The evolution of land plants has been shaped by their interactions with a diverse microbial community. Amongst these interactions, the mutually beneficial symbiosis with arbuscular mycorrhizal (AM) fungi is one of the oldest and most widespread (Remy et al., 1994; Treseder and Cross, 2006). At the start of this symbiosis, fungal spores germinate and extend extraradical hyphae. Then, upon contact with a potential host, AM fungi develop hyphopodia on the surface of epidermal cells, enter the root, and produce intraradical hyphae. Subsequently, AM fungi develop arbuscules in the cortex, and establish vesicles and new spores ( Figure 1 ) (Smith and Read, 2008; van Creij et al., 2020). Arbuscules are the characteristic, highly branched structures which AM fungi produce beneath the walls of plant cells, surrounded by a plant-derived membrane. Across this membrane, plants provide the fungus with sugars and fatty acids, in return for inorganic nutrients such as phosphorous and nitrogen (Luginbuehl and Oldroyd, 2017; Paries and Gutjahr, 2023). In vascular plants, fungal colonization is restricted to root tissue, while in non-vascular plants AM fungi colonise other tissues, such as the thalli of liverworts and hornworts (Edwards et al., 2000; Wang and Qiu, 2006).

AM fungi and their host plants are globally distributed (Treseder and Cross, 2006; Öpik et al., 2010), and AM fungi play important roles in both natural ecosystems and agricultural environments, offering plants both improved access to nutrients and increased abiotic and biotic stress resistance (Begum et al., 2019; Samuel and Veeramani, 2021; Bennett and Groten, 2022). Given their ecological importance, their significance for plant biology, and the potential for agricultural benefit, AM fungi have received increasing scientific and industrial interest.

Most research into AM symbioses requires a means of assessing the abundance of AM fungi within the roots and rhizosphere. The earliest studies of AM fungi were dependent on microscopic analysis of stained plant tissue. While many alternative methods for quantifying AM colonization have since been developed using molecular proxies for the presence of AM fungi ( Table 1 ), microscopy remains the gold-standard. This is because microscopy provides a direct measurement of fungal abundance without the need for homogenization of plant tissue. Consequently, microscopy enables the simultaneous assessment of the abundance, spatial distribution, and morphology of specific fungal structures. Microscopy, however, can be extremely time-intensive, and so represents a bottleneck for research into AM symbioses. Therefore, ways to efficiently record and process data from microscopic observations is of significant value.

To ease data collection by microscopic techniques, we have developed an Excel spreadsheet (AMScorer) which enables a user to easily record the presence or absence of all six fungal structures (extraradical hyphae, hyphopodia, intraradical hyphae, arbuscules, vesicles, and spores) during microscopic assays. An accompanying R package, AMReader, can read the data in AMScorer to rapidly produce graphical outputs with statistical analyses. These tools are user-friendly, compatible with many established methods of quantifying AM fungi, and are available at https://github.com/EJarrattBarnham/AMReader along with a comprehensive manual. Here, we present their key features.

AMScorer has been developed to aid the recording and processing of data from AM root colonization assays. It is paired with AMReader to enable quick statistical analysis and data visualization. They are primarily designed to be compatible with the McGonigle magnified intersection method (McGonigle et al., 1990), or modifications thereof (Torabi et al., 2021). They can also be utilized for methods which quantify AM colonization in similar ways, such as the gridline intersect method (Giovannetti and Mosse, 1980). The specific details of these methods have been reviewed (Ho-Plágaro et al., 2020; Torabi et al., 2021). An example workflow for preparation of this material is outlined in Figure 2 . Following preparation of the root material, AMScorer can be used to record AM colonization levels in either 100 fields-of-view or 100 intersections, recording the presence/absence of each fungal structure. For large root systems, it is critical to collect representative root fragments by thoroughly randomizing their selection. It is also possible to quantify multiple slides of root fragments, and subsequently average the results.

Once microscopic quantification has been performed using AMScorer, the data will be automatically collected within the ‘AM Results’ sheet. In order to demonstrate the downstream functions of AMReader, a simulated dataset was generated ( Supplementary File 2 ). This dataset is accessible within the GitHub repository under ‘man/Data’. Figure 6 demonstrates an example workflow for data analysis, which is also described below.

