The ecological abstraction–ecological association continuum: reframing biodiversity information infrastructures for human–nature reconnection

1 Introduction

1.1 Ecological disconnection in the Anthropocene

Humans evolved within ecological systems and remain fundamentally dependent upon them. Every society relies on biodiversity and ecosystem processes for food production, climate regulation, nutrient cycling, disease regulation, and cultural well-being. Yet despite this dependence, many people increasingly experience nature as something separate from everyday life rather than as the ecological foundation upon which human existence depends (Leopold, 1968).

Over the past several decades, scholars from ecology, environmental humanities, education, psychology, and conservation science have documented growing forms of human-nature disconnection. One of the earliest and most influential descriptions came from Pyle (1978), who introduced the concept of the “extinction of experience” to describe the progressive loss of direct encounters with nature. Subsequent research has shown that opportunities for routine interactions with biodiversity have declined across many regions as urbanization, technological mediation, and changing lifestyles reshape human environments (Pergams and Zaradic, 2008).

The consequences of this disconnection extend beyond recreation or personal well-being. Environmental educators have argued that declining ecological experiences can reduce ecological literacy and weaken understanding of environmental systems (Orr, 1992). Similarly, conservation researchers have suggested that reduced contact with biodiversity may diminish environmental concern and stewardship behaviors over time (Miller, 2005).

Recent work has reinforced these concerns. Soga and Gaston (2016) argue that declining interactions with nature create a feedback loop in which reduced experiences decrease appreciation for biodiversity, thereby reducing support for conservation. Studies of human-nature relationships similarly indicate that individuals who maintain stronger connections with nature tend to exhibit higher levels of environmental responsibility and pro-conservation behavior (Nisbet et al., 2009).

At the same time, environmental change continues to accelerate globally. Defaunation, habitat loss, climate change, pollution, and biological invasions are transforming ecosystems at unprecedented rates (Dirzo et al., 2014). The recent global biodiversity assessment conducted by IPBES concluded that nature is declining globally at rates unprecedented in human history, with approximately one million species threatened with extinction (Díaz et al., 2019).

The challenge facing contemporary societies is therefore not simply a biodiversity crisis. It is also a relationship crisis. This relationship crisis also reflects a deeper conceptual separation of the human animal from the ecological systems of which it remains a biological, cultural, and evolutionary participant. Human societies increasingly depend on ecological systems that many people rarely experience directly and often struggle to understand as interconnected systems. This growing disconnect raises important questions about how ecological understanding is formed and maintained in the modern world.

1.2 Human-nature relationships and ecological understanding

Environmental philosophers and social ecologists have long argued that ecological problems cannot be separated from the ways humans conceptualize their relationships with nature. Leopold’s land ethic proposed that humans should be viewed as members of ecological communities rather than external managers of nature (Leopold, 1968). Later, Naess (1973) argued that ecological sustainability requires recognizing the intrinsic interconnectedness of all living systems.

These ideas have been expanded by scholars who critique the dualistic separation of humans and nature that characterizes many modern societies. Plumwood (2002) argued that Western traditions frequently construct nature as an external object rather than a community of relationships in which humans participate. Similarly, Ingold (2011) emphasized that human life is continuously shaped through interactions with environments rather than existing independently from them.

Recent research has increasingly focused on measuring and understanding these relationships empirically. Schultz (2002) demonstrated that environmental attitudes are influenced by how individuals perceive themselves in relation to nature. The Nature Relatedness framework developed by Nisbet et al. (2009) further showed that stronger connections with nature are associated with greater environmental concern and ecological behavior.

A growing body of literature now recognizes human-nature connection as a multidimensional phenomenon involving emotional, cognitive, cultural, ethical, and experiential dimensions (Ives et al., 2017). Rather than viewing ecological relationships solely through the lens of resource use, these approaches emphasize belonging, participation, stewardship, and reciprocity.

Conservation science has increasingly incorporated similar perspectives. The concept of relational values emphasizes the importance of relationships among people, places, species, and ecosystems as foundations for conservation decision-making (Chan et al., 2016). Likewise, assessments of nature’s contributions to people have highlighted the importance of cultural, social, and relational dimensions of biodiversity alongside material benefits (Díaz et al., 2019).

Together, these perspectives suggest that ecological understanding depends not only on information about biodiversity but also on the relationships through which biodiversity becomes meaningful.

1.3 Information abundance and ecological distance

Paradoxically, ecological disconnection has intensified during a period characterized by unprecedented growth in environmental information. Modern societies possess more information about biodiversity than at any previous point in history. Scientific observations, environmental monitoring systems, remote sensing platforms, genomic databases, and biodiversity repositories collectively generate vast quantities of ecological data (Edwards et al., 2000).

The digitization of biodiversity information has transformed scientific research. Natural history collections, which historically required physical access, are increasingly available through digital infrastructures that support global discovery and analysis (Nelson and Ellis, 2019). Biodiversity occurrence records can now be accessed through global biodiversity repositories that integrate observations from natural history collections, monitoring programs, and community science initiatives worldwide (Robertson et al., 2014; Soberón and Peterson, 2004).

These developments have enabled new forms of ecological research. Biodiversity data are routinely integrated across institutions, countries, and disciplines to support large-scale analyses of ecological change, species distributions, conservation priorities, and environmental forecasting (Graham et al., 2004).

Yet information abundance does not necessarily reduce ecological distance. Several scholars have argued that modern technologies increasingly mediate human experience through representations rather than direct encounters (Borgmann, 2007). Environmental information is often experienced through interfaces, visualizations, databases, and digital platforms rather than through direct ecological engagement.

Floridi (2014) argues that digital information systems increasingly shape how humans perceive and interact with reality. In biodiversity contexts, organisms are often encountered through records, images, metadata, and analytical outputs. These representations are enormously valuable, but they can also create forms of abstraction that separate information from the ecological relationships from which it originates.

Digitization should therefore be understood not only as a technical conversion but also as a representational transformation. Decisions concerning standards, metadata, interfaces, and interoperability determine which aspects of specimens, observations, places, and ecological relationships remain visible as biodiversity information moves across infrastructures (Anderson et al., 2020; Gadelha et al., 2021; Hardisty and Roberts, 2013). The usefulness and interpretability of biodiversity data depend not only on technical quality but also on whether their contextual completeness is adequate for particular uses (Veiga et al., 2017). Abstraction enables scale, interoperability, and synthesis, but it can become ecologically limiting when standardized records are detached from the provenance and relational context required for interpretation.

Consequently, societies may become simultaneously more informed about biodiversity and more distant from direct ecological experience. Understanding this paradox requires examining the infrastructures through which biodiversity knowledge is created, organized, and communicated. Within the Ecological Abstraction–Ecological Association Continuum, this shift toward increasingly decontextualized representations reflects movement toward the abstraction end of the spectrum.

1.4 Ecological re-identification: a theory of human–nature reconnection

A central implication of the framework developed in this paper is that biodiversity information infrastructures can influence not only what people know about nature, but also how they perceive their own place within ecological systems. In this paper, we use ecological re-identification to describe a shift in self-perception through which individuals come to recognize themselves as biological participants embedded within ecological systems rather than as external observers or managers. One potential outcome of ecological association is support for this process of re-identification (Barragan‐Jason et al., 2022). Such recognition is inherently relational; it depends on an individual’s ability to perceive connections to other organisms, places, and the environmental processes that sustain life. Ecological re-identification therefore involves more than generalized awareness of nature; it is more likely to emerge through sustained engagement with particular organisms, places, communities, and ecological relationships (Christens et al., 2025; Riechers et al., 2021).

Biodiversity information infrastructures, as systems of representation, can either obscure or make these connections visible (Anderson et al., 2020). Infrastructures emphasizing ecological abstraction may, in some contexts, reinforce the perception of nature as a collection of external objects. In contrast, infrastructures that preserve ecological context, provenance, place-based relationships, and opportunities for participation may provide cognitive and affective scaffolding for ecological re-identification (Orr, 1992). By enabling people to see not only what exists but also how organisms, environments, places, and people are connected, these systems may support a stronger sense of participation within ecological systems.

These infrastructure characteristics do not automatically produce ecological re-identification, nor does the process necessarily follow a fixed or unidirectional sequence. Rather, they create conditions that may strengthen ecological understanding, perceived interdependence, stewardship, and enduring human–nature relationships (Barragan‐Jason et al., 2022). Within the hypothesis advanced in this paper, biodiversity information infrastructures positioned closer to ecological association are therefore expected to be more likely to support ecological re-identification than infrastructures emphasizing ecological abstraction.

