Public cognition of invasive alien species in China

Public cognition of invasive alien species in China | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Public cognition of invasive alien species in China Liyun Zhang, Jie Huang, Xiaofei Liu This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7581230/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 21 Apr, 2026 Read the published version in Biological Invasions → Version 1 posted 5 You are reading this latest preprint version Abstract Public understanding is critical for invasive alien species (IAS) management, yet national-scale cognition patterns remain understudied. We analyzed TikTok and Sina Microblog data to assess public cognition in China regarding three IAS: Solidago canadensis , Solenopsis invicta , and Trachemys scripta . Three dimensions of cognition, including focus (high-frequency terms), associations (term-pair linkages), and tendencies (sentiment polarity), were quantified using word frequency, semantic co-occurrence networks, and quantile-based sentiment analyses, respectively. Results revealed cross-species disparities: On TikTok, S. canadensis was concerned with its invasiveness (e.g., "Found everywhere", "Fertility → Overlord flower"), while Sina Microblog emphasized its agricultural utility (e.g., "Hu Sheep", "Digestion → Sheep dung"). S. invicta discussions centered on recognition/coping measures on TikTok (e.g., "Red fire ant", "Wound → Soapy water") versus ecological impacts on Sina Microblog (e.g., "Invasive species", "Consequence → Environment"), and T. scripta focused on pet raising issues (TikTok) (e.g., "My home", "Eyes → Can't open eyes") and regulatory policies (Sina Microblog) (e.g., "Release", "Heavy penalty → Country"). These disparities likely stem from species-specific traits and their interactions with human. Cross-platform differences were pronounced: TikTok’s experiential narratives showed polarized sentiment ( S. canadensis median = 0.63; lower/upper quartiles = 0.06/0.94), while Sina Microblog’s policy-driven discourse was predominantly positive (medians: 0.60–0.89). Algorithmic biases amplified these differences: prioritizing emotional content (TikTok) versus institutional narratives (Sina Microblog). A key gap was the scarcity of management strategies, limiting public engagement. Our findings advocate for platform-specific interventions (e.g., TikTok demos of removal techniques) and algorithmic transparency to improve IAS management. Alien species Public cognition Cross-platform disparities Social media platform Species-specific traits Algorithmic biases Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 1. Introduction Biological invasions are inherently linked to human activities (Soto et al. 2024 ), and the management of invasive alien species (IAS) requires consideration of the long-term and complex nature of human influence, necessitating broad public participation (Schwoerer et al. 2024 ). Education and awareness campaigns can enhance public IAS cognition (Miralles et al. 2016 ; Oele et al. 2015 ) and facilitate their understanding of relevant policies and measures, thereby improving engagement in management efforts (Schreck Reis et al. 2013 ). Additionally, the role of education in shaping public cognition is crucial for the development of invasion biology as a discipline, since most foundational research has historically prioritized ecological dimensions over sociological insights (Abrahams et al. 2019 ), particularly regarding public cognition of IAS derived from innovative platforms like social media. Unlike traditional surveys which are often limited by sample size and geographic constraints, social media data offer a dynamic and cost-effective means of monitoring public sentiment over time (Zahra et al. 2025 ). The interactive and decentralized nature of social media facilitates rapid information sharing and dissemination, allowing scientific knowledge, policy updates, and public awareness campaigns to spread efficiently through multimedia content (e.g., videos, infographics) and user engagement mechanisms such as sharing and commenting (Dart et al. 2022 ). Additionally, these platforms serve as valuable sources of public feedback which allow researchers to identify collective cognitions of IAS using data mining techniques (Daume 2016 ). However, extracting relevant data requires overcoming challenges such as noise, misinformation, and selection biases. Advanced computational methods (e.g., natural language processing and machine learning algorithms) are required to automate content filtering and enhance analytical accuracy (Kaplan, Haenlein 2010 ). Thus, researchers must address methodological complexities to ensure data reliability and validity. China represents a critical case in the global context of biological invasion, being simultaneously among the most IAS-vulnerable regions (Zhang et al. 2024 ) and the greatest potential IAS sources worldwide (Paini et al. 2016 ). Thus, effective IAS management requires substantial public participation, as successful strategies depend on community engagement in prevention, monitoring, and reporting (Zhang et al. 2019 ). However, public’s IAS awareness in China remains alarmingly low (Wang 2024 ), as evidenced by irrational public conducts such as the widespread religious release of red-eared slider ( Trachemys scripta ) (Liu et al. 2013 ). No studies have systematically explored how the general public views IAS on a national basis. Additionally, China has the largest social media user base globally, with 1.071 billion users accounting for 76% of its total population in 2024 (KAWO 2025 ), providing a vast data source for studying public cognition of IAS. Given this context, we hypothesize that public cognition of IAS on social media exhibits distinct cross-species disparities and cross-platform differences. To address this, we selected dominant Chinese social media platforms as data sources and focused on highly-discussed IAS as study targets. Through large-scale data collection, processing, and computational analysis, we aimed to reveal public cognitive characteristics toward IAS in China. The findings not only advance understanding of the socio-ecological dimensions of biological invasions but also offer scientific foundations for designing social media-based public education campaigns, thereby better supporting the formulation and implementation of IAS management strategies. 2. Materials and Methods 2.1 Definitions and indicators Cognition represents the mental processes through which individuals acquire, process, and utilize knowledge to understand and respond to their environment, encompassing perception, learning, memory, and decision-making (Bayne et al. 2019 ). These mechanisms fundamentally shape both individual behaviors and collective societal actions, particularly in environmental contexts (Schwoerer et al. 2024 ). In IAS management, public cognition plays a pivotal role by influencing awareness levels and behavioral responses toward IAS (Shackleton et al. 2019a ; Shackleton et al. 2019b ). To operationalize this construct, we quantified public cognition through three empirically validated indicators: (1) cognitive focus (the central themes of public attention regarding IAS), (2) cognitive association (the semantic relations linking these focal points to related concepts), and (3) cognitive tendency (the affective valence triggering potential management behaviors). Together, these metrics form a comprehensive framework for analyzing how public cognition evolves from initial attention (focus) to conceptual connections (association) and ultimately to evaluative judgments (tendency), providing valuable insights for science communication and policy development in biological invasion management. 2.2 Selection of social media platforms and IAS We referred to relevant reports and selected TikTok and Sina Microblog as social media data sources in view of their user base, monthly active users, and content types. TikTok is China's largest short-video sharing platform, with 1,100 million monthly active users as of December 2024 (KAWO 2025 ). Similarly, Sina Microblog is the country's leading microblogging platform, publishing not only short videos but also images and text, with 590 million monthly active users up to the end of 2024 (KAWO 2025 ). We systematically retrieved 2,753 relevant posts from TikTok and Sina Microblog platforms using the following search terms: "invasive organisms", "invasive species", "biological invasion", "alien invasion", "alien organisms", and "alien species". From this dataset, we identified 695 posts containing specific IAS mentions. The five most frequently mentioned IAS were: Solidago canadensis (n = 69), T. scripta (n = 60), Eichhornia crassipes (n = 57), Solenopsis invicta (n = 50), and Pomacea canaliculata (n = 47). To assess public engagement, we selected the most engaged posts (by like counts) for each taxonomic group, including S. canadensis (17,255 likes) for plants, S. invicta (24,257 likes) for insects, and T. scripta (1,812 likes) for animals. 2.3 Data collection, preprocessing, and analysis The fundamental principle of web crawlers is to automate the process of accessing web pages and extracting target information by simulating human browsing behavior. Specifically, a web crawler initiates from one or multiple seed URLs, sends HTTP requests to retrieve web content, and parses the HTML code to extract hyperlinks and other relevant data. During this process, newly discovered URLs are continuously added to a crawling queue, and the crawler follows a predefined strategy to select subsequent pages for scraping until termination conditions are met. The crawler was developed in PyCharm (JetBrains, Prague, Czech Republic) using Python 3.6.0 (Python Software Foundation, Beaverton, OR, USA), with Selenium WebDriver (ThoughtWorks, Chicago, IL, USA) employed for browser automation (see Appendix S1: Codes). Post-data acquisition, redundant information was filtered from the raw HTML source to retain specific content. The Beautiful Soup parser was utilized to transform HTML/XML documents into parse trees, enabling targeted data extraction through Python scripts designed for different webpage structures, particularly for collecting user comments. For multi-page comment sections, the crawler simulated pagination operations to ensure comprehensive data collection. Anti-crawler countermeasures (e.g., CAPTCHAs, IP blocking) were mitigated through strategic implementations including configuring appropriate request intervals, deploying proxy IP rotation, etc. Due to the time-intensive nature of web data crawling, processing, and analytical procedures, we implemented a phased data collection approach: TikTok platform data was initially crawled up to March 31, 2023, followed by subsequent collection of Sina Microblog platform data up to December 31, 2023. Additionally, the raw data obtained through web crawling typically contains substantial noise, including special characters, punctuation marks, and other non-content elements. To address this, we employed regular expressions (regex) - a powerful and flexible pattern-matching tool for text processing - to systematically filter and remove these extraneous elements from the extracted textual data. This preprocessing step utilized Python's built-in re module to implement sophisticated string matching and substitution operations, effectively cleansing the dataset while preserving meaningful content. The regex patterns were carefully designed to account for various character encodings and linguistic peculiarities encountered in the source material, ensuring comprehensive noise removal across diverse text formats. Cognitive focus was determined by analyzing high-frequency terms in social media user feedback. We extracted the top 20 most frequent words using Python's Jieba library for text segmentation (see Appendix S1: Codes), and categorized them into five thematic groups: Identification & Detection (ID), Distribution & Abundance (DA), Impact (IM), Management (MA), and Others (OT). To ensure statistical accuracy, we consolidated synonyms (e.g., merging "roadside" variants into a unified term), and applied stop-word lists (e.g., Harbin Institute of Technology and Baidu stop-word lists) to filter meaningless terms before frequency calculation. Cognitive association was quantitatively analyzed through term co-occurrence relationships in social media comments using Pointwise Mutual Information (PMI) (see Appendix S1: Codes). This statistical measure evaluates lexical association strength by comparing the observed joint probability of term pairs against their expected probabilities under independence assumption, computed as: PMI(x,y) = log₂(P(x,y)/(P(x)×P(y))) where P(x,y) is joint probability of terms x and y co-occurring, and P(x) and P(y) are marginal probabilities of individual term occurrences. The interpretation follows that PMI > 0 indicates semantic association, with magnitude reflecting strength. Our analysis protocol comprised: (1) selecting the lexicon from top 100 high-frequency terms in cognitive focus categories, (2) computing PMI values for all possible term pairs, and (3) identifying the 20 most strongly associated pairs (highest PMI) for cognitive pattern characterization. We quantified cognitive tendency through sentiment analysis of user comments using the open-source Senta system (Baidu, Beijing, China), which integrates RoBERTa (Robustly Optimized BERT Pretraining Approach) with SKEP (Sentiment Knowledge Enhanced Pre-training) for domain-specific sentiment analysis (see Appendix S1: Codes). The system operates by first leveraging RoBERTa's robust contextual understanding to process text, then applying SKEP's specialized sentiment knowledge through techniques including sentiment word masking, aspect-sentiment pair identification, and sentiment reconstruction objectives. This combined approach enables both sentence-level and aspect-level analysis, outputting a normalized sentiment polarity index from 0 (negative) to 1 (positive) that captures nuanced public cognition while maintaining the methodological rigor demonstrated in prior opinion mining research (Liu 2012 ; Tian et al. 2020 ). 