State-of-the-Art of iNaturalist as a Source of Data on Ukrainian Fungi

Data sources

We downloaded an archive of all observations, available on iNaturalist as of 1 June 2024, that met the following criteria: (i) lay within the official state boundary of Ukraine, (ii) belong to the Fungi Kingdom, and (iii) were verifiable (accompanied by photos). Total number of downloaded records was 69,211. The data dump and the R code for reproducing the analysis were archived on the GitHub repository (https://github.com/olehprylutskyi/inaturalist_ua_paper).

The state boundary of Ukraine, as well as its administrative division, was acquired using the rgeoboundaries package v. 1.2.9 (Dicko 2023).

To account for potential observer distribution and how it might influence the spatial patterns of fungal observations, we acquired population density data from the Gridded Population of the World, Version 4 (GPWv4): Population Density, Revision 11 (https://sedac.ciesin.columbia.edu/data/set/gpw-v4-population-density-rev11). Understanding population density is crucial as areas with higher human presence are more likely to generate a greater number of observations, which helps us assess spatial biases.

Additionally, to explore the impact of sociopolitical events on observation activity, we downloaded Ukrainian war events data from ACLED’s Ukraine Conflict Monitor (https://acleddata.com/ukraine-conflict-monitor/). These data were essential for analyzing how conflict zones may have disrupted observation patterns, thus providing context for any anomalies in data distribution.

Analysis and visualization

To model species accumulation over the growth of the number of observations (used as a proxy for sampling effort), we applied rarefaction with Hill numbers, using the “abundance” datatype (Chao et al. 2014), implemented in the iNEXT R package (Hsieh, Ma, and Chao 2016).

To address how accurately iNaturalist reflects real-world fungal species composition, we compare iNaturalist data with the data obtained from curated checklists composed by professional mycologists based on critically revised data from available publications and scientific collections. Since Ukraine does not have an up-to-date checklist for all Fungi, we selected two case groups for which such checklists have been recently published: Order Lecanorales, one of the largest orders among lichen-forming Ascomycota (Kondratyuk et al. 2021), and gilled representatives of Order Agaricales, mushroom-forming Basidiomycota (Prylutskyi et al. 2023). To ensure comparability, we unified taxonomy in checklists and iNaturalist species lists using GBIF Species API (https://www.gbif.org/developer/species) via the rgbif R package (Chamberlain et al. 2024).

To explore whether iNaturalist observations correspond with the species records from other sources, we selected two conspicuous case species, Hericium coralloides and Clathrus archeri. Records for these species were obtained from scientific papers (Heluta and Zykova 2018; Shevchenko et al. 2021; Heluta et al. 2022) and GBIF (GBIF.org 2024a, b, with iNaturalist research grade observation dataset excluded). For records accompanied by only textual location descriptions, we georeferenced them semi-automatically using the Geocode by Awesome Table extension for Google Spreadsheets (https://awesome-table.com/) and then manually checked all data points with Google Maps (https://www.google.com/maps) and OpenStreetMap (https://www.openstreetmap.org/) services.

Data transformations, analysis, and visualizations were performed with R statistical environment (R Core Team 2023). We used tidyverse metapackage (Wickham et al. 2019) along with basic R’s functionality for regular data manipulation. For spatial data manipulations (constructing the spatial grid, calculating observation points within each grid cell, and making maps), we used sf package v. 1.0–14 (Pebesma 2018). All visualizations were made using the ggplot2 package v. 3.4.4 (Wickham et al. 2016).

Species diversity and taxonomic coverage

As of 1 June 2024, the Ukrainian segment of iNaturalist reported 2,090 fungal species, represented by 69,211 observations, with both numbers steadily increasing (Figure 1a).

Figure 1

The community identification approach on iNaturalist allows any user to suggest an identification (ID). For an observation to achieve research grade, it must have at least two IDs, with at least 75% of them in agreement. Due to the lack of moderation, the credibility of identifications is directly related to other users’ engagement in the identification process. As shown in Figure 1b, most Ukrainian fungal observations on iNaturalist lack proper community attention, with many observations reviewed by only one user or none besides the observer.

A direct comparison of species composition reported by iNaturalist and professional mycologists for gilled Agaricales revealed generally symmetrical patterns (Figure 1c). For fungal families with higher species diversity in Ukraine, which is known based on professional data, iNaturalist also generally reports more species. For fungal families where species are primarily identified based on macroscopic features (e.g., Amanitaceae), iNaturalist shows species diversity nearly equivalent to professional data. However, for taxa where microscopy is crucial for identification (e.g., Bolbitiaceae, Psathyrellaceae), iNaturalist understandably lags behind professional data.

Similarly to gilled Agaricales, lichen families with high species diversity in Ukraine, based on professional data, are also represented on iNaturalist with a large number of species (e.g., Cladoniaceae, Lecanoraceae, Parmeliaceae, Ramalinaceae; Figure 1d).

iNaturalist is nearly equivalent to professional data for taxa, where species can be identified only based on macroscopic characteristics (e.g., Cladoniaceae, Parmeliaceae, Ramalinaceae). At the same time, iNaturalist lags behind professional data for lichen taxa identified by microscopic features (Byssolomataceae, Scoliciosporaceae).