1.5 Biodiversity infrastructure and the digital representation of nature

Biodiversity information infrastructures have emerged as critical components of contemporary environmental science. These infrastructures encompass natural history collections, biodiversity databases, standards, repositories, digitization programs, cyberinfrastructure initiatives, and increasingly AI-enabled, interoperable knowledge systems that support machine-actionable biodiversity research (Bakış, 2026; Bisby, 2000).

Natural history collections represent one of the oldest and most influential biodiversity infrastructures. Collections preserve physical evidence of organisms, environments, and historical ecological conditions across centuries (Suarez and Tsutsui, 2004). More recently, mass digitization initiatives have expanded access to these resources by converting specimens and associated materials into globally accessible digital assets (Beaman and Cellinese, 2012).

The emergence of biodiversity informatics has further accelerated this transformation. Biodiversity informatics focuses on the acquisition, integration, management, and analysis of biodiversity information across distributed systems (Soberón and Peterson, 2004). Large-scale infrastructures such as GBIF, VertNet, FishNet2, OBIS, and iDigBio now support unprecedented levels of biodiversity data sharing and interoperability (Constable et al., 2010; FishNet2, 2026; GBIF, 2026; IDigBio, 2026; Klein et al., 2019).

Recent developments further extend these infrastructures beyond data aggregation toward AI-ready research ecosystems that integrate digitized specimens, derivative data products, provenance, environmental context, and analytical workflows into interoperable, machine-actionable biodiversity information systems (Bakış, 2026).

Artificial intelligence represents the latest stage in this evolution. Machine learning systems increasingly support species identification, image analysis, literature mining, ecological forecasting, and biodiversity synthesis (Christin et al., 2019). Biodiversity information is becoming not only digitally accessible but also computationally actionable through AI-enabled biodiversity information infrastructures that integrate digitized specimens, images, environmental observations, publications, provenance, and analytical workflows into interoperable, machine-actionable research ecosystems (Bakış, 2026; Pollock et al., 2025).

These developments have generated extraordinary opportunities for science and conservation. However, they also raise important questions about representation. How biodiversity is represented within information infrastructures influences how biodiversity is perceived, interpreted, and understood.

1.6 Representation shapes relationship

Scholars in information science have long recognized that information systems are not neutral containers of knowledge. Classification systems, metadata standards, databases, and information infrastructures shape how information is organized and interpreted (Bowker and Star, 2000).

Leonelli (2016) argues that contemporary biology increasingly depends upon data infrastructures that mediate scientific understanding. Similarly, Latour (2007) emphasizes that scientific knowledge emerges through networks of representations that connect observations, institutions, technologies, and people.

These insights are particularly relevant for such infrastructures, where every record reflects decisions about what information is preserved, standardized, or omitted (Anderson et al., 2020; Hardisty and Roberts, 2013). Taxonomies, coordinates, images, and metadata emphasize different aspects of biodiversity and shape what remains visible. Consequently, organisms may be represented either as participants in ecological and evolutionary relationships or primarily as data points and computational objects (Jordano, 2021). As biodiversity information infrastructures become increasingly important for research and public engagement, these representational choices carry significant implications for how biodiversity is understood.

In this sense, these infrastructures can either reinforce the perception that humans stand outside ecological systems or support ecological re-identification by helping people recognize themselves as participants within multispecies, place-based, and social-ecological relationships. Rather than serving solely as repositories of biodiversity information, they can become systems of representation that shape not only ecological understanding, but also how people perceive their own place within the living systems they study, document, and depend upon. Through this process, they provide a pathway from ecological understanding toward ecological re-identification, creating a stronger foundation for stewardship, and enduring human-nature relationships.

The principal theoretical contribution of this paper is the proposition that the Ecological Abstraction–Ecological Association Continuum explains how biodiversity information infrastructures support ecological re-identification by shaping whether people perceive themselves as external observers of biodiversity or as participants within ecological systems. This paper therefore advances the hypothesis that infrastructures preserving ecological context, provenance, place-based relationships, and participatory engagement are more likely to support ecological re-identification, thereby strengthening ecological understanding, stewardship, and human–nature relationships.

1.7 The ecological abstraction–ecological association continuum

This paper develops the Ecological Abstraction–Ecological Association Continuum, a conceptual framework describing how biodiversity information infrastructures shape ecological understanding along a spectrum from ecological abstraction to ecological association. The continuum provides a lens for understanding how choices about representation, context, provenance, participation, and information design influence relationships between people and the living world, building upon previous work demonstrating the importance of representation, provenance, and relational perspectives in biodiversity information infrastructures (Anderson et al., 2020; Chan et al., 2016; Hardisty and Roberts, 2013).

At one end of the continuum, ecological abstraction occurs when biodiversity information becomes progressively detached from the ecological, cultural, historical, and place-based relationships that give it meaning (Bowker and Star, 2000; Leonelli, 2016). Organisms become records, localities become coordinates, and ecological systems become collections of information optimized for storage, exchange, and analysis.

At the other end, ecological association occurs when these infrastructures preserve and communicate ecological, historical, provenance, and social relationships alongside information (Chan et al., 2016; Riechers et al., 2021). Organisms remain connected to places, ecosystems, environmental histories, communities, and systems of stewardship. Biodiversity information serves not only as data but also as a mechanism for ecological understanding. Figure 1 introduces the Ecological Abstraction–Ecological Association Continuum by illustrating the two conceptual endpoints and the contrasting pathways through which biodiversity information systems influence ecological understanding and human-nature relationships.

Beyond describing differences among biodiversity information infrastructures, the Ecological Abstraction–Ecological Association Continuum provides a conceptual explanation for how biodiversity information infrastructures support ecological re-identification. By positioning infrastructures along a spectrum according to the extent that they preserve ecological context, provenance, place-based relationships, participation, and relational meaning, the framework explains why some infrastructures are more likely than others to help people recognize themselves as participants within ecological systems rather than as external observers of biodiversity.

2 Ecological abstraction in biodiversity systems

2.1 Representation and the making of biodiversity knowledge

It is important to acknowledge that abstraction is not only inevitable but essential for biodiversity science. Scientific inquiry depends on the ability to simplify complex realities into forms that can be compared, analyzed, and integrated across scales. Biodiversity informatics would be impossible without standards that transform heterogeneous observations into interoperable records, or without models that generalize across species and ecosystems (Edwards et al., 2000; Soberón and Peterson, 2004). Abstraction enables the discovery of large-scale patterns, the integration of data across institutions, and the development of conservation tools that operate at global scales.

The critical question is not whether abstraction occurs, but whether biodiversity information infrastructures preserve sufficient relational context to maintain ecological meaning alongside analytical utility. As the following sections argue, abstraction is most problematic when it operates in isolation from the ecological, historical, and place-based relationships that give biodiversity information its significance. The goal is not to eliminate abstraction but to balance it with ecological association, preserving contextual richness where it matters most for ecological understanding and stewardship.

Biodiversity knowledge is inseparable from representation. Organisms, ecosystems, and ecological processes cannot be observed simultaneously across all places and times. Instead, knowledge about biodiversity is generated through systems of representation that include specimens, field observations, taxonomic classifications, photographs, illustrations, maps, databases, and increasingly, digital infrastructures. These representations enable biodiversity to become visible, measurable, communicable, and actionable across scientific, educational, and policy contexts (Latour, 1987; Leonelli, 2016).

Representation is a necessary condition for ecological knowledge. Scientific understanding depends upon the ability to simplify complex realities into forms that can be observed, compared, classified, and communicated. Species become taxonomic entities, localities become geographic coordinates, habitats become categories, and ecological processes become measurable variables. Such abstractions enable large-scale biodiversity synthesis and conservation planning while making ecological phenomena accessible beyond direct experience (Borgman, 2017).

Yet representation is never neutral. Every representation emphasizes certain aspects of reality while reducing or excluding others. Classification systems determine what distinctions matter. Data standards determine which attributes are preserved. Analytical models determine which variables become visible and which remain hidden. As a result, systems of representation influence not only what is known about biodiversity but also how biodiversity is perceived and understood (Daston, 2017; Hacking, 1999).

