Social media reveals unseen interactions with wildlife during Russia's full‐scale invasion of Ukraine

No war has seen such cyber-complexity as what followed the full-scale Russian invasion of Ukraine in February 2022 and the consequential torrent of related photographic and video content shared. Some of this content inevitably features animals and wild nature in previously unseen, unique wartime situations. This creates a unique opportunity to understand how war affects wildlife. Armed conflicts are still unfortunate but significant hallmarks of human civilization in the 21st century. Modern warfare is intrinsically linked to human technological advancement, and the smartphones era puts immediate frontline footage in the hands of millions (Hauter, 2023; Su et al., 2023). No war has seen such cyber-complexity as what followed the full-scale Russian invasion of Ukraine in February 2022 (Brown, 2022), and the consequential torrent of related photographic and video content shared. As social media continues to be inundated with happenings related to the ongoing invasion, content becomes a primary source documenting historical events. Some of this content inevitably features animals and wild nature in previously unseen, unique wartime situations, which are suddenly widely accessible for the first time in human history (Węgrzyn et al., 2023). This creates a unique data collection opportunity to understand how war affects wildlife (i.e., warfare ecology, Hanson, 2018; Machlis & Hanson, 2008), furthering the evidence base that has already been established from other sources (Polianska, 2023; Russell et al., 2024). Collecting environmental data during war is a dangerous endeavor for researchers (Lawrence et al., 2015; Shevchuk et al., 2022). Avoiding unnecessary risk during research requires finding imaginative solutions to data generation, and social media is already a proven source of data for ecology and conservation (Toivonen et al., 2019). Citizen and community science are now routine initiatives, with proactive public participation in research (Hsing et al., 2022). However, passive data collections also occur, which involve assembling information from public content such as photographs or videos, providing data that are otherwise difficult to collect, or about elusive wildlife (Wright et al., 2023). These records can inform the understandings of species ecology and conservation (e.g., distributions) or interactions with humans (e.g., human–wildlife conflict and mitigation). In this way, social media can even be used for data generation during modern wartime (Węgrzyn et al., 2023). From serendipitous browsing (January–December 2023) on the "X" platform (formerly Twitter), we collected photographic and video content of wild, terrestrial mammals directly impacted by the full-scale Russian invasion of Ukraine. Impacts encompass being killed, injured, or rescued due to military activities, or simply observations where wildlife interacted with personnel or equipment. While the data collection was unstructured, the potential of such content to be widely-shared gave us confidence that we captured a high proportion of it simply through daily browsing, and we believe it is important to promote this data source to the scientific community for more comprehensive analyses in the future. We categorized the content by the taxonomic group, the type of interaction, and the animal's status to show some potential uses of these data. In this short letter, we the authors provide some remarks on rare species and interesting behaviors and observations, as well as biases related to this type of data collection. Finally, we collated the content into an open-science database for scholarship and future research. We ensured that sufficient ethical considerations were met to negate risk of harm from this data collection. The data were public, clearly aimed for broad readership, and we censored identifying features (places, faces, etc.) to protect individuals and operational security. We note that some content may be distressing (violence; death; blood; animal cruelty or animal death). We observed a surprising richness of mammal species (n = 13) from multiple families (n = 5) interacting with the elements of the war (Figure 1, Table S1). Two species stood out: gray-dwarf hamster (Nothocricetulus migratorius) and marbled polecat (Vormela peregusna), as both are Red Data Book species in Ukraine, noting them as species of conservation priority. Indeed, marbled polecat is also listed by the IUCN Red List of Threatened Species as "vulnerable" (Abramov et al., 2015). A substantial part of the Ukrainian distribution of both species is currently under Russian occupation. We observed that fortified military positions like trench lines, dugouts, and bunkers are logical interaction points between wildlife and soldiers. Occasionally, these acted like pitfall traps for animals, while for others they were used to acquire resources. Historical sources report how rodents amass in entrenched living quarters (e.g., during World War I, Moore, 2017), because they host both warmth and food. This is further exaggerated by a year of unharvested crops in frontline areas, allowing booming rodent numbers, which were our most frequently observed taxonomic group. We saw that rodents were often, and unsurprisingly, killed by soldiers if their numbers were problematic, or if they caused disruption by getting into food supplies or sleeping quarters. However, we did not expect to observe abject cruelty toward wildlife by soldiers. We gathered four interactions where Russian soldiers reportedly attached wildlife to weapons. First, an anti-personnel mine, seemingly attached to a European hedgehog (Erinaceus europaeus). The second and third were separate occasions of living mice (Mus spp.) fastened to first-person-view (FPV) drones carrying explosives, both times in the frame of the drone's FPV camera. The fourth was of an already dead mouse, attached by the tail to a drone-dropped explosive. Rescues and animal feeding were frequently seen interactions. Red foxes (Vulpes vulpes) were seen taking food from, or being rescued by, soldiers. Civilians were also involved in rescues, such as a hedgehog rescue from a trench (by French war correspondent and civilian, Arman Soldin, later killed by a Russian rocket attack). A standout period for animal rescue content followed Russia's destruction of the Kakhovka Dam in June 2023. The ensuing floodwaters caused untold environmental damage, including to protected areas and landscapes (Moreland & Darbyshire, 2023; Richardson, 2023; Timmins et al., 2023), and killed thousands of animals (Richardson, 2023). Social media is a vast space of seemingly limitless content, and we show just part of its usefulness in revealing otherwise unseen wildlife interactions in wartime situations. However, the true scale of these interactions may forever be unknown, because metadata (e.g., location, date) are seldom available in a wartime context. We used only serendipitous data collection and a single platform, and some content we collected was reposted there from Telegram and TikTok. Because X is restricted in Russia, the algorithm and sentiment of our data are Ukraine-biased, and Russian point-of-view content is reposted second-hand. However, data collection in this space is in line with conservation principles of investigating, highlighting, and mitigating impacts on wildlife and the environment from war (Machlis & Hanson, 2008; Timmins et al., 2023). Indeed, it is estimated that over 500 billion dollars in damage has been inflicted on the Ukrainian environment (Ministry of Environmental Protection and Natural Resources of Ukraine, 2023), which may be tantamount to ecocide (Polianska, 2023). Our work highlights the possibilities for further data collection during wartime using social media, and displays a brief snapshot of the war's effects on wildlife. These records may further inform ongoing initiatives, such as the Ukraine War Environmental Consequences Work Group, or future restoration of Ukrainian war-torn landscapes and ecology (Timmins et al., 2023). Adam F. Smith: Conceptualization (equal), data curation (lead), writing—original draft (lead), and writing—review and editing (lead). Hanna Kuzo: Conceptualization (equal), data curation (equal), writing—original draft (equal), and writing—review and editing (equal). We highly acknowledge the users and journalists who share this content. We are grateful to colleagues and volunteers, especially in Ukraine, for their help and support. The authors declare no conflicts of interest. The data and archive of the content that we used for this article are available at: https://osf.io/8xck5. Table S1. Observations of wildlife made from social media (on the platform "X") during Russia's full-scale invasion of Ukraine. Please note: The publisher is not responsible for the content or functionality of any supporting information supplied by the authors. Any queries (other than missing content) should be directed to the corresponding author for the article.