Firstly, the experimental data was loaded and displayed in full by activating all three faceting variables, corresponding to the three different experimental variables contained in the simulated data ( Figure 6A ). Subsequently, the data was filtered using the ‘Structure_Include’ and ‘Condition_Include’ variables ( Figure 6B ). Conditions were selected according to the letter codes stored in the ‘Conditions’ sheet ( Figure 3C ), and structures were selected according to their own codes ( Table 3 ). Subsequently, statistical analysis was conducted by defining the statistical test using the ‘Graph_Stat_Test’ variable ( Figure 6C ). The user can choose from a range of parametric and non-parametric statistical tests, including the TukeyHSD and Wilcoxon pairwise tests. Using ‘Stat_Output’ and Stat_File’ would also allow the user to save an excel file containing reports from any statistical tests conducted on the given data.

AMReader currently supports multiple ways of displaying colonization data. For instance, a box plot may be preferable to bar plots, and this option can be chosen using the Graph_Object variable ( Figure 6D ). In addition, it is possible to produce a graph where the quantification of fungal structures is either separated into distinct facets (Graph_Type = “Facets”), or displayed alongside each other (Graph_Type = “Single”) ( Figure 6E ). Further personalization of the graph can be achieved by a great range of variables ( Table 2 ), as illustrated here by changing the color of bars ( Figure 6E ).

We will seek to expand the range of graphical representations available in the near future, and would also welcome contributions from the community. Alternatively, all datafiles generated during the operation of AMReader are saved to the R environment so that more experienced R users can use AMReader for initial analyses, and then generate figures and results according to their own preferences.

Quantification of AM colonization by microscopy represents a significant bottleneck for the study of AM symbioses. Many alternative, high-throughput methods have been developed to quantify AM colonization through molecular and metabolic changes induced by AM colonization. However, these rarely capture information on the spatial distribution, relative abundance, and morphological features of fungal structures. Consequently, microscopy remains a key requirement for many studies into AM symbioses. To reduce the time required for AM colonization assays, we have developed an Excel spreadsheet, AMScorer, which enables the user to record data during microscopy-based assays, and instantly performs the subsequent data processing steps. Over the course of an experiment, this greatly reduces the time required for quantification of AM colonization. In addition, AMScorer enables all data relevant to the experiment to be stored within a single spreadsheet, allowing easier preservation and dissemination of AM colonization data.

Recently, a novel machine learning approach has also been developed to aid the collection of AM colonization data (Evangelisti et al., 2021). Such methods hold promise for enabling full automation of AM colonization assays. Currently, however, they require high-resolution images with good contrast, which can be difficult to obtain in some plant species and are also time-intensive to acquire. Additionally, these tools require significant expertise in the computational methods. Here we provide tools which require limited technical knowledge, and are compatible with currently established methods for conducting colonization assays, and are compatible with downstream analysis pipelines, such as Ramf (Chiapello et al., 2019).

To facilitate data analysis, we have also developed AMReader, an R package which can transform the data collected by AMScorer into graphical representations quickly and easily. We have sought to develop AMReader so that it is usable by those with little experience in R software, creating a user-friendly system in which the user can control the R pipeline, without the need to write any additional layers of code. This distinguishes AMReader from currently available pipelines. We have, however, also ensured that more experienced R users can employ AMScorer and AMReader within their own, personalized, scripts.

Microscopic quantification of AM colonization represents a major bottleneck for research into AM colonization. AMScorer and AMReader represent new tools which can more than halve the time necessary for assays. We hope this will greatly accelerate research and facilitate sharing of data within the community.

The datasets presented in this study can be found in online repositories. The names of the repository/repositories and accession number(s) can be found in the article/ Supplementary Material .

EJ-B: Conceptualization, Software, Validation, Writing – original draft, Writing – review & editing. GO: Funding acquisition, Writing – review & editing, Writing – original draft. JC: Funding acquisition, Writing – original draft, Writing – review & editing.