3. Results 3.1 Cognitive focus The top 20 highest-frequency words and their categories based on TikTok are shown in Fig. 1 (see Appendix S2: Table S1 ). The cognitive focus on S. canadensis is indicated by two ID terms ("Canadian goldenrod" [1486] and "Seeds" [296]), nine DA terms ("Found everywhere" [1013], "Roadside"[308], etc.), two IM terms ("Toxic" [202] and "Biological invasion" [182]), four MA terms ("Research" [279], "Planting" [257], etc.); and three OT terms ("Desert" [1124], "Have seen" [250], and "Childhood" [177]). Further, the focus is comprised mainly of DA (2983) and ID (1782), and rarely of OT (1301), MA (897), and IM (384). For S. invicta , the cognitive focus covers four ID terms ("Ant" [2318], "Red fire ant" [1558], etc.), five DA terms ("A lot" [565], "My home" [226], etc.), eight IM terms ("Got bitten" [487], "Be careful" [389], etc.), two MA terms ("Gasoline" [365] and "Boiling water" [218]), and one OT term ("Childhood" [209]). Further, the focus is comprised mainly of ID (4455), IM (2154), and DA (1360), and rarely of MA (583) and OT (209). Similarly, the cognitive focus surrounding T. scripta includes 11 ID terms ("Brazilian turtle" [2221], "Turtle" [1126], etc.), two DA terms ("My home" [849] and "Deep water" [145]), one IM term ("Biological invasion" [104]), five MA terms ("Raise" [344], "What to do" [236], etc.), and one OT term ("Two" [174]). Further, the focus is comprised mainly of ID (4468), and rarely of DA (994), MA (988), OT (174), and IM (104). The top 20 highest-frequency words and their categories based on Sina Microblog are shown in Fig. 2 (see Appendix S2: Table S2 ). Public cognitive focus on S. canadensis is indicated by five ID terms ("Canadian goldenrod" [205], "Fresh flowers" [123], etc.), six DA terms ("A lot" [73], "Shanghai"[48], etc.), five IM terms ("Hu sheep" [104], "Biological invasion" [75], etc.), and four MA terms ("Planting" [76], "Sina Microblog" [43], etc.). Further, the focus is comprised mainly of ID (480), and rarely of IM (311), DA (261), and MA (182). For S. invicta , the cognitive focus covers five ID terms ("Red fire ant" [79], "Ant" [51], etc.), one DA term ("A lot" [40]), seven IM terms ("Invasive species" [100], "Terrifying" [38], etc.), six MA terms ("Release" [55], "Introduction" 40], etc.), and one OT term ("Childhood" [14]). Further, the focus is comprised mainly of IM (237), ID (197), and MA (179), and rarely of DA (40) and OT (23). Similarly, the cognitive focus surrounding T. scripta includes eight ID terms ("Brazilian turtle" [63], "Turtle" [31], etc.), one DA term ("My home" [20]), five IM terms ("Cute" [27], "Be careful" [27], etc.), four MA terms ("Release" [96], "Science" [18], etc.), and two OT terms ("Two" [12] and "Life" [11]). Further, the focus is comprised mainly of ID (188), MA (142), and IM (113), and rarely of OT (23) and DA (20). 3.2 Cognitive association The top 20 most closely associated word pairs based on TikTok are shown in Fig. 3 (see Appendix S2: Table S3 ). Public cognitive association on S. canadensis is primarily highlighted by ecological and agricultural threats. Strong links between "Fertility → Overlord flower" (7.87) and "Native → Crop" (7.40) highlight concerns over its invasive impact on farmland. Ecological harm ("Ecology → Overlord flower" [7.37]) and weed-like traits (5.59–5.72) further dominate perceptions, while positive attributes (e.g., "Medical value" [5.65]) receive minimal attention. For S. invicta , the cognitive association is primarily linked to health risks and folk remedies. Strong links between "Wound → Soapy water" (5.23) and "Redness → Wound" (4.29) highlight concerns over their painful stings, while home treatments like "Effect → Diesel" (4.31) and "Gasoline → Diesel" (3.99) reflect informal mitigation attempts. Mentions of "Hospital → Whole body" (4.53) and "Toxicity → Child" (3.78) underscore perceived dangers, particularly to children ("Childhood → Hometown" [4.35]). The weak "Childhood → Dirt pile" association (3.52) reflect limited experiential encounters with fire ant habitats in everyday environments. Health issues and morphological traits were identified as the primary aspects of cognitive association regarding T. scripta . The strongest associations highlight eye problems ("Eyes → Can't open eyes" [6.65]) and color variations ("Albino → Darkening" [6.42]), suggesting widespread discussion of these visible abnormalities. Price-related associations ("Price → Mutation" [5.52]; "Pet → Price" [4.89]) indicate significant attention to the commercial aspects of color morphs. Health concerns appear fragmented, with notable links between "Sick → Neck" (5.47) and "Pneumonia → Normal" (4.92), while husbandry practices like "Sunbathe → Stone" (5.46) and "Shallow water → Stone" (4.58) reflect common captive management approaches. The top 20 most closely associated word pairs based on Sina Microblog are shown in Fig. 4 (see Appendix S2: Table S4). Public cognitive association on S. canadensis is primarily highlighted by agricultural and utilization contexts. The strongest associations highlight its potential as livestock feed ("Digestion → Sheep dung" [6.70]; "Seed → Sheep dung" [5.48]) and breeding value ("Ability → Breeding" [6.40]; "Nationwide → Breeding" [6.02]). Commercial aspects appear significant ("Specialized → Price" [6.29]), while ecological concerns manifest through terms like "Harm → Growth" (5.30) and "Large area → Wasteland" (5.17). Notably, some users paradoxically view its spread as beneficial ("Natural enemy → Good thing" [5.74]; "Spread → Good thing" [5.13]), suggesting divided perceptions about this invasive species' impacts. For S. invicta , the cognitive association is primarily linked to ecological threats and public safety concerns. The strongest association ("Hollow → Lotus seed" [7.66]) suggests discussions often link fire ants with agricultural damage, particularly to lotus crops. Ecological impacts are emphasized through terms like "Consequence → Environment" (5.92) and "Harm → Country" (5.92), while public health concerns manifest in "Blisters → Redness" (5.85). Notably, the ants are frequently characterized with negative anthropomorphic terms ("Malicious → Name" [5.85]; "Human → Insidious" [5.60]), reflecting strong public aversion. The association "Helpless shrug → Environment" (5.66) reveals a sense of resignation regarding control efforts. Regulatory and ecological contexts were identified as the primary aspects of cognitive association regarding T. scripta . The strongest associations ("Sina → Hot topic" [7.76]; "Hot topic → Option" [7.76]) highlight active policy debates about this invasive species. Ecological concerns appear through "Heavy penalty → Country" (6.59) and "Environment → Harm" (5.59), reflecting awareness of governmental controls and environmental impacts. Pet trade issues emerge in "Organism → Aquarium" (6.59) and "Animal → Stray" (5.65), indicating concerns about aquarium releases. The association "Helpless shrug → Hometown" (5.85) suggests public resignation regarding established populations, while "Turtle head → Mouth" (6.18) shows anatomical interest. 3.3 Cognitive tendency The distribution of tendency indexes based on TikTok is shown in Fig. 5 (see Appendix S3: Datasets). S. canadensis demonstrates the most polarized tendency distribution (median = 0.63), with highly positive views in the upper quartile (0.94) contrasting sharply with extremely negative views in the lower quartile (0.06) (19,908 comments). S. invicta receives the most negative overall tendency (median = 0.23), though a small segment maintains neutral-to-positive views (upper quartile = 0.85) (23,008 comments). T. scripta shows intermediate but still negative-leaning tendency (median = 0.43), with its upper quartile (0.92) approaching neutral-positive range (12,302 review comments). Notably, all species exhibit strong negative components (lower quartiles 0.02–0.06), with S. invicta being the most uniformly disliked and S. canadensis generating the most divided public opinions among the three species on TikTok. The distribution of tendency indexes based on Sina Microblog is shown in Fig. 6 (see Appendix S3: Datasets). S. canadensis demonstrates the most favorable overall tendency (median = 0.89) with an upper quartile reaching 0.98, indicating a widespread positive view, though a lower quartile of 0.25 suggests a minority holding negative view (2,009 comments). S. invicta shows more polarized tendencies (median = 0.6) with near-universal positivity in the upper quartile (0.97) contrasting sharply with strong negativity in the lower quartile (0.05) (1,189 comments). T. scripta exhibits similar polarization (median = 0.65) with overwhelmingly positive tendency in the upper quartile (0.97) but retains some negative tendencies (lower quartile = 0.09) (868 comments). Notably, all three species maintain high upper quartile values (0.97–0.98), suggesting substantial positive tendency coexists with varying degrees of negative tendencies across these invasive species on Sina Microblog. 4. Discussion Public cognition of IAS exhibited distinct patterns across species. For S. canadensis , identification and distribution patterns dominated discussions, with TikTok emphasizing invasiveness while Sina Microblog focused on agricultural utility. S. invicta cognition centered on morphological identification and health risks, with TikTok highlighting recognition/coping measures and Sina Microblog framing it as an ecological menace. T. scripta discussions primarily involved identification traits, with TikTok addressing pet-related welfare issues and Sina Microblog focusing on regulatory policies. Sentiment analysis revealed species-specific patterns, with S. canadensis showing platform-dependent polarization (TikTok) versus uniform positivity (Sina Microblog), S. invicta exhibiting the most negative tendency on TikTok, and T. scripta displaying moderate negativity on TikTok versus mixed positivity on Sina Microblog. The observed variations likely stem from species-specific traits and their direct human interactions. S. canadensis , as a visible plant, draws attention to its ecological and agricultural impacts, while S. invicta 's health risks and T. scripta 's pet trade associations reflect their immediate human relevance. The dominance of identification features suggests public reliance on easily observable characteristics (Daume 2016 ), whereas limited discussion of ecological impacts and management may indicate knowledge gaps (Liu et al. 2013 ) or lower perceived urgency regarding long-term consequences (Zhang et al. 2019 ). The platforms demonstrated markedly different discourse patterns. TikTok's discussions were more experiential and ecological, featuring extreme tendency polarization for S. canadensis , strong negativity for S. invicta , and moderate negativity for T. scripta . In contrast, Sina Microblog exhibited policy-oriented discourse with consistently positive tendency. Platform-specific emphases emerged clearly: TikTok prioritized ecological dominance and health consequences, while Sina Microblog emphasized agricultural applications, public safety threats, and regulatory governance. This divergence suggests TikTok facilitates more personal, impact-focused narratives with emotional extremes, whereas Sina Microblog fosters more institutional, solution-oriented discussions with positive framing, particularly regarding policy interventions and agricultural utility. Additionally, both platforms shared a common underrepresentation of management strategies despite their differing thematic focuses. Platform disparities may arise from distinct user demographics and content formats. TikTok's experiential focus and emotional polarization likely reflect its short-form, engagement-driven content and its relatively older user base (QuestMobile 2024 ), amplifying personal encounters with IAS. In contrast, Sina Microblog's policy-oriented, positive discourse aligns with its younger, more educated users and text-heavy format (QuestMobile 2024 ), facilitating institutional discussions. The platforms' algorithmic preferences, shown by TikTok favoring emotional content and Sina Microblog promoting authoritative sources (Wang 2023 ), further reinforce these differences in framing IAS issues. Both this study and prior research leverage social media as primary data sources (e.g., TikTok/Sina Microblog vs. Twitter/Facebook/iNaturalist), utilizing user-generated content to analyze public cognition of IAS (Daume 2016 ; Marcenò et al. 2021 ). Common methodologies include web crawling for data collection (Chen et al. 2023 ) and integration of citizen science principles to document species