1.2.1 From Morphology to Molecules
Assessment of AM fungal diversity and dynamics has been one of the major foci of research into the ecology of AM fungi. Microscopy-based studies paved the way for AM fungal research, raising many questions that molecular techniques have since been able to answer. Early work was based solely on microscopical identification of AM fungal spores sampled from field-collected soil or multiplied in trap cultures (Mosse 1973). This is a slow process, relying heavily on expert knowledge. Furthermore, the spore-based detection of AM fungi has its known limitations (Sanders 2004). Importantly, spores of AM fungi are resting and dispersal organs, and factors driving sporulation are not well understood. Thus, the presence of AM fungal spores is evidence of the species present, but the absence of spores is not evidence of the absence of species (e.g., Clapp et al. 1995; Varela-Cervero et al. 2015). Instead, this indicates the absence of sporulation.
The importance of spore-based studies to AM fungal community ecology for gathering observational evidence and developing and answering essential questions cannot be underestimated. Such studies showed that AM fungal diversity can have seasonal and spatial patterns (Merryweather and Fitter 1998; Zangaro et al. 2013), including successional dynamics (Johnson et al. 1991). The first field evidence to which AM fungal diversity and plant diversity were related was based on soil spore identification (Landis et al. 2004). Sporulation dynamics provided data for developing the concept of plant-AM fungal feedback (Bever 1994; Bever et al. 1997) and differential host-AM fungal relationships (Bever et al. 1996). Not only have spore-based studies provided us with important field-based observations, but sporulating and culturable AM fungi are an important source of clean material for conducting experiments in controlled conditions (van der Heijden et al. 1998) and for genomics, genetics, and physiology of AM fungi (e.g., Tisserant et al. 2012, 2013; Lin et al. 2014).
Molecular techniques are currently the prevailing approach for studying AM fungal communities. Compared to studying AM fungal spores, the deoxyribonucleic acid (DNA)- and particularly ribonucleic acid (RNA)-based methods allow active components of the community to be analyzed. Data generated from root samples are currently the norm in molecular AM fungal community ecology. This reflects the interest in the plantâfungus association and also the difficulty in extracting AM fungal DNA directly from soil (Gamper et al. 2004). The major advances brought about by DNA-based studies increased understanding of the global biodiversity of AM fungi (Ăpik et al. 2013; Davison et al. 2015), and improved knowledge of their taxonomy (SchĂźĂler et al. 2001; Oehl et al. 2011; Redecker et al. 2013). These are the prerequisites to studying community dynamics of AM fungi.
1.2.2 From Community Fingerprinting to Deep Sequencing
The first eukaryotic nuclear ribosomal RNA (rRNA) gene sequences (Medlin et al. 1988) and subsequent design of universal polymerase chain reaction (PCR) primers for fungi (White et al. 1990) led to the first eukaryotic 18S rRNA gene sequences from Glomeromycotan fungi (Simon et al. 1992). The development of PCR-based techniques started molecular taxonomy of AM fungi and enabled linking phenotypic data (mostly spore morphology) with genotypic data (DNA sequences) and yielding better understanding about phylogenetic relationships of Glomeromycota. Early studies of AM fungal community diversity were performed using cloning and sequencing (Clapp et al. 1995; Sanders et al. 1995). The next big development was the design of fungal primers that exclude plant sequences (Helgason et al. 1998). The paradigm shift driven by these studies was the unambiguous evidence that multiple colonizations, that is, several AM fungi cocolonizing a root space, even in short root lengths (Helgason et al. 1999), were widespread and that AM fungi are nonrandomly distributed among their host plants (Helgason et al. 2002). The shift from spore identification to study AM fungal DNA and RNA in roots meant that active components of the fungal community could be analyzed. Furthermore, cloning and Sanger sequencing permitted detection and identification of multiple co-occurring AM fungi in situ, without the need for recognizable morphological features.
