sample; n, the total number of individuals (spores).) The estimates were obtained in Pcord5 (McCune & Mefford, 1999).

Haplotype networks were constructed using the network estimation implemented in the software tcs (Clement et al., 2000) tcs uses the probability of parsimony for pairwise sequence differences, known as statistical parsimony, to estimate gene genealogies including multifurcations and/or reticulations (Templeton et al., 1992).

Linkage disequilibrium was determined as an index of association, IA (Maynard-Smith et al., 1993), using the software MultiLocus 1.2 (Agapow & Burt, 2001). IA was estimated from the clone-corrected data set by estimating the number of loci at which all pairs of individuals differ. IA is the variance of these distances (VO) compared with the expected variance if there is no linkage disequilibrium (VE):

  • image(Eqn 4)

If there is linkage equilibrium as a result of frequent recombination events, the expected value of IA is zero. To identify clonal populations with IA values that differ significantly from zero, 1000 randomizations were implemented.

The significance of a difference in genotype frequencies between the fallow and the cultivated fields was determined with a χ2 test using the null hypothesis that the genotype frequencies were similar in the two fields.

Isolation by distance was determined by comparing spatial distances between plots and genetic distances between populations at the plots. As too few spores were recovered, only G. caledonium from the cultivated field (n = 82; 14 populations) and G. mosseae from the fallow field (n = 92; 15 populations) were analyzed. Genetic distances were estimated as Nei's unbiased measures of genetic distance, using the software popgene version 1.32. The correlation of genetic distances with geographical distances for all pairs of populations was determined with the Mantel permutation procedure as implemented in ibd 1.52 (Bohonak, 2002) using 10 000 permutations.

Results

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

Genetic markers from 450 spores were obtained from the two fields, corresponding to an average of 65% of the collected spores. Glomus mosseae predominated in the fallow field, whereas G. caledonium and G. geosporum were most abundant in the cultivated field (Fig. 2). In addition to the analyzed species, Scutellospora pellucida and Acaulospora sp. were abundant in the fallow field but rare in the cultivated field.

image

Figure 2. Number of genotyped spores along the transects in the fallow field (F1–F4) and the cultivated field (C1–C4), given as the number of spores per kg of soil. Gray bars, Glomus caledonium; black bars, Glomus geosporum; white bars, Glomus mosseae.

Download figure to PowerPoint

The number of genotypes did not differ significantly between the two fields (Table 2). Eight genotypes of G. caledonium and 10 genotypes of G. mosseae were identified among all spores collected from the two fields (Table 2). Glomus geosporum was widespread in the cultivated field, but only 13 spores were found in the fallow field. The species was dominated by a single genotype, and only two genotypes were found, although PCR products were obtained from 192 spores. Because of this low number of genotypes, the species was not included in further analyses.

Table 2.  Diversity and heterogeneity estimates of the three species in the cultivated and the fallow field
 Glomus mosseaeGlomus caledoniumGlomus geosporum
TotalCultivatedFallowTotalCultivatedFallowTotalCultivatedFallow
  • *

    Significantly different from 0 (P < 0.001). D, Simpson's diversity index; HT, Nei's gene diversity; GST, diversity attributable to differences between fields; GCS, diversity attributable to differences between plots within the fields; GSP, diversity attributable to differences between subplots; JK, first-order jackknife estimate of the total number of genotypes; IA, index of association; NA, not applicable.

D0.370.510.590.600.550.73NANANA
HT0.790.560.660.420.270.57NANANA
GST0.210.10NA  
GCS0.590.470.170.740.250.60NANANA
GSP 0.280.55 0.580.44 NANA
Spores1303892104822219217913
Genotypes1068886221
JK estimate13.5  11.5  NA  
IA 1.35*1.86* 1.81*1.57*   

The occurrence of the three species along the transect revealed considerable spatial variation (Fig. 2, Table 3). For G. mosseae there was an apparent gradient in spore abundance from the first to the last plot in the fallow field (Fig. 2), but the underlying reason for this is unknown as it did not correlate with field topography or plant cover (Table 1). The jackknife estimates of the number of genotypes (Table 2) indicated that most, but not all, genotypes were sampled.

Table 3. P-values from a two-way ANOVA testing the effects of fields (fallow and cultivated) and plots along the transects on the number of spores recovered
 Glomus mosseaeGlomus caledonium
  1. NS, not significant (P > 0.05).

Field0.030.02
Plot0.02NS
Field × plot0.01NS

The observed index of association for G. mosseae and G. caledonium differed significantly from expected under linkage equilibrium (P < 0.001) in both the fallow and the cultivated fields (Table 2). This strong linkage among the three loci suggests a clonal structure of both species.