It should be noted that the lichen-forming fungi (lichens) included in Lecanorales vary significantly in their size and morphology. Among them are fruticose or foliose macrolichen species (e.g., Cladonia, Evernia, Lecanora, Toninia, Usnea) and small, inconspicuous crustose mіcrolichens (e.g., Micarea, Scoliciosporum), which are identified only by microscopic features and are invisible to the naked eye.

Spatial distribution

Observations are moderately evenly distributed across Ukraine (Figure 2a). Hotspots exist around the two largest cities, Kyiv and Kharkiv (indicated in Figure 2b). The Crimean Mountains and the Ukrainian Carpathians, known for attracting tourists, also show higher-than-average observation density. However, based on the empty grid cell fraction, nearly 51% of Ukraine’s territory still lacks fungal observations.

Figure 2

The critical question is whether mycologists can rely on these data in species distribution analysis. To examine this, we analyzed iNaturalist observations for two easily recognizable fungal species, Hericium coralloides (Scop.) Pers. and Clathrus archeri (Berk.) Dring, comparing them with records from scientific collections and academic publications.

H. coralloides is a wood-inhabiting basidiomycete with large, coral-like fruit bodies. It is currently listed in the Red Data Book of Ukraine, the national list of protected species (Ministry of Environmental Protection and Natural Resources of Ukraine 2021). The distribution of this species observations on iNaturalist (Figure 2c) aligns with the records published by professional mycologists (Shevchenko et al. 2021; Heluta et al. 2022; GBIF.org 2024b).

C. archeri is an alien species in Europe, introduced to western Ukraine from Central Europe in recent decades (Heluta and Zykova 2018). Its conspicuous fruit bodies make it easily recognizable in the field. iNaturalist data not only confirmed its known growing occupancy rate in western Ukraine but revealed an easternmost occurrence near Kyiv (Figure 2d), 350 km from the nearest known one reported in the literature (Heluta and Zykova 2018; GBIF.org 2024a).

Social dimension

Although iNaturalist was launched in 2008, the platform allows users to publish observations made on any date, including historical collections. The first fungal observation for Ukraine on iNaturalist is dated 1927. However, most observations were made after 2008, with the most rapid growth occurring since 2020 (Figure 3a).

Figures 3b and 3c show the dynamics of observation activity in the Ukrainian fungal segment of iNaturalist over the past five years, during which Ukraine underwent the COVID-19 pandemic and the full-scale Russian invasion. The pre-war years (2019–2021) were characterized by a moderate, even growth in the number of observations, with Kyiv Oblast (region), Kharkiv Oblast, and Crimea growing faster than others (Figure 3b). In 2022, the start of the war entailed an abrupt decrease in observation activity in the eastern and southern regions, as well as in the north-central Kyiv Oblast and Kyiv City (Figure 3c). Western and central regions, in contrast, showed faster growth this year. In 2023, after the successful liberation of most of the Kharkiv Oblast, observation activity in this region rose again (Figure 3c).

Figure 3

Species diversity and taxonomic coverage

Low user engagement in the identification process remained a significant issue within the fungal segment of the Ukrainian iNaturalist community. Species identifications suggested by observers often remain the only identifications, resulting in only 45% of fungal observations (31,385) achieving research grade as of 1 June 2024. Given that the observations having only the observer’s initial ID might be identified with an in-built AI-based image recognition tool, those IDs should be considered as not credible until at least one independent ID is suggested (Munzi, Isocrono, and Ravera 2023).

Another limitation of the fungal segment of iNaturalist is that many fungal species require diagnostic features that are not visible in photos of fruit bodies. These features include the shape and size of microstructures, odor, and changes in flesh color during oxidation or exposure to specific chemicals. To facilitate accurate identification, such features should be recorded alongside photos (McMullin and Allen 2022). This requires additional effort from users and profound knowledge of the selected group of fungi. Nevertheless, despite these challenges, iNaturalist has proven to be a valuable resource for documenting fungal species composition, particularly for taxa that can be identified primarily by macroscopic features.

Spatial distribution

iNaturalist observations are known to be spatially biased (Di Cecco et al. 2021; Geurts et al. 2023; Hochmair et al. 2020), and Ukrainian fungal data generally follow this pattern. The concentration of observations near densely populated areas or popular recreational sites creates gaps in data coverage across less-populated or remote regions. This uneven distribution limits the use of iNaturalist as the sole data source for comprehensive fungal diversity assessments, which are crucial for conservation planning or biodiversity management. Effective conservation strategies require accurate data on species distribution across entire landscapes, including underrepresented areas (Chandler et al. 2017). When certain regions lack sufficient data, conservation efforts might overlook important habitats or fail to identify regions at risk, leading to incomplete or skewed management decisions (Wetzel et al. 2018).