It is important to highlight the fact that AM fungi are a monophyletic fungal group (Phylum Glomeromycota), unlike fungi forming other mycorrhizal types; the design of group-specific primers is easier. Such primers helped identify AM fungi in planta, excluding plant sequences and sequences of non-AM fungi colonizing roots and rhizoplane. Several other primer sets specific for AM fungi as a group or smaller subsets of them (e.g., families) have been designed, further improving the detection of AM fungal diversity (Redecker 2000; Lee et al. 2008; KrĂźger et al. 2009; summarized by Hart et al. 2015). Improvements in primer systems have made the large-scale field studies possible by enabling capture of almost all of the diversity of AM fungi in studied ecosystems.
Co-occurrence of multiple AM fungal species in a (root or soil) sample is common. Quantifying community composition necessitates the separation of PCR products of individual fungi by molecular community techniques, by using either fingerprinting or DNA sequencing methods. A range of fingerprinting techniques have been applied to the study of AM fungi: polymerase chain reactionârestriction fragment length polymorphism (PCRâRFLP; Helgason et al. 1999; Husband et al. 2002; Vandenkoornhuyse et al. 2002); single-stranded conformation polymorphism (SSC; Kjøller and Rosendahl 2000; Nielsen et al. 2004); terminal (t)-RFLP (Vandenkoornhuyse et al. 2003); denaturing gradient gel electrophoresis (DGGE; Kowalchuk et al. 2002; Ăpik et al. 2003); and temperature gradient gel electrophoresis (TGGE; Cornejo et al. 2004). The advantage of these techniques was rapid and relatively inexpensive profiling of AM fungal communities; however, without sequence data, the comparison and re-evaluation of individual studies are usually not possible.
In the early molecular AM fungal community studies, there was a trade-off between the high sample throughput but low study-to-study comparability offered by fingerprinting techniques and the costly choice of cloning and Sanger sequencing to identify individual fungal taxa, providing easily comparable and reanalyzable sequence data. High-throughput sequencing (HTS) or next-generation sequencing (although this term is almost outdated as there is now a newer generation of sequencers present, Kircher and Kelso 2010; Venkatesan and Bashir 2011) was initially costly and had low sample throughput. This has now developed to enable both high sample throughput and sufficient sequencing depth per sample at affordable costs to mycorrhizal ecologists. High-throughput sequencing techniques, as they sequence by synthesis, have also removed the need for separation of individual PCR products either via fingerprinting or via cloning techniques. HTS yields more accurate data about rarer members of AM fungal communities via increased sequence numbers per sample (sequencing depth), thus typically reporting higher richness values (Ăpik et al. 2009). It is noteworthy that typical root or soil samples used in AM fungal community surveys contain an average of 5â40 species (operational taxonomic unit [OTUs], molecular taxa; Hart et al. 2015), which is lower than the reported richness values of general fungal (Toju et al. 2013) or bacterial communities (Mantar et al. 2010). Therefore, the sufficient sample-based sequencing depth is lower in the case of AM fungi than that in some other soil microbes (Hart et al. 2015).
The shift from cloning and Sanger sequencing to HTS approaches has been both disruptive, completely changing the scale and design of field-based experiments, and transformative, revealing a new understanding of AM fungal diversity and dynamics. High-throughput sequencing can now be used to relatively rapidly profile dynamics of AM fungal communities in large-scale field studies to describe temporal (Dumbrell et al. 2011; Cotton et al. 2015) and spatial (Ăpik et al. 2013; Davison et al. 2015) variations in these communities. One of the transformative results stemming from HTS data has been the change in understanding about associations between AM fungi and host plant species. Early evidence on AM fungal-host plant species level specificity or preference (Vandenkoornhuyse et al. 2002, 2003) may be better explained by preference among ecological groups of AM fungi and host plants (Ăpik et al. 2009).
The swift accumulation of DNA-based AM fungal community data sets has revealed diversity patterns from local to global scales. The first AM fungal biogeographical meta-analyses described diversity patterns related to biome, spatial (continents), and environmental (edaphic and climatic) factors (Ăpik et al. 2006, 2010; Kivlin et al. 2011). These were followed by HTS-based large-scale case studies, revealing lower global endemism of AM fungi than what was thought earlier (Ăpik et al. 2013; Davison et al. 2015). Observations made in early sequencing studies, such as the dramatic decrease in AM...