The populations of the three species were all dominated by a single genotype (Fig. 3). The haplotype network for G. mosseae showed a clear difference between the fallow and cultivated fields. The haplotype network separated the genotypes of the fallow field from those of the cultivated field. For example, genotype 1 which dominated in the fallow field was only present as a few spores in the cultivated field. Instead, the cultivated field was dominated by genotype 4, which was found infrequently in the fallow field. The G. mosseae haplotypes were separated by relatively few mutational steps, whereas the haplotypes of G. caledonium were genetically more distant (Table 4). The LSU rDNA sequences of G. geosporum could be incorporated into the G. caledonium haplotype network (Fig. 3b), resulting in three separate groups: Gg 1 and 2 which constitute the G. geosporum sensu stricto group, and Gc 6–8 which are similar to known cultures of G. caledonium. Gc 1–5, in contrast, are only known from environmental samples.

image

Figure 3. Large subunit (LSU) rDNA haplotype networks for (a) Glomus mosseae and (b) Glomus caledonium and Glomus geosporum. The lines connecting the circles represent mutational steps, and the black dots are unsampled haplotypes. The size of the circles indicates the frequency of the haplotypes. Dark gray, from the fallow field; light gray, from the cultivated field.

Download figure to PowerPoint

The significance of differences in genotype frequencies between the two fields was determined with a χ2 test. The results showed that the hypothesis of equal distribution of genotypes in the two fields could be rejected (χ2 = 111 for G. mosseae and χ2 = 22.4 for G. caledonium; P < 0.004). The unequal distribution of genotypes between the two fields was also evident from the haplotype networks (Fig. 3a,b).

The clonal diversity (D) of G. mosseae was similar in the two fields (Table 2). The diversity of G. caledonium was slightly higher in the fallow field compared with the cultivated field, but the estimates may have been biased by insufficient sampling in the fallow field.

For both G. mosseae and G. caledonium, GST (representing the heterogeneity attributable to differences between fields) was low whereas GCS (representing the heterogeneity attributable to differences between plots) was high (Table 2), indicating that the spatial variation between plots contributed most to the diversity observed. The highly significant GSP (representing the heterogeneity attributable to differences between subplots) for G. mosseae within the fallow field revealed that a significant proportion of the diversity was attributable to variation between the subplots (GSP = 0.55, P < 0.001), whereas the GSP in the cultivated field was lower (GSP = 0.28). Glomus caledonium showed a similar trend, where the differences between the two fields only accounted for a small proportion of the diversity (GST = 0.10), and most of the diversity was attributable to differences between plots (GCS = 0.74). However, most of the variation could be attributed to differences between subplots (GSP) in the cultivated field and between plots (GCS) in the fallow field (Table 2).

No significant isolation by distance was seen for either G. caledonium (r = −0.09, R2 = 0.007, P = 0.68, where r is the correlation coefficient, R2 is the variance explained by geographic distance, and the P value is for the null hypothesis that r ≤ 0) or G. mosseae (r = 0.10, R2 = 0.01, P = 0.18).

Discussion

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The change in land use clearly affected the abundance and population structure of the mycorrhizal fungi. Glomus mosseae became more abundant after fallowing, whereas G. caledonium and G. geosporum were found more frequently in the cultivated field, indicating that these species responded differently to the discontinuation of tillage. Jansa et al. (2003) studied the effect of tillage in an agricultural system and found that tillage had a strong effect on the abundance of several AMF species. In their study, both G. mosseae and G. caledonium were abundant in both tilled and nontilled systems, but they did not distinguish between genotypes of the species. Previous experiments have also shown that G. mosseae is common in cultured fields (Helgason et al., 1998; Stukenbrock & Rosendahl, 2005b), but no comparisons were made with fallow fields in those studies. The increase in abundance of G. mosseae after fallowing seen in the present study could be linked to the decline in the other two species studied, but the role of competition in these systems is unknown and several other factors are probably involved.

The abundance of G. caledonium was much greater in the cultivated field, compared with the fallow field where only a few spores were found. In spite of this, the most frequent genotypes were found in both fields. This suggests that this species is better adapted to disturbance than G. mosseae, although the mechanism behind this is unknown. It is possible that a fast-growing external mycelium (Warner & Mosse, 1983), and an ability to anastomose (Giovannetti et al., 2004) and establish functional networks may be traits that are selected for in disturbed vegetation systems (Jakobsen, 2004; Rosendahl, 2008).

Gene diversity and genetic structure were greatly influenced by fallowing, which resulted in an increase in genetic heterogeneity of G. caledonium and a significant increase in heterogeneity attributable to differences between plots. The more pronounced homogeneity, expressed as a lower HT, in the cultivated field is interesting and could be a result of the soil treatment, where the genotypes are mixed continuously. A similar result was obtained by Gordon et al. (1992), who found that populations of F. oxysporum were spatially structured in uncultivated soils compared with an adjacent cultivated field. This could indicate that colonies in the fallow field are larger than the colonies in the cultured field.

The heterogeneity of G. mosseae was not influenced as much by fallowing. However, the heterogeneity attributable to differences between subplots became significant after fallowing (Table 2). The underlying reason for this is not easy to discern, but it is important to note that most of the genotypes were not shared by the fields, making the comparison problematic.