distributions (Viana et al. 2025 ; Werenkraut et al. 2020 ). However, this study advances methodological rigor by employing semantic co-occurrence networks and quantile-based sentiment analysis, enabling granular detection of emotional polarization - unlike sentiment categorization (positive/neutral/negative) in earlier works (Zahra et al. 2025 ). Additionally, our systematic cross-platform comparison (TikTok vs. Sina Microblog) reveals algorithmic influences on discourse framing, whereas most literature focuses on single-platform data utility (Daume 2016 ; Izquierdo-Gómez 2022 ). Consistent with global studies, species-specific traits drive public cognition: S. invicta ’s health risks and T. scripta ’s pet trade associations mirror "human-centric" attention patterns observed for Penaeus monodon ’s commercial impacts in Brazil (Viana et al. 2025 ). Both this study and literature confirm social media’s role in filling scientific monitoring gaps, such as undocumented distributions (Schifani, Paolinelli 2018 ). Divergences emerge in identified cognitive blind spots: while studies like Dart et al. ( 2022 ) emphasize distribution data gaps, this study reveals pervasive neglect of management strategies across Chinese platforms, aligning with Zhang et al. ( 2019 )’s observation of "passive strategies" in border regions. Crucially, we expose platform algorithms as amplifiers of emotional extremes (e.g., TikTok’s engagement-driven content promoting polarization), whereas prior research attributes cognitive differences primarily to user demographics (Chen et al. 2023 ) or cultural contexts (Henke et al. 2024 ). The consistent underrepresentation of management strategies in public discourse (Zhang et al. 2019 ) necessitates algorithm-optimized science communication, such as short videos demonstrating control measures. Cross-platform data integration - mirroring Izquierdo-Gómez ( 2022 )’s triangulation of Facebook, questionnaires, and local knowledge - can mitigate single-platform biases (e.g., Sina Microblog’s policy focus overlooking ecological experiences). Furthermore, algorithmic transparency in content curation must be addressed through policy frameworks akin to public-private partnerships (Markell et al. 2020 ), given its demonstrated role in shaping polarized perceptions. This study underscores that algorithmic governance is as critical as species traits or user demographics in understanding contemporary IAS cognition. To enhance public education on IAS through social media, our findings suggest the need for platform-tailored science communication strategies that address cognitive gaps while leveraging algorithmic dynamics. Given TikTok's dominance in ecological experience-sharing and emotional polarization, authorities should collaborate with content creators to develop engaging short videos that visually demonstrate species identification, ecological impacts, and simple mitigation measures (e.g., safe removal of S. canadensis or first aid for S. invicta bites), capitalizing on the platform's preference for visceral narratives. For Sina Microblog, policy-aligned messaging should emphasize regulatory successes and agricultural co-benefits of IAS management, utilizing its text-heavy format to disseminate expert-endorsed guidelines and participatory monitoring initiatives. Both platforms must prioritize algorithmic transparency to counteract extreme sentiment amplification, potentially through partnerships with tech companies to promote authoritative content (Markell et al. 2020 ). Crucially, educational campaigns should bridge the observed management strategy gap by integrating actionable advice (e.g., reporting protocols, biological controls) into trending formats - such as TikTok challenges or Sina Microblog Q&A threads - while addressing species-specific cognitive biases (e.g., T. scripta 's pet trade appeal). Cross-platform synergies, akin to citizen science triangulation (Izquierdo-Gómez 2022 ), could further enhance reach, combining TikTok's broad visibility with Sina Microblog's policy influence to foster a cohesive, evidence-based public discourse on IAS mitigation. 5. Conclusion This study demonstrates that public cognition of IAS in China exhibits distinct cross-species ( S. canadensis : ecological vs. agricultural focus; S. invicta : health risks; T. scripta : pet trade features) and cross-platform disparities (TikTok: experiential/polarized; Sina Microblog: policy/positive), shaped by species traits and algorithmic amplification of emotional or institutional narratives. Critically, the pervasive underrepresentation of management strategies across platforms necessitates integrating actionable control measures (e.g., reporting protocols, biological controls) into algorithm-optimized education - leveraging TikTok’s short videos for experiential learning and Sina Microblog’s text-based format for policy frameworks. Future research should prioritize cross-platform data triangulation (e.g., combining citizen science records with social media analytics) to mitigate sampling biases and explore algorithmic governance models that counter emotional extremism while promoting science communication. These steps are essential to transform fragmented public cognition into coordinated IAS management. Declarations Acknowledgements The authors thank Mr. Shengqing Yang (Yunnan University) and Ms. Jun Wan (Yunnan University) for their assistance in data collection and analyses. The authors are also grateful to the editor and anonymous reviewers for their invaluable contributions. Authors Contribution Statement Liyun Zhang and Xiaofei Liu contributed to the study conception and design. Jie Huang conducted material preparation, field surveys, and data collection and analysis. Liyun Zhang wrote the first draft of this manuscript, and Jie Huang and Xiaofei Liu made comments. All authors have read and approved the final manuscript. Funding This work was supported by the National Key R&D Program of China (grant number 2024YFF1306700), the Yunnan International Joint Laboratory of Fruit-Vegetable-Flower Invasive Insect Pests Management (grant number 202303AP140018), and the National Natural Science Foundation of China (grant number 42161011). Code availability The codes we developed for data collection and analyses during the current study can be found online in the Supporting Material section at the end of this article. Competing Interests The authors have no relevant financial or non-financial interests to disclose. Data Availability The datasets generated during and/or analyzed during the current study can be found online in the Supporting Material section at the end of this article. Ethics approval This is an observational study and no ethical approval is required. Consent to participate Informed consent was obtained from all individual participants included in the study. Consent to publish The authors affirm that human research participants provided informed consent for publication of the manuscript. Supporting Information Additional supporting information can be found online in the Supporting Material section at the end of this article. References Abrahams B, Sitas N, Esler KJ (2019) Exploring the dynamics of research collaborations by mapping social networks in invasion science. Journal of Environmental Management 229:27-37. https://doi.org/10.1016/j.jenvman.2018.06.051 Bayne T, Brainard D, Byrne RW, et al. (2019) What is cognition? Current Biology 29:R608-R615. https://doi.org/10.1016/j.cub.2019.05.044 Chen D, Zhang L, Peng Y, et al. (2023) Exploring the association between social media and farmers' knowledge of a worldwide invasive agricultural pest, Spodoptera frugiperda (Lepidoptera: Noctuidae). Pest Management Science 80:678-686. https://doi.org/10.1002/ps.7796 Dart K, Latty T, Greenville A (2022) Citizen science reveals current distribution, predicted habitat suitability and resource requirements of the introduced African Carder Bee Pseudoanthidium ( Immanthidium ) repetitum in Australia. Biological Invasions 24:1827-1838. https://doi.org/10.1007/s10530-022-02753-2 Daume S (2016) Mining Twitter to monitor invasive alien species — An analytical framework and sample information topologies. Ecological Informatics 31:70-82. https://doi.org/10.1016/j.ecoinf.2015.11.014 Henke T, Novoa A, Bárðarson H, et al. (2024) Let’s talk aliens - Stakeholder perceptions of an alien species differ in time and space. NeoBiota 93:117-141. https://doi.org/10.3897/neobiota.93.117200 Izquierdo-Gómez D (2022) Synergistic use of facebook, online questionnaires and local ecological knowledge to detect and reconstruct the bioinvasion of the Iberian Peninsula by Callinectes sapidus Rathbun, 1896. Biological Invasions 24:1059-1082. https://doi.org/10.1007/s10530-021-02696-0 Kaplan AM, Haenlein M (2010) Users of the world, unite! The challenges and opportunities of Social Media. Business Horizons 53:59-68. https://doi.org/10.1016/j.bushor.2009.09.003 KAWO (2025) The ultimate guide to China social media platform 2025. KAWO, Shanghai, China Liu B (2012) Sentiment analysis and opinion mining. Springer, Cham, Switzerland Liu X, McGarrity ME, Bai C, et al. (2013) Ecological knowledge reduces religious release of invasive species. Ecosphere 4:art21. https://doi.org/10.1890/ES12-00368.1 Marcenò C, Padullés Cubino J, Chytrý M, et al. (2021) Facebook groups as citizen science tools for plant species monitoring. Journal of Applied Ecology 58:2018-2028. https://doi.org/10.1111/1365-2664.13896 Markell SG, Tylka GL, Anderson EJ, et al. (2020) Developing Public–Private Partnerships in plant pathology extension: Case studies and opportunities in the United States. Annual Review of Phytopathology 58:161-180. https://doi.org/10.1146/annurev-phyto-030320-041359 Miralles L, Dopico E, Devlo-Delva F, et al. (2016) Controlling populations of invasive pygmy mussel ( Xenostrobus securis ) through citizen science and environmental DNA. Marine Pollution Bulletin 110:127-132. https://doi.org/10.1016/j.marpolbul.2016.06.072 Oele DL, Wagner KI, Mikulyuk A, et al. (2015) Effecting compliance with invasive species regulations through outreach and education of live plant retailers. Biological Invasions 17:2707-2716. https://doi.org/10.1007/s10530-015-0907-2 Paini DR, Sheppard AW, Cook DC, et al. (2016) Global threat to agriculture from invasive species. Proceedings of the National Academy of Sciences 113:7575-9. https://doi.org/10.1073/pnas.1602205113 QuestMobile (2024) QuestMobile 2024 new media ecosystem report. QuestMobile, Beijing, China Schifani E, Paolinelli R (2018) Forums and social media help to discover exotic species in Europe and monitor their spread: the case of Exaireta spinigera (Wiedemann, 1830) (Diptera, Stratiomyidae) in the Italian peninsula and Sicily. Graellsia 74. https://doi.org/10.3989/graellsia.2018.v74.213 Schreck Reis C, Marchante H, Freitas H, et al. (2013) Public perception of invasive plant species: assessing the impact of workshop activities to promote young students' awareness. International Journal of Science Education 35:690-712. https://doi.org/10.1080/09500693.2011.610379 Schwoerer T, Schmidt JI, Davis TJ, et al. (2024) Human-connected wild lands: How network analysis can inform invasive species management. Biological Conservation 299. https://doi.org/10.1016/j.biocon.2024.110797 Shackleton RT, Adriaens T, Brundu G, et al. (2019a) Stakeholder engagement in the study and management of invasive alien species. Journal of Environmental Management 229:88-101. https://doi.org/10.1016/j.jenvman.2018.04.044 Shackleton RT, Richardson DM, Shackleton CM, et al. (2019b) Explaining people's perceptions of invasive alien species: A conceptual framework. Journal of Environmental Management 229:10-26. https://doi.org/10.1016/j.jenvman.2018.04.045 Soto I, Balzani P, Carneiro L, et al. (2024) Taming the terminological tempest in invasion science. Biological Reviews 99:1357-1390. https://doi.org/10.1111/brv.13071 Tian H, Gao C, Xiao X, et al. (2020) SKEP: Sentiment knowledge enhanced pre-training for sentiment analysis. Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. Association for Computational Linguistics, pp. 4067–4076 Viana JT, de Freitas JO, Marinho-Júnior M, et al. (2025) Using social media to identify records and commercialization of exotic Asian shrimp Penaeus monodon (Fabricius, 1798) In Brazil. Thalassas: An International Journal of Marine Sciences 41. https://doi.org/10.1007/s41208-025-00885-5 Wang D (2023) Research on dissemination strategies for youth online help-seeking behavior in the social media era. Journal of News Research 14:33-35. https://doi.org/10.3969/j.issn.1674-8883.2023.24.010 Wang L (2024) Current situation and future of legislation on precaution of alien species invasion in China. In: Antonelli G, Qin T, Ferroni MV and Erwin A (eds) Biodiversity Laws, Policies and Science in Europe, the United States and China. Springer Nature Switzerland, Cham, pp. 305-320 Werenkraut V, Baudino F, Roy HE (2020) Citizen science reveals the distribution of the invasive harlequin ladybird ( Harmonia axyridis Pallas) in Argentina. Biological Invasions 22:2915-2921. https://doi.org/10.1007/s10530-020-02312-7 Zahra A, Ma L, Khong KW (2025) Advancing sentiment analysis of social media data: Unveiling public perception of environmental challenges in Malaysia. In: Pon Selvan C, Sehgal N, Ruhela S and Rizvi NU (eds) International Conference on Innovation, Sustainability, and Applied Sciences. Springer Nature Switzerland, Cham, pp. 159-167 Zhang L, Li Y, Huang J, et al. (2019) Evaluation of the short-term and long-term performance of biological invasion management in the China-Myanmar border region. Journal of Environmental Management 240:1-8. https://doi.org/10.1016/j.jenvman.2019.03.061 Zhang Q, Wang Y, Liu X (2024) Risk of introduction and establishment of alien vertebrate species in transboundary neighboring areas. Nature Communications 15. https://doi.org/10.1038/s41467-024-45025-4 Supplementary Files AppendixS1Codes.pdf AppendixS2TableS1S4.pdf AppendixS3Datasets.xlsx Cite Share Download PDF Status: Published Journal Publication published 21 Apr, 2026 Read the published version in Biological Invasions → Version 1 posted Reviewers agreed at journal 29 Sep, 2025 Reviewers invited by journal 29 Sep, 2025 Editor invited by journal 27 Sep, 2025 Editor assigned by journal 10 Sep, 2025 First submitted to journal 10 Sep, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-7581230","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":522182668,"identity":"0d292ce7-d3e3-4f41-a296-fde226fd4de4","order_by":0,"name":"Liyun Zhang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA0UlEQVRIiWNgGAWjYBACPgYGNhAtB8SMBxgYEghrYYNqMWZgYGYgTUtiA/FaJNKfPfi5oza9XyL/wIEPFWkM/O3d+PWxSSSkG/aeOZ47c0Yyw8EZZ3IYJM6c3UBIyzEJ3rZjuRtuJDMc5m2rYDCQyCWkJbFN8m/bsXQDErQks0nzttUkQLXkEKGF5xmbtGzbAcOZPY8NgH5J4yHoF3729GeSb9vq5PnZEx8++FCRLMff3otfC4NAAog8DOfz4FcOtuYAiKwjrHAUjIJRMApGLgAABZ1G7F57h04AAAAASUVORK5CYII=","orcid":"https://orcid.org/0000-0002-0994-4994","institution":"Yunnan University","correspondingAuthor":true,"prefix":"","firstName":"Liyun","middleName":"","lastName":"Zhang","suffix":""},{"id":522182669,"identity":"89afa745-94ef-489a-a600-2389109b3c2c","order_by":1,"name":"Jie Huang","email":"","orcid":"","institution":"Yunnan University","correspondingAuthor":false,"prefix":"","firstName":"Jie","middleName":"","lastName":"Huang","suffix":""},{"id":522182670,"identity":"513a915c-7cfc-4eed-8d16-9d5aac20fb26","order_by":2,"name":"Xiaofei Liu","email":"","orcid":"https://orcid.org/0000-0002-1968-3644","institution":"Yunnan University","correspondingAuthor":false,"prefix":"","firstName":"Xiaofei","middleName":"","lastName":"Liu","suffix":""}],"badges":[],"createdAt":"2025-09-10 09:26:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7581230/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7581230/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s10530-026-03830-6","type":"published","date":"2026-04-21T15:57:41+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":93337340,"identity":"4382079e-607c-4cb5-97b5-b5a520951c7f","added_by":"auto","created_at":"2025-10-12 14:08:43","extension":"xml","order_by":2,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":9314,"visible":true,"origin":"","legend":"","description":"","filename":"binvBINVD2500678.xml","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/712b3a24001ec3fe8098b7b6.xml"},{"id":93337331,"identity":"7e252c7d-ea28-4b8e-9386-99a55baca852","added_by":"auto","created_at":"2025-10-12 14:08:42","extension":"xml","order_by":3,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":1098,"visible":true,"origin":"","legend":"","description":"","filename":"BINVD250067816541.go.xml","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/ae4f361174feb2235a8aac4a.xml"},{"id":93337314,"identity":"ee89438a-b9d0-4494-8010-5c5f2095ff3e","added_by":"auto","created_at":"2025-10-12 14:08:41","extension":"xml","order_by":4,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":766,"visible":true,"origin":"","legend":"","description":"","filename":"BINVD2500678Import.xml","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/ff04d657e0238dfcf24d15db.xml"},{"id":93337329,"identity":"7e0fb083-3912-47d2-b4e7-7412f53bb9ef","added_by":"auto","created_at":"2025-10-12 14:08:42","extension":"xml","order_by":8,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":91797,"visible":true,"origin":"","legend":"","description":"","filename":"BINVD25006780enriched.xml","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/b83cbdb95a2d0ff2339b81dc.xml"},{"id":93337344,"identity":"8b754ed3-14f1-4b80-9eba-9ab854538e66","added_by":"auto","created_at":"2025-10-12 14:08:43","extension":"jpeg","order_by":9,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":257368,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/ee042cd8e0d2206bb5e2e9e7.jpeg"},{"id":93337324,"identity":"b716f9d8-cd65-4a4a-a03d-180e4d1465db","added_by":"auto","created_at":"2025-10-12 14:08:41","extension":"jpeg","order_by":10,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":245632,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/9fc3de2edce0615431f9d826.jpeg"},{"id":93337343,"identity":"e789c5f1-599b-460c-8dab-5a25a6d33920","added_by":"auto","created_at":"2025-10-12 14:08:43","extension":"jpeg","order_by":11,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":611776,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage3.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/2336da78a310d811f7e54cb4.jpeg"},{"id":93337341,"identity":"08e46f83-fea9-487d-9c37-c6adf859be6b","added_by":"auto","created_at":"2025-10-12 14:08:43","extension":"jpeg","order_by":12,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":692206,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage4.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/99580a184da9d22d103f5f01.jpeg"},{"id":93337342,"identity":"2ae5411d-e580-4ff4-be74-65fe8351e259","added_by":"auto","created_at":"2025-10-12 14:08:43","extension":"jpeg","order_by":13,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":162224,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/9d9bc38f0cab2640a6149c2e.jpeg"},{"id":93337350,"identity":"53e288f2-8d5f-4555-ae57-f0902e89a402","added_by":"auto","created_at":"2025-10-12 14:08:44","extension":"jpeg","order_by":14,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":159720,"visible":true,"origin":"","legend":"","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/a920ff8b71df688b4254e414.jpeg"},{"id":93337325,"identity":"e9914451-cdb6-4b17-9b4f-22611a275df9","added_by":"auto","created_at":"2025-10-12 14:08:41","extension":"png","order_by":15,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":37094,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage1.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/fbcee5fd945bbb966b959d72.png"},{"id":93337338,"identity":"83c4368a-5cf8-4d80-9803-63fa2cd80ac8","added_by":"auto","created_at":"2025-10-12 14:08:42","extension":"png","order_by":16,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":35546,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/6281f770d774850031ea4856.png"},{"id":93337335,"identity":"3a970b18-d0e0-46de-87c5-d7cd287b43bf","added_by":"auto","created_at":"2025-10-12 14:08:42","extension":"png","order_by":17,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":75198,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/5922ea19e10a0791fda49e78.png"},{"id":93337347,"identity":"80ec75fa-0561-4849-a786-5a2634180b86","added_by":"auto","created_at":"2025-10-12 14:08:43","extension":"png","order_by":18,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":85289,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/2225e61c34a5c2fde8012f2a.png"},{"id":93338076,"identity":"0f84b23b-cb07-447f-a2bd-42e6f009514c","added_by":"auto","created_at":"2025-10-12 14:16:42","extension":"png","order_by":19,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":19378,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage5.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/97bb9d19cc1eb25653df937b.png"},{"id":93338082,"identity":"424a590e-7814-4c00-ae8f-d85ffb32536f","added_by":"auto","created_at":"2025-10-12 14:16:44","extension":"png","order_by":20,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":19935,"visible":true,"origin":"","legend":"","description":"","filename":"Onlinefloatimage6.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/bfee9db7c4b1cf41df00dbfb.png"},{"id":93337352,"identity":"22170330-60fb-46d2-8595-b0f0ecf1af93","added_by":"auto","created_at":"2025-10-12 14:08:44","extension":"xml","order_by":21,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":90025,"visible":true,"origin":"","legend":"","description":"","filename":"BINVD25006780structuring.xml","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/e541d9cf78d45b1d5d8287b2.xml"},{"id":93337330,"identity":"e193e2ef-0e38-442a-ad70-179d0e87ea88","added_by":"auto","created_at":"2025-10-12 14:08:42","extension":"html","order_by":22,"title":"","display":"","copyAsset":false,"role":"acdc-reference","size":99713,"visible":true,"origin":"","legend":"","description":"","filename":"earlyproof.html","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/52b877ca457127ffb5a8eb1d.html"},{"id":93337334,"identity":"200a9feb-7176-494e-8385-cc56fab7cec8","added_by":"auto","created_at":"2025-10-12 14:08:42","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":152787,"visible":true,"origin":"","legend":"\u003cp\u003eThe top 20 highest-frequency words and their categories regarding \u003cem\u003eSolidago canadensis\u003c/em\u003e, \u003cem\u003eSolenopsis invicta\u003c/em\u003e, and \u003cem\u003eTrachemys scripta\u003c/em\u003e based on TikTok. ID, identification; DA, distribution \u0026amp; abundance; IM, impact; MA, management; OT, others.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/00bce82959525179cf38871f.png"},{"id":93337328,"identity":"d4c0d19d-575e-4607-9111-d55bfcd3851b","added_by":"auto","created_at":"2025-10-12 14:08:41","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":145382,"visible":true,"origin":"","legend":"\u003cp\u003eThe top 20 highest-frequency words and their categories regarding \u003cem\u003eS. canadensis\u003c/em\u003e, \u003cem\u003eS. invicta\u003c/em\u003e, and \u003cem\u003eT. scripta\u003c/em\u003e based on Sina Microblog. ID, identification; DA, distribution \u0026amp; abundance; IM, impact; MA, management; OT, others.\u003c/p\u003e","description":"","filename":"2.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/24b4f33e39c9c831a1e024f1.png"},{"id":93337305,"identity":"a77923a2-966e-4e2b-b0a3-49c0dfe34391","added_by":"auto","created_at":"2025-10-12 14:08:41","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":274000,"visible":true,"origin":"","legend":"\u003cp\u003eThe top 20 most closely associated word pairs regarding \u003cem\u003eS. canadensis\u003c/em\u003e, \u003cem\u003eS. invicta\u003c/em\u003e, and \u003cem\u003eT. scripta\u003c/em\u003e based on TikTok\u003c/p\u003e","description":"","filename":"3.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/abeb87727061e01d0c0e0fd6.png"},{"id":93338079,"identity":"3bb3d65a-f85c-4753-9d47-32e07a7d65c5","added_by":"auto","created_at":"2025-10-12 14:16:43","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":305354,"visible":true,"origin":"","legend":"\u003cp\u003eThe top 20 most closely associated word pairs regarding \u003cem\u003eS. canadensis\u003c/em\u003e, \u003cem\u003eS. invicta\u003c/em\u003e, and \u003cem\u003eT. scripta\u003c/em\u003e based on Sina Microblog\u003c/p\u003e","description":"","filename":"4.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/e452ce9851d41ce9753006df.png"},{"id":93338080,"identity":"694f168f-87f8-453e-8557-7b4328ff0625","added_by":"auto","created_at":"2025-10-12 14:16:43","extension":"png","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":57591,"visible":true,"origin":"","legend":"\u003cp\u003eThe distribution of tendency indexes regarding \u003cem\u003eS. canadensis\u003c/em\u003e, \u003cem\u003eS. invicta\u003c/em\u003e, and \u003cem\u003eT. scripta\u003c/em\u003e based on TikTok\u003c/p\u003e","description":"","filename":"5.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/92e45c925d751915d8f2ee19.png"},{"id":93338078,"identity":"f7802d63-8a81-4999-abb9-8d270bca1359","added_by":"auto","created_at":"2025-10-12 14:16:43","extension":"png","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":58244,"visible":true,"origin":"","legend":"\u003cp\u003eThe distribution of tendency indexes regarding \u003cem\u003eS. canadensis\u003c/em\u003e, \u003cem\u003eS. invicta\u003c/em\u003e, and \u003cem\u003eT. scripta\u003c/em\u003e based on Sina Microblog\u003c/p\u003e","description":"","filename":"6.png","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/fa50a9b07fac844f7dfb4f44.png"},{"id":107927780,"identity":"efeb6d2d-3ea2-4389-bdda-55de09112bca","added_by":"auto","created_at":"2026-04-27 16:04:25","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1154473,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/d1b5a60a-07c0-4460-b004-d46aec868bb6.pdf"},{"id":93338077,"identity":"81081a5d-6375-47c2-b92d-5df9082091ae","added_by":"auto","created_at":"2025-10-12 14:16:42","extension":"pdf","order_by":5,"title":"","display":"","copyAsset":false,"role":"supplement","size":256069,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixS1Codes.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/e94f7f05502023d13dee8bd8.pdf"},{"id":93337336,"identity":"c9946309-3a9b-462e-a17b-8511693594fe","added_by":"auto","created_at":"2025-10-12 14:08:42","extension":"pdf","order_by":6,"title":"","display":"","copyAsset":false,"role":"supplement","size":67303,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixS2TableS1S4.