This influence becomes increasingly important as human experiences of nature become more mediated. For many individuals, biodiversity is encountered primarily through digital representations rather than through direct ecological engagement. The characteristics of those representations therefore shape how organisms, ecosystems, and ecological relationships are imagined and valued.

2.2 Datafication and the transformation of living systems

The rapid growth of biodiversity informatics has accelerated a broader societal process known as datafication, the transformation of complex phenomena into structured and computationally manageable forms (Kitchin, 2014; Mayer-Schönberger and Cukier, 2014). Through digitization, standardization, and data integration, biodiversity increasingly becomes accessible as digital information.

These developments have transformed biodiversity science. Global infrastructures now integrate billions of occurrence records, specimen records, images, genomic sequences, environmental measurements, and trait observations originating from thousands of institutions worldwide (Nelson and Ellis, 2019). Such resources have expanded opportunities for ecological modeling, biodiversity assessment, conservation planning, and environmental forecasting at unprecedented scales.

However, datafication also alters how biodiversity is represented. Organisms become records, images become datasets, traits become variables, and ecosystems become collections of interoperable data objects. Rich ecological realities are translated into forms optimized for storage, retrieval, exchange, and computation. Information that cannot be easily standardized may become less visible within these representations.

This process is not unique to biodiversity science. Scholars have observed similar transformations across numerous domains where digital systems increasingly mediate human understanding of complex realities (Borgmann, 2007; Floridi, 2014). Information systems excel at organizing and transmitting knowledge, but they can also encourage perceptions that privilege discrete informational units over relationships, histories, and contexts. Within biodiversity information infrastructures, this tendency may contribute to viewing organisms as isolated data objects rather than participants in ecological systems.

2.3 Ecological distance and the loss of relational context

A growing body of research suggests that modern societies are experiencing increasing forms of ecological distance. Urbanization, technological mediation, and reduced opportunities for everyday encounters with biodiversity contribute to what has been described as an extinction of experience, a decline in routine human interactions with nature that may weaken ecological awareness and stewardship (Gaston and Soga, 2020; Miller, 2005; Soga and Gaston, 2016).

Biodiversity information infrastructures do not create this phenomenon, but they operate within it. As ecological knowledge becomes increasingly mediated through databases, models, portals, and digital interfaces, the preservation of contextual relationships becomes increasingly important. Within the Ecological Abstraction–Ecological Association Continuum, this shift toward increasingly decontextualized representations reflects movement toward the abstraction end of the spectrum. A specimen record may retain species identity while losing habitat context. A biodiversity image may preserve morphology while omitting ecological interactions. A species distribution map may reveal geographic patterns while obscuring environmental histories and human relationships with place.

Environmental philosophers and scholars in the environmental humanities have argued that modern forms of knowledge often separate humans from ecological systems by emphasizing observation, classification, and management while downplaying participation, reciprocity, and relationship (Cronon, 1992; Plumwood, 2002). These concerns are particularly relevant to biodiversity information infrastructures because they influence how biodiversity is represented to scientists, conservation practitioners, educators, and the public.

The consequence is not necessarily misinformation or scientific error. Rather, it is the gradual reduction of relational context. Biodiversity remains visible, but the ecological relationships that connect organisms to places, histories, communities, and environmental processes may become progressively less apparent.

2.4 Artificial intelligence and new forms of abstraction

Artificial intelligence introduces additional layers of representation between ecological reality and human understanding. Contemporary machine learning systems rely on highly structured representations including images, labels, annotations, traits, environmental variables, and occurrence records. Organisms are transformed into computational objects that can be classified, compared, and predicted at scales beyond human capacity (Wäldchen and Mäder, 2018).

The benefits of these technologies are substantial. AI systems are increasingly used for automated species identification, camera-trap analysis, bioacoustic monitoring, ecological forecasting, biodiversity assessment, and conservation decision support (Norouzzadeh et al., 2018; Tuia et al., 2022; Weinstein, 2018). These applications expand the reach of biodiversity science and enable new forms of ecological discovery.

At the same time, AI systems operate on representations that have already undergone multiple layers of abstraction. Organisms become image embeddings, feature vectors, training examples, or classification targets. These representations are optimized for prediction and recognition rather than for preserving ecological meaning. As a result, AI can amplify existing tendencies toward abstraction if biodiversity information infrastructures focus exclusively on machine readability and computational performance.

The challenge is therefore not whether AI should be used in biodiversity science. Rather, it is how these infrastructures can retain ecological relationships, contextual understanding, and relational meaning within increasingly computational forms of representation.

2.5 Ecological abstraction

Within the Ecological Abstraction–Ecological Association Continuum, ecological abstraction represents one endpoint of the spectrum, where biodiversity representations progressively lose the ecological, historical, cultural, and place-based relationships that give them meaning. It occurs when biodiversity is encountered primarily as information rather than as relationships, as inventory rather than as ecological communities, or as computational resources rather than as living systems.

Ecological abstraction is not inherently negative. Scientific inquiry necessarily relies on abstraction, and biodiversity information infrastructures provide indispensable capabilities for conservation, ecological research, and scientific discovery. Biodiversity informatics depends on standards, classifications, and models that transform heterogeneous observations into forms that can be compared, integrated, and analyzed across institutions and scales (Edwards et al., 2000; Hardisty and Roberts, 2013; Soberón and Peterson, 2004). The critical question is therefore not whether abstraction occurs, but whether systems of representation preserve the relationships that connect biodiversity information to ecological reality.

When locality becomes merely a coordinate, habitat becomes merely metadata, and organisms become merely records, opportunities for relational understanding may diminish. Conversely, such infrastructures that preserve ecological context, environmental history, provenance, and place-based meaning help maintain connections between biodiversity knowledge and ecological experience. Within the Ecological Abstraction–Ecological Association Continuum, this movement toward preserving relationships defines ecological association, which is developed in the following sections.

Recent syntheses have highlighted that biodiversity science remains characterized by fragmented data sources, incomplete knowledge of traits, species interactions, ecosystem functions, and substantial challenges in discovering, reusing, and integrating heterogeneous information types (Poisot et al., 2019; Pollock et al., 2025). These limitations illustrate that increasing volumes of biodiversity information do not necessarily translate into a coherent understanding of ecological relationships, underscoring the need for biodiversity information infrastructures that preserve relational context alongside increasingly machine-actionable data.

3 Museum collections as ecocultural memory

3.1 Beyond archives: collections as relationship repositories

Natural history collections are traditionally viewed as scientific archives that preserve specimens for taxonomic, systematic, ecological, and evolutionary research. Their scientific importance is unquestionable. Collections document biodiversity across centuries, provide material for species discovery and identification, support environmental monitoring, and preserve irreplaceable records of life on Earth (Pyke and Ehrlich, 2010; Rocha et al., 2014; Suarez and Tsutsui, 2004). Yet focusing exclusively on their scientific utility understates their broader significance. Digitized natural history collections increasingly function as a globally distributed research and educational infrastructure, sometimes described as the “Global Museum” (Bakker et al., 2020). This distributed infrastructure extends the scientific and public value of individual collections by connecting specimens, associated data, institutions, and users across geographic and disciplinary boundaries.

Every specimen embodies a network of relationships. It records an organism at a particular place and time while preserving traces of ecological conditions, collecting practices, institutional histories, and human experiences of place (Bennett, 2004; Daston, 2004).

Collections function as repositories of ecological and cultural memory as much as repositories of objects, preserving connections among organisms, places, communities, institutions, and generations of researchers. Within the Ecological Abstraction–Ecological Association Continuum, museum collections exemplify ecological association by preserving relational context that might otherwise be lost as biodiversity becomes increasingly digitized and computationally mediated. Rather than reducing organisms to isolated records, they maintain connections between organisms and their broader ecological, historical, and cultural contexts.

As biodiversity is increasingly encountered through digitized collections and online platforms, these representational choices become increasingly important. Figure 2 illustrates how museum collections function as ecocultural memory systems by preserving relationships among organisms, places, people, institutions, and ecological history across successive stages of curation and knowledge generation.

3.2 Collections as ecological memory

Natural history collections are among humanity’s most important repositories of ecological memory. Ecological memory refers to the persistence of information, experiences, and biological records that connect present ecosystems with past ecological conditions (Barthel et al., 2010; Folke et al., 2001). Collections preserve such memory by documenting biodiversity across temporal scales that far exceed individual human lifetimes.