The spatial heterogeneity of AMF populations is in concordance with the results of earlier studies of the same species (Stukenbrock & Rosendahl, 2005b) but also with the results obtained by Croll et al. (2008), who found a similar spatial structure of G. intraradices. Croll et al. (2008) proposed that the spatial structure could be attributable to isolation by distance resulting from a restricted dispersal of fungal propagules, but argued that the finding of similar genotypes at different locations made this hypothesis less likely. We also found similar genotypes at different locations, and the population differentiation did not follow an isolation-by-distance pattern, as geographic distance only explained 1% of the genetic distance. The most distant sampling plots were only 50 m apart, and further large-scale sampling of isolates is necessary to determine distances that may limit dispersal of AMF propagules. If dispersal of AMF within fields is not limited by distance, the spatial heterogeneity is more likely to be a result of selection for different genotypes in a heterogeneous soil environment (Stukenbrock & Rosendahl, 2005b). The soil in the fallow field is probably more heterogeneous than that in the cultivated field because of the diverse plant cover. In spite of this, only G. caledonium showed more spatial heterogeneity in the fallow field, and the heterogeneity of G. mosseae was similar in the two fields. If the plant community was driving the genetic heterogeneity of AMF populations, a more spatial genetic heterogeneity would have been expected in the fallow field compared with the cultivated field. However, it is important to take in consideration that the two fields only shared a few G. mosseae genotypes, and that these may have different dispersal strategies and vegetative growth patterns.

The frequency of genotypes was estimated based on the number of spores found in the fields. It is possible, therefore, that a change in frequency of the genotypes reflects a change in sporulation of some genotypes and not overall abundance as a result of the change in land use. Experiments have shown that there are generally more spores in disturbed soils than in undisturbed soils (Jasper et al., 1991) and that several nonsporulating species may dominate in undisturbed soils (Rosendahl & Stukenbrock, 2004). Genotypes of AMF may differ in sporulation, and heavily sporulating genotypes may have been replaced by genotypes that produce fewer spores in the fallow field. This could be particularly relevant for G. mosseae, where there was a limited overlap in occurrence of genotypes between the two fields (Fig. 3). The plant cover was also different between the two fields (Table 1), which may also have affected sporulation (Ahn-Heum et al., 2000) and genotype distribution. However, before any conclusions are drawn from the present results, it is important to stress that the experiment was not replicated. Only two fields were compared, and future experiments should compare more fields to see if this pattern is a general trend.

The use of spore numbers as a quantitative measure of AMF has been criticized by Young (2008), who claimed that the approach used by (Croll et al., 2008) to study spatial variation of G. intraradices using single spores from trap cultures would give a better representation of the biomass. However, the procedure is laborious and only a few individuals can be studied. Croll et al. (2008) were able to establish cultures representing 41 spores, whereas we analyzed 450 spores. The advantage of using in vitro cultures is that more markers can be included in the experiment. The nested multiplex approach has been used with five primer pairs (Stukenbrock & Rosendahl, 2005a), but so far these primers can only be applied to the species used in the present study.

A strong linkage between the alleles was demonstrated in G. mosseae populations in both fields. This is in agreement with previous results (Rosendahl & Taylor, 1997; Stukenbrock & Rosendahl, 2005b), suggesting that the studied species are strictly clonal. Recombination has been detected in AMF. A single study found recombination in a data set based on inter simple sequence repeats (ISSR) markers obtained from field-collected spores (Vandenkoornhuyse et al., 2001). However, dominant markers such as ISSR are difficult to use to estimate linkage as the homology between the alleles is difficult to verify (Rosendahl, 2008). Although no indication of recombination was found in the present study, rare recombination events may still take place between AMF genomes following anastomosis (Giovannetti et al., 1999); however, more information on the nuclear processes following anastomosis events is needed to support this hypothesis. The clonal structure of the Glomus spp. studied in the present experiment makes it problematic to define populations and communities (Rosendahl, 2008). The haplotype networks revealed that the species constitute evolutionary units, but as the genotypes do not exchange genetic material, they may correspond to separate species in a community of recombining species showing intraspecific competition. The replacement of ‘cultivated’G. mosseae genotypes by ‘fallow’G. mosseae genotypes in the fallow field may suggest that such competition is taking place. The haplotype network of G. mosseae showed that the genotypes in the fallow and in the cultivated fields may have had different evolutionary histories. Based on the number of mutations (Table 4), the evolution must have taken place long before the fields were established, but the genotypes may represent isolates with different adaptations to disturbance. The genotypes found in the fallow field may also occur in the cultivated field, and land use practices may determine their relative abundances. Comparative functional studies should explore the heritability of such traits among isolates from disturbed and undisturbed sites.

Acknowledgements

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References

The work was supported by a grant from the Danish National Research Council. Ylva Lekberg and three anonymous referees are acknowledged for their valuable comments on the manuscript.

References

  1. Top of page
  2. Summary
  3. Introduction
  4. Materials and Methods
  5. Results
  6. Discussion
  7. Acknowledgements
  8. References