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/ae8a971bbce03384f9667c91.pdf"},{"id":93337326,"identity":"3def7eb4-ff6f-4981-a110-9b4db1c8c34c","added_by":"auto","created_at":"2025-10-12 14:08:41","extension":"xlsx","order_by":7,"title":"","display":"","copyAsset":false,"role":"supplement","size":676632,"visible":true,"origin":"","legend":"","description":"","filename":"AppendixS3Datasets.xlsx","url":"https://assets-eu.researchsquare.com/files/rs-7581230/v1/40f3b4f61d4798f1b8573edc.xlsx"}],"financialInterests":"","formattedTitle":"Public cognition of invasive alien species in China","fulltext":[{"header":"1. Introduction","content":"\u003cp\u003eBiological invasions are inherently linked to human activities (Soto et al. \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), and the management of invasive alien species (IAS) requires consideration of the long-term and complex nature of human influence, necessitating broad public participation (Schwoerer et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Education and awareness campaigns can enhance public IAS cognition (Miralles et al. \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Oele et al. \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2015\u003c/span\u003e) and facilitate their understanding of relevant policies and measures, thereby improving engagement in management efforts (Schreck Reis et al. \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). Additionally, the role of education in shaping public cognition is crucial for the development of invasion biology as a discipline, since most foundational research has historically prioritized ecological dimensions over sociological insights (Abrahams et al. \u003cspan citationid=\"CR1\" class=\"CitationRef\"\u003e2019\u003c/span\u003e), particularly regarding public cognition of IAS derived from innovative platforms like social media.\u003c/p\u003e\u003cp\u003eUnlike traditional surveys which are often limited by sample size and geographic constraints, social media data offer a dynamic and cost-effective means of monitoring public sentiment over time (Zahra et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). The interactive and decentralized nature of social media facilitates rapid information sharing and dissemination, allowing scientific knowledge, policy updates, and public awareness campaigns to spread efficiently through multimedia content (e.g., videos, infographics) and user engagement mechanisms such as sharing and commenting (Dart et al. \u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Additionally, these platforms serve as valuable sources of public feedback which allow researchers to identify collective cognitions of IAS using data mining techniques (Daume \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). However, extracting relevant data requires overcoming challenges such as noise, misinformation, and selection biases. Advanced computational methods (e.g., natural language processing and machine learning algorithms) are required to automate content filtering and enhance analytical accuracy (Kaplan, Haenlein \u003cspan citationid=\"CR8\" class=\"CitationRef\"\u003e2010\u003c/span\u003e). Thus, researchers must address methodological complexities to ensure data reliability and validity.\u003c/p\u003e\u003cp\u003eChina represents a critical case in the global context of biological invasion, being simultaneously among the most IAS-vulnerable regions (Zhang et al. \u003cspan citationid=\"CR31\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and the greatest potential IAS sources worldwide (Paini et al. \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2016\u003c/span\u003e). Thus, effective IAS management requires substantial public participation, as successful strategies depend on community engagement in prevention, monitoring, and reporting (Zhang et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). However, public\u0026rsquo;s IAS awareness in China remains alarmingly low (Wang \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), as evidenced by irrational public conducts such as the widespread religious release of red-eared slider (\u003cem\u003eTrachemys scripta\u003c/em\u003e) (Liu et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2013\u003c/span\u003e). No studies have systematically explored how the general public views IAS on a national basis. Additionally, China has the largest social media user base globally, with 1.071\u0026nbsp;billion users accounting for 76% of its total population in 2024 (KAWO \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e), providing a vast data source for studying public cognition of IAS.\u003c/p\u003e\u003cp\u003eGiven this context, we hypothesize that public cognition of IAS on social media exhibits distinct cross-species disparities and cross-platform differences. To address this, we selected dominant Chinese social media platforms as data sources and focused on highly-discussed IAS as study targets. Through large-scale data collection, processing, and computational analysis, we aimed to reveal public cognitive characteristics toward IAS in China. The findings not only advance understanding of the socio-ecological dimensions of biological invasions but also offer scientific foundations for designing social media-based public education campaigns, thereby better supporting the formulation and implementation of IAS management strategies.\u003c/p\u003e"},{"header":"2. Materials and Methods","content":"\u003cdiv id=\"Sec3\" class=\"Section2\"\u003e\u003ch2\u003e2.1 Definitions and indicators\u003c/h2\u003e\u003cp\u003eCognition represents the mental processes through which individuals acquire, process, and utilize knowledge to understand and respond to their environment, encompassing perception, learning, memory, and decision-making (Bayne et al. \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). These mechanisms fundamentally shape both individual behaviors and collective societal actions, particularly in environmental contexts (Schwoerer et al. \u003cspan citationid=\"CR20\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In IAS management, public cognition plays a pivotal role by influencing awareness levels and behavioral responses toward IAS (Shackleton et al. \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2019a\u003c/span\u003e; Shackleton et al. \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2019b\u003c/span\u003e). To operationalize this construct, we quantified public cognition through three empirically validated indicators: (1) cognitive focus (the central themes of public attention regarding IAS), (2) cognitive association (the semantic relations linking these focal points to related concepts), and (3) cognitive tendency (the affective valence triggering potential management behaviors). Together, these metrics form a comprehensive framework for analyzing how public cognition evolves from initial attention (focus) to conceptual connections (association) and ultimately to evaluative judgments (tendency), providing valuable insights for science communication and policy development in biological invasion management.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\u003ch2\u003e2.2 Selection of social media platforms and IAS\u003c/h2\u003e\u003cp\u003eWe referred to relevant reports and selected TikTok and Sina Microblog as social media data sources in view of their user base, monthly active users, and content types. TikTok is China's largest short-video sharing platform, with 1,100\u0026nbsp;million monthly active users as of December 2024 (KAWO \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Similarly, Sina Microblog is the country's leading microblogging platform, publishing not only short videos but also images and text, with 590\u0026nbsp;million monthly active users up to the end of 2024 (KAWO \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eWe systematically retrieved 2,753 relevant posts from TikTok and Sina Microblog platforms using the following search terms: \"invasive organisms\", \"invasive species\", \"biological invasion\", \"alien invasion\", \"alien organisms\", and \"alien species\". From this dataset, we identified 695 posts containing specific IAS mentions. The five most frequently mentioned IAS were: \u003cem\u003eSolidago canadensis\u003c/em\u003e (n\u0026thinsp;=\u0026thinsp;69), \u003cem\u003eT. scripta\u003c/em\u003e (n\u0026thinsp;=\u0026thinsp;60), \u003cem\u003eEichhornia crassipes\u003c/em\u003e (n\u0026thinsp;=\u0026thinsp;57), \u003cem\u003eSolenopsis invicta\u003c/em\u003e (n\u0026thinsp;=\u0026thinsp;50), and \u003cem\u003ePomacea canaliculata\u003c/em\u003e (n\u0026thinsp;=\u0026thinsp;47). To assess public engagement, we selected the most engaged posts (by like counts) for each taxonomic group, including \u003cem\u003eS. canadensis\u003c/em\u003e (17,255 likes) for plants, \u003cem\u003eS. invicta\u003c/em\u003e (24,257 likes) for insects, and \u003cem\u003eT. scripta\u003c/em\u003e (1,812 likes) for animals.\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\u003ch2\u003e2.3 Data collection, preprocessing, and analysis\u003c/h2\u003e\u003cp\u003eThe fundamental principle of web crawlers is to automate the process of accessing web pages and extracting target information by simulating human browsing behavior. Specifically, a web crawler initiates from one or multiple seed URLs, sends HTTP requests to retrieve web content, and parses the HTML code to extract hyperlinks and other relevant data. During this process, newly discovered URLs are continuously added to a crawling queue, and the crawler follows a predefined strategy to select subsequent pages for scraping until termination conditions are met. The crawler was developed in PyCharm (JetBrains, Prague, Czech Republic) using Python 3.6.0 (Python Software Foundation, Beaverton, OR, USA), with Selenium WebDriver (ThoughtWorks, Chicago, IL, USA) employed for browser automation (see Appendix S1: Codes). Post-data acquisition, redundant information was filtered from the raw HTML source to retain specific content. The Beautiful Soup parser was utilized to transform HTML/XML documents into parse trees, enabling targeted data extraction through Python scripts designed for different webpage structures, particularly for collecting user comments. For multi-page comment sections, the crawler simulated pagination operations to ensure comprehensive data collection. Anti-crawler countermeasures (e.g., CAPTCHAs, IP blocking) were mitigated through strategic implementations including configuring appropriate request intervals, deploying proxy IP rotation, etc. Due to the time-intensive nature of web data crawling, processing, and analytical procedures, we implemented a phased data collection approach: TikTok platform data was initially crawled up to March 31, 2023, followed by subsequent collection of Sina Microblog platform data up to December 31, 2023. Additionally, the raw data obtained through web crawling typically contains substantial noise, including special characters, punctuation marks, and other non-content elements. To address this, we employed regular expressions (regex) - a powerful and flexible pattern-matching tool for text processing - to systematically filter and remove these extraneous elements from the extracted textual data. This preprocessing step utilized Python's built-in re module to implement sophisticated string matching and substitution operations, effectively cleansing the dataset while preserving meaningful content. The regex patterns were carefully designed to account for various character encodings and linguistic peculiarities encountered in the source material, ensuring comprehensive noise removal across diverse text formats.\u003c/p\u003e\u003cp\u003eCognitive focus was determined by analyzing high-frequency terms in social media user feedback. We extracted the top 20 most frequent words using Python's Jieba library for text segmentation (see Appendix S1: Codes), and categorized them into five thematic groups: Identification \u0026amp; Detection (ID), Distribution \u0026amp; Abundance (DA), Impact (IM), Management (MA), and Others (OT). To ensure statistical accuracy, we consolidated synonyms (e.g., merging \"roadside\" variants into a unified term), and applied stop-word lists (e.g., Harbin Institute of Technology and Baidu stop-word lists) to filter meaningless terms before frequency calculation.\u003c/p\u003e\u003cp\u003eCognitive association was quantitatively analyzed through term co-occurrence relationships in social media comments using Pointwise Mutual Information (PMI) (see Appendix S1: Codes). This statistical measure evaluates lexical association strength by comparing the observed joint probability of term pairs against their expected probabilities under independence assumption, computed as:\u003c/p\u003e\u003cp\u003ePMI(x,y)\u0026thinsp;=\u0026thinsp;log₂(P(x,y)/(P(x)\u0026times;P(y)))\u003c/p\u003e\u003cp\u003ewhere P(x,y) is joint probability of terms x and y co-occurring, and P(x) and P(y) are marginal probabilities of individual term occurrences. The interpretation follows that PMI\u0026thinsp;\u0026gt;\u0026thinsp;0 indicates semantic association, with magnitude reflecting strength. Our analysis protocol comprised: (1) selecting the lexicon from top 100 high-frequency terms in cognitive focus categories, (2) computing PMI values for all possible term pairs, and (3) identifying the 20 most strongly associated pairs (highest PMI) for cognitive pattern characterization.