Specimens provide direct evidence of historical species distributions, environmental conditions, phenological events, genetic variation, morphological change, contaminant exposure, and community composition (Holmes et al., 2016; Meineke et al., 2019; Shaffer et al., 1998). In many cases, collections preserve the only remaining evidence of ecological states that no longer exist. They allow scientists to reconstruct changing ecosystems, track biological responses to environmental change, and establish historical baselines for conservation.

Ecological memory extends beyond biological records. Field notes, specimen labels, photographs, and archival materials document how previous generations encountered and understood ecological systems, preserving both biological and cultural memory.

3.3 Place, locality, and environmental change

One of the most powerful aspects of collections is their ability to preserve connections between biodiversity and place. Every specimen originates from a locality. Although localities are frequently represented as coordinates within digital systems, they represent far more than geographic positions. Localities connect organisms to habitats, ecosystems, environmental conditions, land-use histories, and human communities.

Place-based information is increasingly important as global environmental change accelerates. Historical collections have become essential resources for understanding shifts in species distributions, climate responses, habitat transformation, invasive species dynamics, extinction risk, and ecological resilience (Heberling and Isaac, 2017; Tingley and Beissinger, 2009). By linking organisms to specific places and times, collections provide opportunities to reconstruct ecological change across decades or centuries.

Environmental change is not experienced solely through ecological processes. It is also experienced through changing human relationships with landscapes. As habitats are transformed, urbanized, fragmented, restored, or conserved, collections preserve evidence of both ecological and social histories. Locality records therefore function as bridges connecting biodiversity, environmental change, and human experience.

3.4 Collections, public engagement, and ecological reconnection

Museums have increasingly embraced roles that extend beyond research and preservation. Contemporary natural history institutions are important centers for public education, biodiversity communication, environmental literacy, and conservation engagement (Ballard et al., 2017; Cook et al., 2014). Through exhibitions, outreach programs, community science initiatives, and digital platforms, collections provide opportunities for people to encounter biodiversity in meaningful ways.

Digitization has dramatically expanded these opportunities. Millions of specimen records, images, field notes, and collection objects are now accessible through online portals and biodiversity platforms (Hedrick et al., 2020; Nelson and Ellis, 2019). This accessibility allows individuals to engage with biodiversity across geographic and institutional boundaries. At the same time, digitization introduces a challenge central to this paper: whether digital representations preserve relationships or merely information.

The value of digitized collections depends on whether contextual information remains visible alongside specimen records. Collections are uniquely positioned to support ecological reconnection because they preserve evidence of ecological change and human engagement across time. They demonstrate that biodiversity is not simply a catalog of species but a history of interactions among organisms, environments, and people. Through these preserved relationships, collections help transform biodiversity information into ecological understanding.

3.5 Collections as ecocultural memory systems

Viewed through the framework developed in this paper, natural history collections are best understood not simply as scientific archives but as ecocultural memory systems. They preserve ecological information, environmental histories, cultural practices, scientific knowledge, and human encounters with biodiversity within integrated records that span generations.

Collections preserve ecological information, cultural memory, and historical continuity across generations. As biodiversity increasingly becomes mediated through digital infrastructures, the role of collections becomes even more important. They provide a counterbalance to ecological abstraction by preserving relational context and ecological memory. In doing so, collections demonstrate how biodiversity information infrastructures can move along the Ecological Abstraction–Ecological Association Continuum by functioning not merely as repositories of information, but as mechanisms for maintaining and restoring relationships between human societies and the living world.

4 Provenance as relationship preservation

4.1 Beyond technical metadata

Within biodiversity informatics, provenance is typically discussed as information describing the origin, history, ownership, custody, and transformation of data, specimens, and digital products. Provenance enables transparency, reproducibility, attribution, quality assessment, and workflow documentation, making it a foundational component of trustworthy scientific infrastructures (Buneman et al., 2001; Cheney et al., 2009; Ram and Liu, 2007; Simmhan et al., 2005). Provenance records help answer questions about who collected a specimen, where and when it was collected, how it was processed, what transformations it underwent, and how it ultimately became part of scientific and digital information systems.

While these functions are essential, they represent only part of provenance’s significance. Provenance does more than document the history of information. It preserves the relationships through which knowledge is produced. Every biodiversity record emerges from interactions among organisms, environments, places, collectors, institutions, technologies, and communities. Provenance captures traces of these interactions, preserving connections that might otherwise disappear as biodiversity information moves through increasingly complex infrastructures (Moreau and Groth, 2013).

Viewed through this lens, provenance becomes more than metadata. It becomes a mechanism for preserving meaning. Biodiversity records without provenance risk becoming detached informational objects whose origins, contexts, and relationships are difficult to reconstruct. In contrast, records with rich provenance maintain links to the ecological, historical, and social circumstances that produced them.

This distinction is central to the argument developed in this paper. If ecological abstraction occurs when biodiversity becomes detached from relational context, provenance represents one of the primary mechanisms through which those relationships can be maintained.

4.2 Provenance and relational continuity

Provenance can be understood as a form of ecological memory. Just as museum collections preserve biological evidence of past ecosystems, provenance preserves information about how ecological knowledge was generated, interpreted, and transmitted through time.

Ecological systems are fundamentally relational. Organisms exist within networks of interactions involving habitats, environmental conditions, evolutionary histories, species assemblages, and human activities. Biodiversity records capture only fragments of these broader systems. Provenance helps preserve links between those fragments and the ecological realities from which they originate (Moreau and Groth, 2013).

A specimen record, for example, may include locality descriptions, collection dates, habitat information, collecting methods, associated species, field observations, collector identities, and curatorial histories. Individually, these elements appear as metadata. Collectively, they document a network of relationships connecting an organism to a place, a time, an ecosystem, and a chain of human interactions (Hardisty et al., 2022). Provenance therefore functions as a relational bridge between biodiversity information and ecological reality.

This role becomes increasingly important as biodiversity information is digitized, aggregated, integrated, and reused. Data infrastructures excel at moving information across systems, but mobility can come at the cost of contextual richness. Records that travel far from their origins may gradually lose connections to the circumstances under which they were created. Provenance mitigates this loss by preserving pathways back to ecological, historical, and institutional contexts.

From this perspective, provenance is not merely a record of information flow. It is a record of ecological encounters. It preserves traces of relationships among organisms, environments, and people that remain attached to biodiversity information long after the original encounter has passed.

4.3 Place, history, and relational context

One of the most important contributions of provenance is its ability to preserve connections to place. Biodiversity information derives much of its meaning from locality and environmental context. Organisms do not exist independently of places; they are embedded within ecosystems, landscapes, climatic conditions, and environmental histories.

Yet many digital systems represent locality primarily through coordinates or administrative units. Although such representations are scientifically useful, they often provide limited insight into the broader ecological and cultural significance of place. Provenance enriches locality by documenting the circumstances surrounding collection and observation, including habitat descriptions, environmental conditions, historical land use, collection motivations, and associated observations.

This contextual richness becomes particularly valuable when studying environmental change. Historical biodiversity records gain interpretive power when provenance preserves information about the conditions under which organisms were encountered. Provenance therefore supports not only scientific reproducibility but also ecological understanding by maintaining connections among biodiversity, place, and environmental history.

Relational context extends beyond ecological conditions. Biodiversity records are also products of human histories. Collectors, researchers, institutions, expeditions, and communities all contribute to the creation and interpretation of biodiversity knowledge. Provenance preserves these contributions, making visible the social and historical dimensions of scientific information that might otherwise remain hidden.

4.4 Provenance, stewardship, and accountability

Stewardship depends upon understanding relationships and responsibilities. Biodiversity information is not produced in isolation; it emerges through networks of labor, expertise, institutional support, environmental engagement, and long-term care. Provenance provides a mechanism for documenting these networks and making them visible.

Recent discussions of FAIR data principles have emphasized the importance of transparency, traceability, and reproducibility within scientific infrastructures (Wilkinson et al., 2016). Provenance is fundamental to achieving these goals because it enables users to understand where information originates, how it has been transformed, and what assumptions may influence its interpretation.

However, accountability extends beyond technical reproducibility. Provenance can also support ethical stewardship by preserving awareness of the relationships that underpin biodiversity knowledge. When records remain connected to their origins, users are better positioned to recognize responsibilities associated with data interpretation, environmental management, conservation decision-making, and knowledge reuse.

This perspective shifts provenance from a purely technical concern to a relational one. Provenance does not simply document workflows; it documents responsibilities. It helps maintain awareness that biodiversity information originates from real organisms, real places, real communities, and real ecological systems.