However, the data distribution does not always adhere to this bias. Sometimes, professional biologists use iNaturalist as a data repository in collaborative biodiversity survey projects (Kuzemko et al. 2021). In these instances, the spatial distribution of observations becomes more systematic, deviating from the usual trend. The observed distribution of Hericium coralloides and Clathrus archeri further demonstrates the value of iNaturalist data in documenting the presence of rare or invasive fungal species. Both species show higher numbers of observations on iNaturalist compared with other sources, highlighting the platform’s potential for monitoring purposes.

Social dimension

Citizen science reflects social processes (Albagli and Iwama 2022). In Ukraine, we observe the intersection of several of them. Like the rest of the world, Ukraine faced severe restrictions during the COVID-19 pandemic, which might have impacted observation activity and the spatial distribution of observations. On a national level, though, the exponential growth in iNaturalist observations since 2019 remained consistent with global trends (Di Cecco et al. 2021). Since 24 February 2022, Ukraine has been living under the conditions of a full-scale Russian invasion. Nearly 6.5 million people were forced to leave the country (UN High Commissioner for Refugees 2024), and 3.5 million became internally displaced persons (IDPs; International Organization for Migration 2024).

On a finer scale, war-caused migrations drastically reshaped citizen science participation. The start of the war entailed an abrupt decrease in observation activity in the eastern and southern regions, as well as in the north-central Kyiv Oblast and Kyiv City. These regions were the main theatres of war, resulting in a decrease in the number of observers (due to the outflow of the population) and reduced observation activity (due to the higher risk of any outdoor pursuit). Western and central regions, which accepted the majority of IDPs and suffered less from the actions, showed faster growth in observation activity that year. In 2023, after the successful liberation of most of the Kharkiv Oblast, observation activity in this region rose again.

This study did not aim to trace individual observers’ activity. However, the data show that only 86 unique users (3.7% of observers) have made 100 or more fungal observations each, and these users have contributed 74.5% of the total observations. Thus, on a regional level, even a single active user moving to another region or country can make a significant difference.

Strengths and weaknesses

iNaturalist has proven to be a powerful tool for accumulating data on fungi in Ukraine, with the vast amount of data already gathered. Among possible strengths of the fungal part of the Ukrainian segment of iNaturalist, we can outline the following:

1. Egalitarianism: Anyone can contribute to biodiversity observations.

2. Scalability: The number of users and observations continues to grow. For instance, for Hericium coralloides and Clathrus archeri, the total number of observations available through iNaturalist has already surpassed all available records from other sources.

3. High spatial and temporal coverage.

These strengths give iNaturalist (and CS in general) an advantage over professional mycology. There are few mycologists in Ukraine, as well as globally (Hawksworth and Lücking 2017). With limited human resources, professional mycologists must narrow their focus to specific taxa, local surveys, and scientific collections. As Ukraine lacks mycological societies to curate regular broadscale surveys, highly technological and distributed CS projects like iNaturalist can fill this gap.

However, iNaturalist fungal data remain underutilized by Ukrainian mycologists. Based on personal observation and discussion with Ukrainian colleagues, we can outline three main limitations of the tool, leashing data use and reducing the overall interest of professional mycologists:

1. Identification uncertainty, especially for taxa requiring microscopy.

2. Lack of voucher specimens, which are often essential for precise identification and integrating observations into professional taxonomic research.

3. Sampling bias, both taxonomic and spatial.

Projects like FunDiS or FungiMap show that the first two limitations can be addressed through increased engagement from professional mycologists, who can provide advice and venues for collected specimens. Aside from improving overall identification quality through revising existing observations, the broader participation of experts might motivate amateurs to upload more observations. An excellent example of this might be the FoU Facebook group, which offers active interaction among amateurs and professionals through the comment section under each post. This interaction stimulates amateurs to collect specimens of interest, transfer them to professionals, and even prepare collaborative scientific publications (Martyniuk et al. 2024; Prydiuk and Safina 2024).

Taxonomic bias is inevitable given the diversity of fungi, and it is unrealistic to expect amateurs to pay equal attention to micromycetes and fungal species with conspicuous fruit bodies. Thus, species lists derived solely from iNaturalist data may not fully represent fungal diversity. However, for groups identified primarily based on macroscopic features, iNaturalist can provide a valuable addition to professional checklists.

Despite the spatial bias demonstrated above, the current state of iNaturalist fungal data allows its use for monitoring rare or invasive species in Ukraine, as evidenced by the cases of Hericium coralloides and Clathrus archeri.

The limitations we list likely contribute to the underuse of iNaturalist data in Ukrainian fungal research compared with data on vascular plants or animals. In our view, the key reason for this is the lack of attention from professional mycologists, even compared with botanists and zoologists. The only available comparative review on the Ukrainian segment of iNaturalist (Balashov 2023) showed a clear dominance of zoologists and botanists among the most active observers and identifiers. This situation creates a vicious circle: A lack of professional engagement leads to low identification credibility, which in turn devalues iNaturalist as a data source among professional mycologists. As we have shown, the potential of iNaturalist data is high, and more active participation of professional mycologists could unlock even more of its value.