\u003c/p\u003e\u003cp\u003eWe quantified cognitive tendency through sentiment analysis of user comments using the open-source Senta system (Baidu, Beijing, China), which integrates RoBERTa (Robustly Optimized BERT Pretraining Approach) with SKEP (Sentiment Knowledge Enhanced Pre-training) for domain-specific sentiment analysis (see Appendix S1: Codes). The system operates by first leveraging RoBERTa's robust contextual understanding to process text, then applying SKEP's specialized sentiment knowledge through techniques including sentiment word masking, aspect-sentiment pair identification, and sentiment reconstruction objectives. This combined approach enables both sentence-level and aspect-level analysis, outputting a normalized sentiment polarity index from 0 (negative) to 1 (positive) that captures nuanced public cognition while maintaining the methodological rigor demonstrated in prior opinion mining research (Liu \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2012\u003c/span\u003e; Tian et al. \u003cspan citationid=\"CR24\" class=\"CitationRef\"\u003e2020\u003c/span\u003e).\u003c/p\u003e\u003c/div\u003e"},{"header":"3. Results","content":"\u003cdiv id=\"Sec7\" class=\"Section2\"\u003e\u003ch2\u003e3.1 Cognitive focus\u003c/h2\u003e\u003cp\u003eThe top 20 highest-frequency words and their categories based on TikTok are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig1\" class=\"InternalRef\"\u003e1\u003c/span\u003e (see Appendix S2: Table \u003cspan refid=\"MOESM1\" class=\"InternalRef\"\u003eS1\u003c/span\u003e). The cognitive focus on \u003cem\u003eS. canadensis\u003c/em\u003e is indicated by two ID terms (\"Canadian goldenrod\" [1486] and \"Seeds\" [296]), nine DA terms (\"Found everywhere\" [1013], \"Roadside\"[308], etc.), two IM terms (\"Toxic\" [202] and \"Biological invasion\" [182]), four MA terms (\"Research\" [279], \"Planting\" [257], etc.); and three OT terms (\"Desert\" [1124], \"Have seen\" [250], and \"Childhood\" [177]). Further, the focus is comprised mainly of DA (2983) and ID (1782), and rarely of OT (1301), MA (897), and IM (384). For \u003cem\u003eS. invicta\u003c/em\u003e, the cognitive focus covers four ID terms (\"Ant\" [2318], \"Red fire ant\" [1558], etc.), five DA terms (\"A lot\" [565], \"My home\" [226], etc.), eight IM terms (\"Got bitten\" [487], \"Be careful\" [389], etc.), two MA terms (\"Gasoline\" [365] and \"Boiling water\" [218]), and one OT term (\"Childhood\" [209]). Further, the focus is comprised mainly of ID (4455), IM (2154), and DA (1360), and rarely of MA (583) and OT (209). Similarly, the cognitive focus surrounding \u003cem\u003eT. scripta\u003c/em\u003e includes 11 ID terms (\"Brazilian turtle\" [2221], \"Turtle\" [1126], etc.), two DA terms (\"My home\" [849] and \"Deep water\" [145]), one IM term (\"Biological invasion\" [104]), five MA terms (\"Raise\" [344], \"What to do\" [236], etc.), and one OT term (\"Two\" [174]). Further, the focus is comprised mainly of ID (4468), and rarely of DA (994), MA (988), OT (174), and IM (104).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe top 20 highest-frequency words and their categories based on Sina Microblog are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig2\" class=\"InternalRef\"\u003e2\u003c/span\u003e (see Appendix S2: Table \u003cspan refid=\"MOESM2\" class=\"InternalRef\"\u003eS2\u003c/span\u003e). Public cognitive focus on \u003cem\u003eS. canadensis\u003c/em\u003e is indicated by five ID terms (\"Canadian goldenrod\" [205], \"Fresh flowers\" [123], etc.), six DA terms (\"A lot\" [73], \"Shanghai\"[48], etc.), five IM terms (\"Hu sheep\" [104], \"Biological invasion\" [75], etc.), and four MA terms (\"Planting\" [76], \"Sina Microblog\" [43], etc.). Further, the focus is comprised mainly of ID (480), and rarely of IM (311), DA (261), and MA (182). For \u003cem\u003eS. invicta\u003c/em\u003e, the cognitive focus covers five ID terms (\"Red fire ant\" [79], \"Ant\" [51], etc.), one DA term (\"A lot\" [40]), seven IM terms (\"Invasive species\" [100], \"Terrifying\" [38], etc.), six MA terms (\"Release\" [55], \"Introduction\" 40], etc.), and one OT term (\"Childhood\" [14]). Further, the focus is comprised mainly of IM (237), ID (197), and MA (179), and rarely of DA (40) and OT (23). Similarly, the cognitive focus surrounding \u003cem\u003eT. scripta\u003c/em\u003e includes eight ID terms (\"Brazilian turtle\" [63], \"Turtle\" [31], etc.), one DA term (\"My home\" [20]), five IM terms (\"Cute\" [27], \"Be careful\" [27], etc.), four MA terms (\"Release\" [96], \"Science\" [18], etc.), and two OT terms (\"Two\" [12] and \"Life\" [11]). Further, the focus is comprised mainly of ID (188), MA (142), and IM (113), and rarely of OT (23) and DA (20).\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec8\" class=\"Section2\"\u003e\u003ch2\u003e3.2 Cognitive association\u003c/h2\u003e\u003cp\u003eThe top 20 most closely associated word pairs based on TikTok are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig3\" class=\"InternalRef\"\u003e3\u003c/span\u003e (see Appendix S2: Table \u003cspan refid=\"MOESM3\" class=\"InternalRef\"\u003eS3\u003c/span\u003e). Public cognitive association on \u003cem\u003eS. canadensis\u003c/em\u003e is primarily highlighted by ecological and agricultural threats. Strong links between \"Fertility \u0026rarr; Overlord flower\" (7.87) and \"Native \u0026rarr; Crop\" (7.40) highlight concerns over its invasive impact on farmland. Ecological harm (\"Ecology \u0026rarr; Overlord flower\" [7.37]) and weed-like traits (5.59\u0026ndash;5.72) further dominate perceptions, while positive attributes (e.g., \"Medical value\" [5.65]) receive minimal attention. For \u003cem\u003eS. invicta\u003c/em\u003e, the cognitive association is primarily linked to health risks and folk remedies. Strong links between \"Wound \u0026rarr; Soapy water\" (5.23) and \"Redness \u0026rarr; Wound\" (4.29) highlight concerns over their painful stings, while home treatments like \"Effect \u0026rarr; Diesel\" (4.31) and \"Gasoline \u0026rarr; Diesel\" (3.99) reflect informal mitigation attempts. Mentions of \"Hospital \u0026rarr; Whole body\" (4.53) and \"Toxicity \u0026rarr; Child\" (3.78) underscore perceived dangers, particularly to children (\"Childhood \u0026rarr; Hometown\" [4.35]). The weak \"Childhood \u0026rarr; Dirt pile\" association (3.52) reflect limited experiential encounters with fire ant habitats in everyday environments. Health issues and morphological traits were identified as the primary aspects of cognitive association regarding \u003cem\u003eT. scripta\u003c/em\u003e. The strongest associations highlight eye problems (\"Eyes \u0026rarr; Can't open eyes\" [6.65]) and color variations (\"Albino \u0026rarr; Darkening\" [6.42]), suggesting widespread discussion of these visible abnormalities. Price-related associations (\"Price \u0026rarr; Mutation\" [5.52]; \"Pet \u0026rarr; Price\" [4.89]) indicate significant attention to the commercial aspects of color morphs. Health concerns appear fragmented, with notable links between \"Sick \u0026rarr; Neck\" (5.47) and \"Pneumonia \u0026rarr; Normal\" (4.92), while husbandry practices like \"Sunbathe \u0026rarr; Stone\" (5.46) and \"Shallow water \u0026rarr; Stone\" (4.58) reflect common captive management approaches.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe top 20 most closely associated word pairs based on Sina Microblog are shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig4\" class=\"InternalRef\"\u003e4\u003c/span\u003e (see Appendix S2: Table S4). Public cognitive association on \u003cem\u003eS. canadensis\u003c/em\u003e is primarily highlighted by agricultural and utilization contexts. The strongest associations highlight its potential as livestock feed (\"Digestion \u0026rarr; Sheep dung\" [6.70]; \"Seed \u0026rarr; Sheep dung\" [5.48]) and breeding value (\"Ability \u0026rarr; Breeding\" [6.40]; \"Nationwide \u0026rarr; Breeding\" [6.02]). Commercial aspects appear significant (\"Specialized \u0026rarr; Price\" [6.29]), while ecological concerns manifest through terms like \"Harm \u0026rarr; Growth\" (5.30) and \"Large area \u0026rarr; Wasteland\" (5.17). Notably, some users paradoxically view its spread as beneficial (\"Natural enemy \u0026rarr; Good thing\" [5.74]; \"Spread \u0026rarr; Good thing\" [5.13]), suggesting divided perceptions about this invasive species' impacts. For \u003cem\u003eS. invicta\u003c/em\u003e, the cognitive association is primarily linked to ecological threats and public safety concerns. The strongest association (\"Hollow \u0026rarr; Lotus seed\" [7.66]) suggests discussions often link fire ants with agricultural damage, particularly to lotus crops. Ecological impacts are emphasized through terms like \"Consequence \u0026rarr; Environment\" (5.92) and \"Harm \u0026rarr; Country\" (5.92), while public health concerns manifest in \"Blisters \u0026rarr; Redness\" (5.85). Notably, the ants are frequently characterized with negative anthropomorphic terms (\"Malicious \u0026rarr; Name\" [5.85]; \"Human \u0026rarr; Insidious\" [5.60]), reflecting strong public aversion. The association \"Helpless shrug \u0026rarr; Environment\" (5.66) reveals a sense of resignation regarding control efforts. Regulatory and ecological contexts were identified as the primary aspects of cognitive association regarding \u003cem\u003eT. scripta\u003c/em\u003e. The strongest associations (\"Sina \u0026rarr; Hot topic\" [7.76]; \"Hot topic \u0026rarr; Option\" [7.76]) highlight active policy debates about this invasive species. Ecological concerns appear through \"Heavy penalty \u0026rarr; Country\" (6.59) and \"Environment \u0026rarr; Harm\" (5.59), reflecting awareness of governmental controls and environmental impacts. Pet trade issues emerge in \"Organism \u0026rarr; Aquarium\" (6.59) and \"Animal \u0026rarr; Stray\" (5.65), indicating concerns about aquarium releases. The association \"Helpless shrug \u0026rarr; Hometown\" (5.85) suggests public resignation regarding established populations, while \"Turtle head \u0026rarr; Mouth\" (6.18) shows anatomical interest.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e\u003cdiv id=\"Sec9\" class=\"Section2\"\u003e\u003ch2\u003e3.3 Cognitive tendency\u003c/h2\u003e\u003cp\u003eThe distribution of tendency indexes based on TikTok is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig5\" class=\"InternalRef\"\u003e5\u003c/span\u003e (see Appendix S3: Datasets). \u003cem\u003eS. canadensis\u003c/em\u003e demonstrates the most polarized tendency distribution (median\u0026thinsp;=\u0026thinsp;0.63), with highly positive views in the upper quartile (0.94) contrasting sharply with extremely negative views in the lower quartile (0.06) (19,908 comments). \u003cem\u003eS. invicta\u003c/em\u003e receives the most negative overall tendency (median\u0026thinsp;=\u0026thinsp;0.23), though a small segment maintains neutral-to-positive views (upper quartile\u0026thinsp;=\u0026thinsp;0.85) (23,008 comments). \u003cem\u003eT. scripta\u003c/em\u003e shows intermediate but still negative-leaning tendency (median\u0026thinsp;=\u0026thinsp;0.43), with its upper quartile (0.92) approaching neutral-positive range (12,302 review comments). Notably, all species exhibit strong negative components (lower quartiles 0.02\u0026ndash;0.06), with \u003cem\u003eS. invicta\u003c/em\u003e being the most uniformly disliked and \u003cem\u003eS. canadensis\u003c/em\u003e generating the most divided public opinions among the three species on TikTok.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003cp\u003eThe distribution of tendency indexes based on Sina Microblog is shown in Fig.\u0026nbsp;\u003cspan refid=\"Fig6\" class=\"InternalRef\"\u003e6\u003c/span\u003e (see Appendix S3: Datasets). \u003cem\u003eS. canadensis\u003c/em\u003e demonstrates the most favorable overall tendency (median\u0026thinsp;=\u0026thinsp;0.89) with an upper quartile reaching 0.98, indicating a widespread positive view, though a lower quartile of 0.25 suggests a minority holding negative view (2,009 comments). \u003cem\u003eS. invicta\u003c/em\u003e shows more polarized tendencies (median\u0026thinsp;=\u0026thinsp;0.6) with near-universal positivity in the upper quartile (0.97) contrasting sharply with strong negativity in the lower quartile (0.05) (1,189 comments). \u003cem\u003eT. scripta\u003c/em\u003e exhibits similar polarization (median\u0026thinsp;=\u0026thinsp;0.65) with overwhelmingly positive tendency in the upper quartile (0.97) but retains some negative tendencies (lower quartile\u0026thinsp;=\u0026thinsp;0.09) (868 comments). Notably, all three species maintain high upper quartile values (0.97\u0026ndash;0.98), suggesting substantial positive tendency coexists with varying degrees of negative tendencies across these invasive species on Sina Microblog.