4.5 Indigenous data governance and relational integrity

The relational significance of provenance becomes particularly evident in discussions of Indigenous data governance. Many Indigenous knowledge systems emphasize relationships among people, lands, waters, species, ancestors, and future generations. Knowledge is often understood not as a detached informational resource but as something embedded within ongoing relationships and responsibilities (Berkes, 2012; Whyte, 2018).

These perspectives challenge conventional approaches that prioritize unrestricted information mobility while paying limited attention to context, authority, and relational obligations. The CARE Principles for Indigenous Data Governance - Collective Benefit, Authority to Control, Responsibility, and Ethics - were developed in part to address these concerns by ensuring that data remain connected to the communities and relationships from which they originate (Carroll et al., 2020).

Although biodiversity information infrastructures operate across diverse scientific and cultural contexts, the underlying lesson is broadly relevant. Information cannot be fully separated from the relationships that produce it. Preserving relational integrity requires maintaining connections among knowledge, communities, places, and ecological systems.

Provenance provides one practical mechanism for doing so. By documenting origins, histories, responsibilities, and contexts, provenance helps prevent biodiversity information from becoming detached from the relationships that give it meaning.

4.6 Provenance as relationship preservation

The conventional view of provenance emphasizes transparency, reproducibility, and data quality. These functions remain critically important. However, the framework developed in this paper suggests a broader interpretation.

It preserves relationships between organisms and places, specimens and ecosystems, observations and histories, communities and knowledge, researchers and institutions, and past ecological conditions and present understanding. Through these preserved connections, provenance helps maintain relational meaning as biodiversity information moves through scientific and digital infrastructures.

Viewed in this way, provenance represents one of the most important mechanisms through which biodiversity information infrastructures can resist ecological abstraction. It helps ensure that biodiversity information remains connected to the ecological, historical, cultural, and social relationships from which it emerges. In doing so, provenance contributes to ecological association by preserving pathways through which people can understand biodiversity not merely as information, but as evidence of living relationships. Within the Ecological Abstraction–Ecological Association Continuum, provenance functions as one of the principal mechanisms through which biodiversity information infrastructures move toward ecological association by preserving relational continuity across people, places, organisms, and ecological processes.

5 AI-ready biodiversity infrastructure: risks and opportunities

5.1 Artificial intelligence as a mediator of ecological knowledge

Artificial intelligence is rapidly becoming a primary mediator between people and biodiversity knowledge. Increasingly, ecological information is encountered not through direct engagement with organisms, landscapes, or natural history collections, but through digital systems that identify species, summarize information, generate explanations, recommend observations, and interpret environmental data (Floridi, 2014). As a result, AI systems are beginning to shape not only biodiversity analysis but also public and scientific perceptions of biodiversity.

This transformation represents a significant shift in the production and communication of ecological knowledge. Historically, ecological understanding emerged through direct observation, field experience, local knowledge, natural history practice, and long-term engagement with places and organisms. Today, computational systems increasingly participate in these processes by filtering information, prioritizing observations, and generating representations of ecological reality (Leonelli, 2016).

Viewing AI as a mediator of ecological knowledge shifts attention from technological capability to how AI-mediated representations shape ecological understanding. Within the Ecological Abstraction–Ecological Association Continuum, artificial intelligence is best understood as a mediating technology rather than an inherently abstracting or associating one. Depending on how biodiversity information infrastructures preserve ecological context, provenance, and relational meaning, AI can move representations toward either ecological abstraction or ecological association.

5.2 Practical risks: misidentification and data quality

AI-enabled identification can compromise biodiversity data quality when taxonomic complexity exceeds what images alone can resolve. Cryptic, poorly resolved, and morphologically complex taxa may depend on subtle or internal diagnostic characters that are not visible in photographs. Image-based systems can therefore produce confident but incorrect identifications, particularly for specimens outside their training distribution (Campos et al., 2025). AI should augment rather than replace expert taxonomic judgment in such cases.

Once incorporated into biodiversity databases, erroneous identifications can propagate through global infrastructures and affect species distribution models, ecological analyses, conservation assessments, and subsequent research (Pollock et al., 2025). Verification becomes more difficult when records lack voucher specimens, which remain essential for re-examining diagnostic characters and preserving the reproducibility and long-term scientific value of biodiversity data (Rocha et al., 2014; Suarez and Tsutsui, 2004).

Generative AI creates an additional risk by producing realistic images that may not correspond to actual organisms, specimens, or observations. When synthetic or modified images enter biodiversity systems without clear disclosure and verifiable provenance, they may be mistaken for authentic evidence and weaken trust in biodiversity documentation (Campos et al., 2023). Preserving links to voucher material, provenance records, verification procedures, and explicit documentation of AI involvement is therefore essential.

These risks show that ecological abstraction can have practical as well as conceptual consequences. Machine-learning systems may also reproduce geographic, taxonomic, and temporal biases, while uncertainty remains difficult to communicate reliably in many applications (Pollock et al., 2025). Questions of accountability, transparency, and oversight further require governance structures that support human verification, uncertainty communication, and institutional responsibility (Dignum, 2019; UNESCO, 2021).

5.3 Opportunities for ecological understanding and ecological literacy

Artificial intelligence offers significant opportunities for expanding understanding of ecological relationships. Species-identification systems based on computer vision have dramatically lowered barriers to biodiversity observation, allowing individuals with little formal training to engage with local ecosystems through photographs and mobile applications (Wäldchen and Mäder, 2018). These tools can transform casual encounters with organisms into opportunities for ecological learning.

AI also enables new forms of environmental interpretation. Machine-learning systems can detect patterns within large ecological datasets, identify species in images and acoustic recordings, monitor environmental change, and support conservation planning (Christin et al., 2019). By revealing ecological patterns that would otherwise remain difficult to perceive, AI can contribute to ecological understanding rather than merely automate analysis. AI may also contribute to ecological literacy by helping users explore ecological complexity, species relationships, and environmental change (Orr, 1992).

These possibilities suggest that AI can support movement toward ecological association within the Ecological Abstraction–Ecological Association Continuum when it helps reveal relationships, dependencies, and ecological interconnectedness rather than merely providing information about individual organisms.

5.4 Risks of ecological abstraction

Despite these opportunities, AI systems can also reinforce ecological abstraction. Machine learning requires biodiversity to be represented in forms that are computationally tractable. Organisms become labels, images become training examples, habitats become variables, and ecological processes become statistical patterns. Although such abstractions are necessary for computation, they can obscure ecological meaning when detached from broader contexts (Leonelli, 2016).

Recent reviews of AI in ecology and biodiversity science have identified recurring challenges associated with applying machine-learning methods to ecological problems, including insufficient integration of ecological domain knowledge, uncertainty estimation, ecological context, and quality control (Pollock et al., 2025; Tuia et al., 2022). These observations reinforce the need for biodiversity information infrastructures that treat AI not merely as a mechanism for classification, prediction, or automation, but as a mediator of ecological knowledge whose outputs remain connected to organisms, places, interactions, uncertainty, and ecological processes.

Information infrastructures frequently privilege what can be standardized, measured, and processed efficiently. Bowker and Star (2000) demonstrated that classification systems inevitably emphasize certain forms of knowledge while rendering others less visible. Similar dynamics can emerge within AI systems when ecological complexity is reorganized according to digital logics that prioritize measurability, standardization, prediction, and computational tractability (Westerlaken, 2024).

These limitations are not merely technical. They influence ecological perception. Species-identification applications may successfully identify organisms while providing little understanding of ecological interactions. Predictive models may reveal patterns without communicating historical, cultural, or environmental contexts. Generative AI systems may summarize ecological information while obscuring uncertainty, provenance, and competing interpretations.

Such outcomes can contribute to ecological distancing. Individuals may interact with representations of biodiversity while remaining disconnected from the ecological relationships that sustain living systems. As biodiversity becomes increasingly machine-mediated, there is a risk that ecological understanding becomes narrower even as information availability expands.

Environmental philosophers have long warned that technological mediation can alter human relationships with nature. Heidegger and Lovitt (1977) argued that technological systems encourage viewing the world as a resource to be managed and utilized. Plumwood (2002) similarly criticized modes of thought that separate humans from ecological communities. Contemporary AI systems may unintentionally reproduce these tendencies when biodiversity is represented primarily as information to be classified, predicted, and optimized.