\u003c/p\u003e\u003cp\u003e\u003c/p\u003e\u003c/div\u003e"},{"header":"4. Discussion","content":"\u003cp\u003ePublic cognition of IAS exhibited distinct patterns across species. For \u003cem\u003eS. canadensis\u003c/em\u003e, identification and distribution patterns dominated discussions, with TikTok emphasizing invasiveness while Sina Microblog focused on agricultural utility. \u003cem\u003eS. invicta\u003c/em\u003e cognition centered on morphological identification and health risks, with TikTok highlighting recognition/coping measures and Sina Microblog framing it as an ecological menace. \u003cem\u003eT. scripta\u003c/em\u003e discussions primarily involved identification traits, with TikTok addressing pet-related welfare issues and Sina Microblog focusing on regulatory policies. Sentiment analysis revealed species-specific patterns, with \u003cem\u003eS. canadensis\u003c/em\u003e showing platform-dependent polarization (TikTok) versus uniform positivity (Sina Microblog), \u003cem\u003eS. invicta\u003c/em\u003e exhibiting the most negative tendency on TikTok, and \u003cem\u003eT. scripta\u003c/em\u003e displaying moderate negativity on TikTok versus mixed positivity on Sina Microblog. The observed variations likely stem from species-specific traits and their direct human interactions. \u003cem\u003eS. canadensis\u003c/em\u003e, as a visible plant, draws attention to its ecological and agricultural impacts, while \u003cem\u003eS. invicta\u003c/em\u003e's health risks and \u003cem\u003eT. scripta\u003c/em\u003e's pet trade associations reflect their immediate human relevance. The dominance of identification features suggests public reliance on easily observable characteristics (Daume \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2016\u003c/span\u003e), whereas limited discussion of ecological impacts and management may indicate knowledge gaps (Liu et al. \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2013\u003c/span\u003e) or lower perceived urgency regarding long-term consequences (Zhang et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2019\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe platforms demonstrated markedly different discourse patterns. TikTok's discussions were more experiential and ecological, featuring extreme tendency polarization for \u003cem\u003eS. canadensis\u003c/em\u003e, strong negativity for \u003cem\u003eS. invicta\u003c/em\u003e, and moderate negativity for \u003cem\u003eT. scripta\u003c/em\u003e. In contrast, Sina Microblog exhibited policy-oriented discourse with consistently positive tendency. Platform-specific emphases emerged clearly: TikTok prioritized ecological dominance and health consequences, while Sina Microblog emphasized agricultural applications, public safety threats, and regulatory governance. This divergence suggests TikTok facilitates more personal, impact-focused narratives with emotional extremes, whereas Sina Microblog fosters more institutional, solution-oriented discussions with positive framing, particularly regarding policy interventions and agricultural utility. Additionally, both platforms shared a common underrepresentation of management strategies despite their differing thematic focuses. Platform disparities may arise from distinct user demographics and content formats. TikTok's experiential focus and emotional polarization likely reflect its short-form, engagement-driven content and its relatively older user base (QuestMobile \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), amplifying personal encounters with IAS. In contrast, Sina Microblog's policy-oriented, positive discourse aligns with its younger, more educated users and text-heavy format (QuestMobile \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2024\u003c/span\u003e), facilitating institutional discussions. The platforms' algorithmic preferences, shown by TikTok favoring emotional content and Sina Microblog promoting authoritative sources (Wang \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2023\u003c/span\u003e), further reinforce these differences in framing IAS issues.\u003c/p\u003e\u003cp\u003eBoth this study and prior research leverage social media as primary data sources (e.g., TikTok/Sina Microblog vs. Twitter/Facebook/iNaturalist), utilizing user-generated content to analyze public cognition of IAS (Daume \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Marcen\u0026ograve; et al. \u003cspan citationid=\"CR12\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Common methodologies include web crawling for data collection (Chen et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) and integration of citizen science principles to document species distributions (Viana et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Werenkraut et al. \u003cspan citationid=\"CR28\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). However, this study advances methodological rigor by employing semantic co-occurrence networks and quantile-based sentiment analysis, enabling granular detection of emotional polarization - unlike sentiment categorization (positive/neutral/negative) in earlier works (Zahra et al. \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Additionally, our systematic cross-platform comparison (TikTok vs. Sina Microblog) reveals algorithmic influences on discourse framing, whereas most literature focuses on single-platform data utility (Daume \u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2016\u003c/span\u003e; Izquierdo-G\u0026oacute;mez \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2022\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eConsistent with global studies, species-specific traits drive public cognition: \u003cem\u003eS. invicta\u003c/em\u003e\u0026rsquo;s health risks and \u003cem\u003eT. scripta\u003c/em\u003e\u0026rsquo;s pet trade associations mirror \"human-centric\" attention patterns observed for \u003cem\u003ePenaeus monodon\u003c/em\u003e\u0026rsquo;s commercial impacts in Brazil (Viana et al. \u003cspan citationid=\"CR25\" class=\"CitationRef\"\u003e2025\u003c/span\u003e). Both this study and literature confirm social media\u0026rsquo;s role in filling scientific monitoring gaps, such as undocumented distributions (Schifani, Paolinelli \u003cspan citationid=\"CR18\" class=\"CitationRef\"\u003e2018\u003c/span\u003e). Divergences emerge in identified cognitive blind spots: while studies like Dart et al. (\u003cspan citationid=\"CR4\" class=\"CitationRef\"\u003e2022\u003c/span\u003e) emphasize distribution data gaps, this study reveals pervasive neglect of management strategies across Chinese platforms, aligning with Zhang et al. (\u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2019\u003c/span\u003e)\u0026rsquo;s observation of \"passive strategies\" in border regions. Crucially, we expose platform algorithms as amplifiers of emotional extremes (e.g., TikTok\u0026rsquo;s engagement-driven content promoting polarization), whereas prior research attributes cognitive differences primarily to user demographics (Chen et al. \u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2023\u003c/span\u003e) or cultural contexts (Henke et al. \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe consistent underrepresentation of management strategies in public discourse (Zhang et al. \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2019\u003c/span\u003e) necessitates algorithm-optimized science communication, such as short videos demonstrating control measures. Cross-platform data integration - mirroring Izquierdo-G\u0026oacute;mez (\u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2022\u003c/span\u003e)\u0026rsquo;s triangulation of Facebook, questionnaires, and local knowledge - can mitigate single-platform biases (e.g., Sina Microblog\u0026rsquo;s policy focus overlooking ecological experiences). Furthermore, algorithmic transparency in content curation must be addressed through policy frameworks akin to public-private partnerships (Markell et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2020\u003c/span\u003e), given its demonstrated role in shaping polarized perceptions. This study underscores that algorithmic governance is as critical as species traits or user demographics in understanding contemporary IAS cognition.\u003c/p\u003e\u003cp\u003eTo enhance public education on IAS through social media, our findings suggest the need for platform-tailored science communication strategies that address cognitive gaps while leveraging algorithmic dynamics. Given TikTok's dominance in ecological experience-sharing and emotional polarization, authorities should collaborate with content creators to develop engaging short videos that visually demonstrate species identification, ecological impacts, and simple mitigation measures (e.g., safe removal of \u003cem\u003eS. canadensis\u003c/em\u003e or first aid for \u003cem\u003eS. invicta\u003c/em\u003e bites), capitalizing on the platform's preference for visceral narratives. For Sina Microblog, policy-aligned messaging should emphasize regulatory successes and agricultural co-benefits of IAS management, utilizing its text-heavy format to disseminate expert-endorsed guidelines and participatory monitoring initiatives. Both platforms must prioritize algorithmic transparency to counteract extreme sentiment amplification, potentially through partnerships with tech companies to promote authoritative content (Markell et al. \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). Crucially, educational campaigns should bridge the observed management strategy gap by integrating actionable advice (e.g., reporting protocols, biological controls) into trending formats - such as TikTok challenges or Sina Microblog Q\u0026amp;A threads - while addressing species-specific cognitive biases (e.g., \u003cem\u003eT. scripta\u003c/em\u003e's pet trade appeal). Cross-platform synergies, akin to citizen science triangulation (Izquierdo-G\u0026oacute;mez \u003cspan citationid=\"CR7\" class=\"CitationRef\"\u003e2022\u003c/span\u003e), could further enhance reach, combining TikTok's broad visibility with Sina Microblog's policy influence to foster a cohesive, evidence-based public discourse on IAS mitigation.\u003c/p\u003e"},{"header":"5. Conclusion","content":"\u003cp\u003eThis study demonstrates that public cognition of IAS in China exhibits distinct cross-species (\u003cem\u003eS. canadensis\u003c/em\u003e: ecological vs. agricultural focus; \u003cem\u003eS. invicta\u003c/em\u003e: health risks; \u003cem\u003eT. scripta\u003c/em\u003e: pet trade features) and cross-platform disparities (TikTok: experiential/polarized; Sina Microblog: policy/positive), shaped by species traits and algorithmic amplification of emotional or institutional narratives. Critically, the pervasive underrepresentation of management strategies across platforms necessitates integrating actionable control measures (e.g., reporting protocols, biological controls) into algorithm-optimized education - leveraging TikTok\u0026rsquo;s short videos for experiential learning and Sina Microblog\u0026rsquo;s text-based format for policy frameworks. Future research should prioritize cross-platform data triangulation (e.g., combining citizen science records with social media analytics) to mitigate sampling biases and explore algorithmic governance models that counter emotional extremism while promoting science communication. These steps are essential to transform fragmented public cognition into coordinated IAS management.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors thank Mr. Shengqing Yang (Yunnan University) and Ms. Jun Wan (Yunnan University) for their assistance in data collection and analyses. The authors are also grateful to the editor and anonymous reviewers for their invaluable contributions.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthors Contribution Statement\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eLiyun Zhang and Xiaofei Liu contributed to the study conception and design. Jie Huang conducted material preparation, field surveys, and data collection and analysis. Liyun Zhang wrote the first draft of this manuscript, and Jie Huang and Xiaofei Liu made comments. All authors have read and approved the final manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis work was supported by the National Key R\u0026amp;D Program of China (grant number 2024YFF1306700), the Yunnan International Joint Laboratory of Fruit-Vegetable-Flower Invasive Insect Pests Management (grant number 202303AP140018), and the National Natural Science Foundation of China (grant number 42161011).\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCode availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe codes we developed for data collection and analyses during the current study can be found online in the Supporting Material section at the end of this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting Interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors have no relevant financial or non-financial interests to disclose.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData Availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe datasets generated during and/or analyzed during the current study can be found online in the Supporting Material section at the end of this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics approval\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThis is an observational study and no ethical approval is required.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to participate\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eInformed consent was obtained from all individual participants included in the study.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsent to publish\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe authors affirm that human research participants provided informed consent for publication of the manuscript.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSupporting Information\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAdditional supporting information can be found online in the Supporting Material section at the end of this article.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003e\u003cbr\u003e\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003eAbrahams B, Sitas N, Esler KJ (2019) Exploring the dynamics of research collaborations by mapping social networks in invasion science. Journal of Environmental Management 229:27-37. https://doi.org/10.1016/j.jenvman.2018.06.051\u003c/li\u003e\n \u003cli\u003eBayne T, Brainard D, Byrne RW, et al. (2019) What is cognition? Current Biology 29:R608-R615. https://doi.org/10.1016/j.cub.2019.05.044\u003c/li\u003e\n \u003cli\u003eChen D, Zhang L, Peng Y, et al. (2023) Exploring the association between social media and farmers\u0026apos; knowledge of a worldwide invasive agricultural pest, \u003cem\u003eSpodoptera frugiperda\u003c/em\u003e (Lepidoptera: Noctuidae). Pest Management Science 80:678-686. https://doi.org/10.1002/ps.7796\u003c/li\u003e\n \u003cli\u003eDart K, Latty T, Greenville A (2022) Citizen science reveals current distribution, predicted habitat suitability and resource requirements of the introduced African Carder Bee\u003cem\u003e\u0026nbsp;Pseudoanthidium\u003c/em\u003e (\u003cem\u003eImmanthidium\u003c/em\u003e) \u003cem\u003erepetitum\u003c/em\u003e in Australia. Biological Invasions 24:1827-1838. https://doi.org/10.1007/s10530-022-02753-2\u003c/li\u003e\n \u003cli\u003eDaume S (2016) Mining Twitter to monitor invasive alien species \u0026mdash; An analytical framework and sample information topologies. Ecological Informatics 31:70-82. https://doi.org/10.1016/j.ecoinf.2015.11.014\u003c/li\u003e\n \u003cli\u003eHenke T, Novoa A, B\u0026aacute;r\u0026eth;arson H, et al. (2024) Let\u0026rsquo;s talk aliens - Stakeholder perceptions of an alien species differ in time and space. NeoBiota 93:117-141. https://doi.org/10.3897/neobiota.93.117200\u003c/li\u003e\n \u003cli\u003eIzquierdo-G\u0026oacute;mez D (2022) Synergistic use of facebook, online questionnaires and local ecological knowledge to detect and reconstruct the bioinvasion of the Iberian Peninsula by \u003cem\u003eCallinectes sapidus\u003c/em\u003e Rathbun, 1896. Biological Invasions 24:1059-1082. https://doi.org/10.1007/s10530-021-02696-0\u003c/li\u003e\n \u003cli\u003eKaplan AM, Haenlein M (2010) Users of the world, unite! The challenges and opportunities of Social Media. Business Horizons 53:59-68. https://doi.org/10.1016/j.bushor.2009.09.003\u003c/li\u003e\n \u003cli\u003eKAWO (2025) The ultimate guide to China social media platform 2025. \u0026nbsp; KAWO, Shanghai, China\u003c/li\u003e\n \u003cli\u003eLiu B (2012) Sentiment analysis and opinion mining. Springer, Cham, Switzerland\u003c/li\u003e\n \u003cli\u003eLiu X, McGarrity ME, Bai C, et al. (2013) Ecological knowledge reduces religious release of invasive species. Ecosphere 4:art21. https://doi.org/10.1890/ES12-00368.1\u003c/li\u003e\n \u003cli\u003eMarcen\u0026ograve; C, Padull\u0026eacute;s Cubino J, Chytr\u0026yacute; M, et al. (2021) Facebook groups as citizen science tools for plant species monitoring. Journal of Applied Ecology 58:2018-2028. https://doi.org/10.1111/1365-2664.13896\u003c/li\u003e\n \u003cli\u003eMarkell SG, Tylka GL, Anderson EJ, et al. (2020) Developing Public\u0026ndash;Private Partnerships in plant pathology extension: Case studies and opportunities in the United States. Annual Review of Phytopathology 58:161-180. https://doi.org/10.1146/annurev-phyto-030320-041359\u003c/li\u003e\n \u003cli\u003eMiralles L, Dopico E, Devlo-Delva F, et al. (2016) Controlling populations of invasive pygmy mussel (\u003cem\u003eXenostrobus securis\u003c/em\u003e) through citizen science and environmental DNA. Marine Pollution Bulletin 110:127-132. https://doi.org/10.1016/j.marpolbul.2016.06.072\u003c/li\u003e\n \u003cli\u003eOele DL, Wagner KI, Mikulyuk A, et al. (2015) Effecting compliance with invasive species regulations through outreach and education of live plant retailers. Biological Invasions 17:2707-2716. https://doi.org/10.1007/s10530-015-0907-2\u003c/li\u003e\n \u003cli\u003ePaini DR, Sheppard AW, Cook DC, et al. (2016) Global threat to agriculture from invasive species. Proceedings of the National Academy of Sciences 113:7575-9. https://doi.org/10.1073/pnas.1602205113\u003c/li\u003e\n \u003cli\u003eQuestMobile (2024) QuestMobile 2024 new media ecosystem report. \u0026nbsp; QuestMobile, Beijing, China\u003c/li\u003e\n \u003cli\u003eSchifani E, Paolinelli R (2018) Forums and social media help to discover exotic species in Europe and monitor their spread: the case of \u003cem\u003eExaireta spinigera\u003c/em\u003e (Wiedemann, 1830) (Diptera, Stratiomyidae) in the Italian peninsula and Sicily. Graellsia 74. https://doi.org/10.3989/graellsia.2018.v74.213\u003c/li\u003e\n \u003cli\u003eSchreck Reis C, Marchante H, Freitas H, et al. (2013) Public perception of invasive plant species: assessing the impact of workshop activities to promote young students\u0026apos; awareness. International Journal of Science Education 35:690-712. https://doi.org/10.1080/09500693.2011.610379\u003c/li\u003e\n \u003cli\u003eSchwoerer T, Schmidt JI, Davis TJ, et al. (2024) Human-connected wild lands: How network analysis can inform invasive species management. Biological Conservation 299. https://doi.org/10.1016/j.biocon.2024.110797\u003c/li\u003e\n \u003cli\u003eShackleton RT, Adriaens T, Brundu G, et al. (2019a) Stakeholder engagement in the study and management of invasive alien species. Journal of Environmental Management 229:88-101. https://doi.org/10.1016/j.jenvman.2018.04.044\u003c/li\u003e\n \u003cli\u003eShackleton RT, Richardson DM, Shackleton CM, et al. (2019b) Explaining people\u0026apos;s perceptions of invasive alien species: A conceptual framework. Journal of Environmental Management 229:10-26. https://doi.org/10.1016/j.jenvman.2018.04.045\u003c/li\u003e\n \u003cli\u003eSoto I, Balzani P, Carneiro L, et al. (2024) Taming the terminological tempest in invasion science. Biological Reviews 99:1357-1390. https://doi.org/10.1111/brv.13071\u003c/li\u003e\n \u003cli\u003eTian H, Gao C, Xiao X, et al. (2020) SKEP: Sentiment knowledge enhanced pre-training for sentiment analysis. \u0026nbsp;Proceedings of the 58th Annual Meeting of the Association for Computational Linguistics. \u0026nbsp;Association for Computational Linguistics, pp. 4067\u0026ndash;4076\u003c/li\u003e\n \u003cli\u003eViana JT, de Freitas JO, Marinho-J\u0026uacute;nior M, et al. (2025) Using social media to identify records and commercialization of exotic Asian shrimp \u003cem\u003ePenaeus monodon\u003c/em\u003e (Fabricius, 1798) In Brazil. Thalassas: An International Journal of Marine Sciences 41. https://doi.org/10.1007/s41208-025-00885-5\u003c/li\u003e\n \u003cli\u003eWang D (2023) Research on dissemination strategies for youth online help-seeking behavior in the social media era. Journal of News Research 14:33-35. https://doi.org/10.3969/j.issn.1674-8883.2023.24.010\u003c/li\u003e\n \u003cli\u003eWang L (2024) Current situation and future of legislation on precaution of alien species invasion in China. In: Antonelli G, Qin T, Ferroni MV and Erwin A (eds) Biodiversity Laws, Policies and Science in Europe, the United States and China. \u0026nbsp;Springer Nature Switzerland, Cham, pp. 305-320\u003c/li\u003e\n \u003cli\u003eWerenkraut V, Baudino F, Roy HE (2020) Citizen science reveals the distribution of the invasive harlequin ladybird (\u003cem\u003eHarmonia axyridis\u003c/em\u003e Pallas) in Argentina. Biological Invasions 22:2915-2921. https://doi.org/10.1007/s10530-020-02312-7\u003c/li\u003e\n \u003cli\u003eZahra A, Ma L, Khong KW (2025) Advancing sentiment analysis of social media data: Unveiling public perception of environmental challenges in Malaysia. In: Pon Selvan C, Sehgal N, Ruhela S and Rizvi NU (eds) International Conference on Innovation, Sustainability, and Applied Sciences. \u0026nbsp;Springer Nature Switzerland, Cham, pp. 159-167\u003c/li\u003e\n \u003cli\u003eZhang L, Li Y, Huang J, et al. (2019) Evaluation of the short-term and long-term performance of biological invasion management in the China-Myanmar border region. Journal of Environmental Management 240:1-8. https://doi.org/10.1016/j.jenvman.2019.03.061\u003c/li\u003e\n \u003cli\u003eZhang Q, Wang Y, Liu X (2024) Risk of introduction and establishment of alien vertebrate species in transboundary neighboring areas. Nature Communications 15. https://doi.org/10.1038/s41467-024-45025-4\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":true,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"biological-invasions","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"binv","sideBox":"Learn more about [Biological Invasions](https://www.springer.com/journal/10530)","snPcode":"10530","submissionUrl":"https://submission.nature.com/new-submission/10530/3","title":"Biological Invasions","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Alien species, Public cognition, Cross-platform disparities, Social media platform, Species-specific traits, Algorithmic biases","lastPublishedDoi":"10.21203/rs.3.rs-7581230/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7581230/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003ePublic understanding is critical for invasive alien species (IAS) management, yet national-scale cognition patterns remain understudied. We analyzed TikTok and Sina Microblog data to assess public cognition in China regarding three IAS: \u003cem\u003eSolidago canadensis\u003c/em\u003e, \u003cem\u003eSolenopsis invicta\u003c/em\u003e, and \u003cem\u003eTrachemys scripta\u003c/em\u003e. Three dimensions of cognition, including focus (high-frequency terms), associations (term-pair linkages), and tendencies (sentiment polarity), were quantified using word frequency, semantic co-occurrence networks, and quantile-based sentiment analyses, respectively. Results revealed cross-species disparities: On TikTok, \u003cem\u003eS. canadensis\u003c/em\u003e was concerned with its invasiveness (e.g., \"Found everywhere\", \"Fertility \u0026rarr; Overlord flower\"), while Sina Microblog emphasized its agricultural utility (e.g., \"Hu Sheep\", \"Digestion \u0026rarr; Sheep dung\"). \u003cem\u003eS. invicta\u003c/em\u003e discussions centered on recognition/coping measures on TikTok (e.g., \"Red fire ant\", \"Wound \u0026rarr; Soapy water\") versus ecological impacts on Sina Microblog (e.g., \"Invasive species\", \"Consequence \u0026rarr; Environment\"), and \u003cem\u003eT. scripta\u003c/em\u003e focused on pet raising issues (TikTok) (e.g., \"My home\", \"Eyes \u0026rarr; Can't open eyes\") and regulatory policies (Sina Microblog) (e.g., \"Release\", \"Heavy penalty \u0026rarr; Country\"). These disparities likely stem from species-specific traits and their interactions with human. Cross-platform differences were pronounced: TikTok\u0026rsquo;s experiential narratives showed polarized sentiment (\u003cem\u003eS. canadensis\u003c/em\u003e median\u0026thinsp;=\u0026thinsp;0.63; lower/upper quartiles\u0026thinsp;=\u0026thinsp;0.06/0.94), while Sina Microblog\u0026rsquo;s policy-driven discourse was predominantly positive (medians: 0.60\u0026ndash;0.89). Algorithmic biases amplified these differences: prioritizing emotional content (TikTok) versus institutional narratives (Sina Microblog). A key gap was the scarcity of management strategies, limiting public engagement. Our findings advocate for platform-specific interventions (e.g., TikTok demos of removal techniques) and algorithmic transparency to improve IAS management.\u003c/p\u003e","manuscriptTitle":"Public cognition of invasive alien species in China","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-10-12 14:08:31","doi":"10.21203/rs.3.rs-7581230/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"reviewerAgreed","content":"","date":"2025-09-29T12:02:14+00:00","index":0,"fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-09-29T10:45:48+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"Biological Invasions","date":"2025-09-27T12:41:44+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-09-10T12:32:40+00:00","index":"","fulltext":""},{"type":"submitted","content":"Biological Invasions","date":"2025-09-10T05:25:36+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"biological-invasions","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"binv","sideBox":"Learn more about [Biological Invasions](https://www.springer.com/journal/10530)","snPcode":"10530","submissionUrl":"https://submission.nature.com/new-submission/10530/3","title":"Biological Invasions","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"d142fe40-145a-4e38-a8a8-d7083065c5b1","owner":[],"postedDate":"October 12th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2026-04-27T16:02:07+00:00","versionOfRecord":{"articleIdentity":"rs-7581230","link":"https://doi.org/10.1007/s10530-026-03830-6","journal":{"identity":"biological-invasions","isVorOnly":false,"title":"Biological Invasions"},"publishedOn":"2026-04-21 15:57:41","publishedOnDateReadable":"April 21st, 2026"},"versionCreatedAt":"2025-10-12 14:08:31","video":"","vorDoi":"10.1007/s10530-026-03830-6","vorDoiUrl":"https://doi.org/10.1007/s10530-026-03830-6","workflowStages":[]},"version":"v1","identity":"rs-7581230","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7581230","identity":"rs-7581230","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}