5.5 Relational values and ecological meaning

The implications of AI extend beyond questions of information accuracy because biodiversity possesses meanings that cannot be fully captured through technical representations alone. Conservation scholars increasingly emphasize relational values, which arise from relationships among people, species, places, and ecosystems rather than from purely instrumental or intrinsic valuations of nature (Chan et al., 2016).

Relational values highlight dimensions of biodiversity that include belonging, stewardship, cultural identity, responsibility, and care. These values often emerge through lived interactions with ecological systems and may be difficult to represent within conventional data structures (Himes and Muraca, 2018). Yet they remain essential for conservation and ecological engagement.

Environmental humanities scholars have similarly argued that ecological understanding emerges through relationships rather than detached observation. Haraway (2016) emphasizes multispecies entanglements as fundamental components of human existence. Tsing (2021) demonstrates how ecological relationships connect humans and nonhuman organisms through shared histories and environments. Rose et al. (2012) likewise argues that ethical relationships with the living world depend upon recognizing interconnectedness and mutual dependence.

These perspectives suggest that biodiversity information infrastructures should preserve relational meaning alongside information. AI systems built upon infrastructures that retain ecological context, provenance, locality, and historical relationships are more likely to support ecological association than systems that focus exclusively on classification and prediction.

5.6 Provenance, context, and responsible AI

The effects of AI depend heavily upon the infrastructures from which AI systems learn. Artificial intelligence does not create ecological knowledge independently; it inherits assumptions, structures, and limitations from the data and knowledge systems that support it.

Consequently, context becomes critical. Biodiversity observations accompanied by locality information, habitat descriptions, collection histories, environmental context, and provenance provide richer foundations for ecological interpretation than isolated records lacking contextual information (Carroll et al., 2020).

Provenance is particularly important because it preserves relationships among organisms, places, people, institutions, and histories. Provenance allows ecological knowledge to remain connected to the conditions under which it was produced and interpreted.

Recent discussions surrounding FAIR and CARE principles similarly emphasize accountability, transparency, stewardship, and contextual integrity within data systems (Carroll et al., 2020; Wilkinson et al., 2016). These principles become increasingly important as AI systems gain greater influence over how biodiversity knowledge is generated, interpreted, and communicated.

Responsible AI frameworks also stress transparency, explainability, accountability, and human oversight as essential components of trustworthy AI systems (Dignum, 2019). Within biodiversity contexts, these principles help ensure that AI supports ecological understanding rather than simply producing technically efficient outputs.

5.7 Toward ecology-centered artificial intelligence

The challenge facing biodiversity science is therefore not whether artificial intelligence should be adopted. AI is already becoming deeply integrated into biodiversity research, conservation practice, environmental monitoring, and public engagement. The more important question concerns the forms of ecological relationship that AI systems encourage.

An ecology-centered approach to AI would prioritize relationships alongside information. Such systems would communicate ecological context, provenance, environmental histories, uncertainty, and stewardship relevance in addition to classifications and predictions. Rather than treating biodiversity solely as computational input, ecology-centered AI would help users understand organisms as participants in broader ecological systems.

This perspective aligns closely with emerging work on human-nature connection and ecological reconnection. Studies consistently demonstrate that meaningful relationships with nature are associated with stronger stewardship behaviors, conservation engagement, and environmental responsibility (Ives et al., 2017; Richardson et al., 2020). AI systems that support understanding of ecological relationships may therefore contribute to ecological literacy and conservation outcomes.

The future significance of AI in biodiversity science may therefore depend less on the sophistication of algorithms and more on whether biodiversity information infrastructures preserve the relationships that make ecological knowledge meaningful.

6 Biodiversity data as ecological association

6.1 From information systems to relationship systems

If ecological abstraction describes the detachment of biodiversity information from the relationships that give it meaning, ecological association represents the opposite trajectory. Ecological association occurs when biodiversity information infrastructures preserve, communicate, and strengthen relationships among organisms, places, ecosystems, people, and histories. In such systems, biodiversity data function not merely as informational resources but as pathways through which ecological understanding and environmental stewardship can emerge.

Ecological association emphasizes preserving context, connection, and relational meaning so that biodiversity information systems communicate relationships rather than merely store observations.

6.2 Participation as a form of ecological connection

One of the most powerful mechanisms for ecological association is participation. People often develop stronger relationships with ecological systems when they actively engage in observing, documenting, interpreting, or caring for biodiversity (Ballard et al., 2017; Bonney et al., 2014; Danielsen et al., 2010).

Community science initiatives illustrate this principle clearly. Projects such as iNaturalist, breeding bird surveys, phenology networks, and local biodiversity monitoring programs encourage individuals to participate directly in biodiversity observation and documentation. These activities do more than generate scientific data. They create opportunities for learning, curiosity, place-based awareness, and ecological engagement (Chandler et al., 2017; Cooper et al., 2007).

Individuals who contribute observations often become more aware of species diversity, seasonal change, habitat conditions, and ecological processes occurring within their local environments (Jordan et al., 2011). In this way, such information infrastructures can function as platforms for ecological engagement rather than simply repositories of information.

Importantly, participation can occur at many levels. Individuals may contribute observations, annotate records, transcribe historical documents, validate identifications, share local knowledge, assist with digitization projects, or engage with biodiversity narratives derived from collections and databases. Each of these activities creates opportunities for ecological association by strengthening relationships between people and biodiversity information. Rather than functioning solely as contributors of observations, participants become active contributors to the relational context that biodiversity information infrastructures preserve and communicate.

6.3 Biodiversity education and ecological literacy

Biodiversity information infrastructures also play increasingly important roles in environmental education and ecological literacy. Ecological literacy refers broadly to the capacity to understand ecological relationships, environmental processes, and the interconnected nature of living systems (Orr, 1992).

Traditional approaches to ecological education often emphasize direct experiences in nature. Such experiences remain critically important. However, biodiversity information systems now provide additional pathways through which ecological knowledge can be acquired and explored. Digital collections, biodiversity portals, species distribution systems, ecological visualizations, and AI-assisted educational tools can help individuals investigate ecological systems across scales that would otherwise be inaccessible.

The educational value of these systems depends heavily on how information is represented. Systems that focus exclusively on taxonomic names, occurrence records, or isolated observations may contribute little to ecological understanding. In contrast, infrastructures that reveal relationships among organisms, habitats, environmental change, and human activities can help users develop richer ecological perspectives (Krasny and Tidball, 2015; Monroe et al., 2019).

Digital biodiversity information infrastructures cannot replace lived ecological experiences, but they can help support ecological awareness and understanding when designed to emphasize relationships rather than isolated facts.

6.4 Narratives, context, and ecological meaning

Humans often understand the world through narratives rather than through isolated pieces of information. Ecological association therefore depends not only on preserving data but also on preserving the stories, histories, and contexts that give biodiversity information meaning (Cronon, 1992; Rose et al., 2012).

Narratives help connect biodiversity information to human experience. They transform records into stories of place, environmental change, ecological resilience, and stewardship. Such narratives are particularly important because they help bridge the gap between abstract information and lived ecological understanding.

Digital biodiversity systems increasingly possess the capacity to support narrative forms of engagement through interactive visualizations, storytelling platforms, educational resources, and AI-assisted interpretation tools. When contextual information is preserved, biodiversity systems can communicate not only what exists but also why it matters.

6.5 Biodiversity information infrastructures as pathways to ecological reconnection

The concept of ecological association ultimately reframes the role of biodiversity information infrastructures. Rather than viewing these systems solely as tools for scientific data management, ecological association recognizes them as infrastructures that influence relationships between people and ecological systems.

This perspective does not diminish the importance of scientific rigor, interoperability, or computational capability. Instead, it expands the objectives of biodiversity information infrastructures. Alongside supporting research and conservation, these systems can also contribute to ecological literacy, environmental stewardship, community participation, and place-based understanding.

Such contributions are increasingly important in a world characterized by biodiversity loss, environmental change, and declining opportunities for direct ecological experience. These information infrastructures alone cannot solve these challenges. However, they can help shape how biodiversity is represented, understood, and valued.

7 Toward human-centered and ecology-centered biodiversity systems

7.1 Beyond information infrastructure

The preceding sections have argued that biodiversity information infrastructures are not merely technical systems for managing information. They are also systems of representation that influence how people understand, value, and relate to the living world. As biodiversity science becomes increasingly digitized and mediated through artificial intelligence, databases, and global information networks, the design of these infrastructures carries consequences that extend beyond scientific efficiency. Infrastructure design increasingly influences ecological perception, environmental values, and opportunities for ecological engagement (Bowker and Star, 2000; Leonelli, 2016).

Historically, biodiversity informatics has focused on goals such as data mobilization, interoperability, accessibility, standardization, and analytical capacity (Edwards et al., 2000). These objectives remain essential and have contributed substantially to biodiversity research and conservation. However, the arguments developed throughout this paper suggest that future biodiversity information systems should also be evaluated according to their ability to preserve ecological meaning, support stewardship, facilitate ecological learning, and strengthen human relationships with ecological systems (Ives et al., 2017).

This shift does not require abandoning technological innovation. Rather, it requires recognizing that biodiversity information infrastructures simultaneously serve scientific, educational, cultural, ethical, and ecological functions. Information systems that influence how people encounter biodiversity inevitably influence how people understand their place within ecological systems and their responsibilities toward them (Haraway, 2016).

The challenge therefore becomes one of design. What principles should guide the development of these information infrastructures that move along the Ecological Abstraction–Ecological Association Continuum toward ecological association? Figure 3 summarizes this continuum and presents a conceptual framework for designing biodiversity information infrastructures that preserve ecological relationships while supporting scientific discovery.

7.2 Relational values and ecological meaning

One promising foundation comes from the growing literature on relational values. Traditional environmental debates often distinguish between instrumental values, which emphasize the usefulness of nature to humans, and intrinsic values, which emphasize the inherent worth of nonhuman organisms and ecosystems. Relational values introduce a third perspective by focusing on relationships among people, species, places, and ecological systems (Chan et al., 2016).

Relational values arise through belonging, stewardship, responsibility, identity, reciprocity, care, and attachment to place. They are not properties of organisms alone, nor are they reducible to economic or utilitarian benefits. Rather, they emerge through ongoing interactions between humans and the living world and through the relationships that sustain individual and collective well-being (Himes and Muraca, 2018).

This perspective has important implications for biodiversity information infrastructures. Information systems frequently represent organisms as records, observations, images, or data points. While such representations support scientific analysis, they may fail to communicate the relationships through which biodiversity becomes meaningful. Relational values suggest that biodiversity systems should preserve and communicate ecological relationships rather than simply store ecological information.

Recent biodiversity assessments have similarly emphasized the importance of nature’s contributions to people as multidimensional relationships that include cultural, social, ethical, and place-based dimensions alongside material benefits (Díaz et al., 2019). Biodiversity information systems capable of representing these dimensions may contribute to richer forms of ecological understanding than systems focused solely on biological attributes.

Environmental humanities scholars have likewise argued that ecological meaning emerges through relationships rather than detached observation. Rose et al. (2012) emphasizes ethical responsibilities that arise through human relationships with other species, while Tsing (2021) demonstrates how ecological communities emerge through networks of interdependence among organisms, landscapes, and histories. Together, these perspectives suggest that such information infrastructures should preserve relationships as carefully as they preserve information.

7.3 Stewardship and ecological responsibility

Stewardship provides a second important foundation for ecology-centered biodiversity information infrastructures. Stewardship refers broadly to the responsibility of caring for ecological systems in ways that sustain biodiversity and ecosystem functioning across generations (Berkes, 2012).

Conservation outcomes are strongly influenced by how individuals and communities perceive their relationships with ecological systems. People are more likely to engage in conservation and environmental action when they view themselves as participants within ecological communities rather than external observers or managers (Leopold, 1968). Stewardship depends not only on ecological knowledge but also on ecological identity, belonging, and responsibility (Berkes, 2012; Ives et al., 2017).

Research in social-ecological systems has repeatedly demonstrated that long-term sustainability depends upon relationships among ecological processes, institutions, cultural practices, governance systems, and community engagement (Folke, 2006). Effective stewardship emerges when ecological understanding remains connected to local knowledge, environmental experience, and collective responsibility (Berkes, 2012).

An ecology-centered infrastructure therefore treats stewardship not as a secondary outcome but as a design objective. Biodiversity systems should help cultivate responsibility, participation, and engagement alongside scientific knowledge.

7.4 Ecological learning and relational infrastructure

Nature connectedness and ecological literacy research suggest that meaningful relationships with nature are strengthened through observation, participation, and engagement rather than passive access to information (Nisbet et al., 2009; Richardson et al., 2020). Similarly, ecological literacy depends not only on understanding ecological processes but also on recognizing human embeddedness within them (Orr, 1992). Biodiversity information infrastructures increasingly function as educational environments through collections, community science platforms, and digital knowledge systems, influencing how ecological understanding develops (Bonney et al., 2016).

From a social-ecological systems perspective, these information infrastructures can be viewed as relational infrastructures that connect organisms, places, communities, institutions, and knowledge systems (Cash et al., 2003; Folke, 2006). Their value extends beyond data management when they preserve ecological context, historical continuity, and connections among people and the living world (Carroll et al., 2020; Cumming, 2014). Designing infrastructures that support learning, participation, and relational understanding may therefore strengthen ecological association and stewardship alongside scientific discovery.

7.5 Principles for ecology-centered biodiversity infrastructure

The Ecological Abstraction–Ecological Association Continuum suggests several design principles for future biodiversity information infrastructures.

First, biodiversity systems should preserve context alongside information. Ecological observations become more meaningful when connected to locality, habitat, environmental conditions, historical circumstances, and provenance (Carroll et al., 2020; Leonelli, 2016).

Second, infrastructures should preserve relationships alongside records. Organisms should remain connected to ecological communities, environmental processes, and systems of stewardship rather than existing solely as isolated observations (Chan et al., 2016; Himes and Muraca, 2018).

Third, biodiversity systems should support participation alongside access. Opportunities for observation, annotation, interpretation, and community engagement strengthen ecological association and stewardship (Ballard et al., 2017; Bonney et al., 2014).

Fourth, biodiversity information systems should support multiple forms of knowledge. Scientific information remains essential, but local knowledge, Indigenous knowledge, cultural knowledge, historical knowledge, and community knowledge also contribute to ecological understanding (Berkes, 2012).

Fifth, AI-enabled systems should be evaluated not only according to analytical performance but also according to their ability to communicate ecological relationships, uncertainty, context, and meaning (Dignum, 2019).

The design principles synthesized from the framework developed throughout this paper are summarized in Figure 4, which illustrates key foundations, infrastructure components, and practices for ecology-centered biodiversity information infrastructures. Collectively, these principles move these infrastructures along the Ecological Abstraction–Ecological Association Continuum toward ecology-centered and human-centered models that support ecological association rather than ecological abstraction. These principles align with ecocentric learning, relational values, and ecocultural restoration by treating biodiversity knowledge not simply as information, but as a pathway for reassociating people with the ecological communities and places they inhabit.

7.6 Ethical implications and future directions

The future of biodiversity informatics is not solely a technical question. It is also an ethical and ecological one. Decisions regarding what information is preserved, how biodiversity is represented, and which relationships remain visible influence how people understand and interact with the living world (Bowker and Star, 2000; Leonelli, 2016). Because these representational choices influence how people perceive and relate to ecological communities, they are not only technical decisions but also ethical ones, consistent with the broader tradition of environmental ethics established by Leopold (1968) and further developed by Callicott (1989).

Environmental ethics has long emphasized that sustainability depends upon recognizing relationships among humans, other species, and ecological systems (Plumwood, 2002). Biodiversity information infrastructures increasingly participate in shaping these relationships by mediating how ecological knowledge is created, organized, and communicated (Leonelli, 2016).

As biodiversity science enters an era characterized by global digitization and artificial intelligence, the challenge is not simply to make biodiversity information more accessible, interoperable, or machine-readable. The challenge is to ensure that ecological relationships remain visible within the systems through which biodiversity becomes known.

Human-centered and ecology-centered biodiversity information systems offer one pathway toward that goal. By preserving context, provenance, stewardship, participation, and relational meaning alongside scientific information, biodiversity information systems can contribute not only to scientific discovery but also to ecological understanding, environmental responsibility, and ecological reconnection (Ives et al., 2017).

8 Discussion

Human societies are experiencing unprecedented access to biodiversity information. Billions of biodiversity records, images, observations, environmental measurements, and scientific publications are now accessible through digital infrastructures. At the same time, biodiversity loss continues to accelerate, ecological systems are undergoing rapid transformation, and opportunities for direct engagement with nature are declining for many people (Díaz et al., 2019; Soga and Gaston, 2016). This apparent paradox raises an important question: why does increasing information about nature not necessarily translate into stronger ecological relationships?

The central argument of this paper is that biodiversity information infrastructures influence not only ecological understanding but also whether individuals recognize themselves as participants within ecological systems. We argue that ecological re-identification is the principal mechanism through which biodiversity information infrastructures shape long-term human–nature relationships. The Ecological Abstraction–Ecological Association Continuum explains how representational choices either obscure or reveal ecological relationships, thereby inhibiting or supporting ecological re-identification.

Viewed through this perspective, the Ecological Abstraction–Ecological Association Continuum does more than describe alternative approaches to biodiversity information infrastructure. It provides a conceptual explanation for why infrastructures that preserve ecological context, provenance, place-based relationships, participation, and relational meaning are more likely to support ecological re-identification than infrastructures emphasizing ecological abstraction. The framework therefore helps explain how representational choices within biodiversity information infrastructures influence whether people encounter biodiversity as disconnected information or as evidence of ecological systems in which they themselves participate. More specifically, the continuum explains how biodiversity information infrastructures support ecological re-identification through four complementary features: preserving ecological relationships rather than isolated data objects, restoring place-based ecological context, supporting participatory engagement with biodiversity, and using AI to preserve and reveal ecological relationships rather than obscure them.

To explore this challenge, the paper developed the Ecological Abstraction–Ecological Association Continuum, a conceptual framework describing how such information infrastructures shape ecological understanding by positioning biodiversity representations along a spectrum from ecological abstraction to ecological association. Ecological abstraction occurs when organisms and ecosystems become detached from the ecological, historical, cultural, and place-based relationships that give them meaning. Within increasingly digital and AI-enabled environments, biodiversity may be encountered primarily as records, labels, datasets, images, or computational resources. These representations provide enormous scientific value, but they can also reduce visibility of the relationships that connect organisms to ecosystems, places, histories, and human communities.

At the opposite end of the continuum, ecological association occurs when biodiversity information infrastructures preserve and communicate relationships alongside information. Rather than representing organisms as isolated data objects, ecology-centered infrastructures maintain connections among biodiversity, locality, provenance, environmental history, ecological interactions, stewardship, and human experience. In this framework, biodiversity information becomes a pathway to ecological interpretation rather than an endpoint in itself. Ecological association therefore represents the representational conditions necessary for ecological re-identification. Within the Ecological Abstraction–Ecological Association Continuum, ecological association describes the relational characteristics of biodiversity information infrastructures, ecological understanding describes recognition of relationships among organisms and ecological systems, and ecological re-identification describes the subsequent shift in which individuals recognize themselves as participants within those same systems. Stewardship emerges as a downstream consequence of this change in self-perception. The Ecological Abstraction–Ecological Association Continuum extends beyond describing biodiversity information infrastructures by explaining how those representational conditions influence whether people come to recognize themselves as participants within ecological systems.

Biodiversity information infrastructures should be evaluated not only according to technical performance but also according to their capacity to preserve ecological meaning, support stewardship, and sustain relationships among organisms, places, and people (Ives et al., 2017). The central challenge is ensuring that biodiversity knowledge remains connected to the ecological realities it represents. Infrastructures that preserve relationships alongside information can help support ecological understanding, stewardship, and conservation.

We hypothesize that biodiversity information infrastructures can be positioned along the Ecological Abstraction–Ecological Association Continuum, and that infrastructures preserving ecological context, provenance, place-based relationships, and participatory engagement will foster stronger ecological understanding and human–nature relationships than infrastructures that emphasize information abstraction alone. This hypothesis is testable within current biodiversity informatics practice. Future studies can operationalize dimensions of the Ecological Abstraction–Ecological Association Continuum, such as preservation of provenance, ecological context, place-based relationships, participatory engagement, and relational richness, and compare biodiversity information infrastructures to evaluate whether systems positioned closer to ecological association foster stronger ecological understanding and human–nature relationships. Possible outcome measures include ecological literacy, nature connectedness, stewardship attitudes, conservation behaviors, user interpretation of biodiversity information, and the preservation of ecological context and provenance within biodiversity information systems.

By designing biodiversity information infrastructures that preserve context, provenance, place, participation, and ecological meaning, biodiversity informatics can contribute not only to scientific discovery but also to strengthening conservation, ecological literacy, stewardship, and human–nature relationships. The Ecological Abstraction–Ecological Association Continuum provides one conceptual framework for guiding that transition.

8.1 Illustrative dimensions for operationalizing the continuum

The hypothesis advanced in this paper - that biodiversity information infrastructures positioned closer to ecological association are more likely to support ecological understanding and stronger human–nature relationships - can be evaluated at two related levels. Infrastructure-level assessment can examine whether systems preserve provenance, ecological context, place-based relationships, ecological interactions, participation, and human–ecological linkages. User-level assessment can then test whether exposure to these characteristics is associated with ecological literacy, systems thinking, nature connectedness, ecological re-identification, stewardship attitudes, and conservation behavior.

The illustrative dimensions summarized in Table 1 should therefore be understood as characteristics that may enable ecological association rather than as direct measures of human–nature relationships. For example, rich provenance may situate records within ecological encounters and histories; ecological and place-based context may help users interpret organisms as components of living systems; representations of species interactions may make ecological interdependence visible; and participatory features may position users as contributors rather than passive observers. These proposed mechanisms remain hypotheses that require empirical testing.

Table 1. Illustrative dimensions for operationalizing the ecological abstraction–ecological association continuum.

Comparative, experimental, or longitudinal studies could examine whether infrastructures or interfaces with stronger ecological-association characteristics produce different user outcomes than systems emphasizing decontextualized information. Such studies could also test whether ecological understanding mediates the relationship between infrastructure design and ecological re-identification.

Ecological re-identification is therefore not equivalent to acquiring more biodiversity knowledge. It represents a shift in self-perception through which individuals recognize themselves as participants within ecological systems. Biodiversity information infrastructures may support this shift by preserving relational context, making ecological interdependence visible, connecting information to place and history, and enabling participation in knowledge production. These characteristics do not guarantee stronger human–nature relationships, but they may create conditions that support ecological understanding, stewardship, and more enduring forms of ecological association.

9 Conclusions

The primary theoretical contribution of this paper is the proposition that the Ecological Abstraction–Ecological Association Continuum explains how biodiversity information infrastructures support ecological re-identification by shaping how people perceive their own place within ecological systems. Rather than viewing stewardship as the principal outcome of biodiversity information systems, the framework proposes that stewardship emerges from a prior shift in self-perception in which individuals increasingly recognize themselves as participants embedded within living ecological systems. The Ecological Abstraction–Ecological Association Continuum therefore explains the representational mechanisms through which biodiversity information infrastructures either reinforce ecological abstraction or support ecological re-identification.

As biodiversity information becomes increasingly digitized and AI-mediated, designing infrastructures that preserve ecological context, provenance, place-based relationships, participation, and relational meaning will become increasingly important. By integrating these principles alongside computational capabilities, biodiversity information infrastructures can contribute not only to scientific discovery but also to supporting ecological re-identification, strengthening the recognition that humans remain integral participants within the living systems upon which they depend, and ultimately fostering more enduring commitments to biodiversity conservation.

Ultimately, the Ecological Abstraction–Ecological Association Continuum reframes biodiversity information infrastructures not simply as systems for managing biodiversity knowledge, but as systems whose representational characteristics influence ecological re-identification—the process through which people come to recognize themselves as participants within the living systems they study, document, and depend upon.

Generative AI statement

The author(s) declared that generative AI was used in the creation of this manuscript. The author used generative AI during the preparation of this manuscript to assist with language editing and refinement of the text. All scientific content, conceptual development, interpretations, and final editorial decisions were performed and verified by the author, who assumes full responsibility for the content of the manuscript.

Any alternative text (alt text) provided alongside figures in this article has been generated by Frontiers with the support of artificial intelligence and reasonable efforts have been made to ensure accuracy, including review by the authors wherever possible. If you identify any issues, please contact us.

Publisher’s note

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.

Footnotes

Edited by: Thomas H. Beery, Kristianstad University, Sweden

Reviewed by: Consuelo Sendino, Natural History Museum (United Kingdom), United Kingdom

Felipe Polivanov Ottoni, Federal University of Maranhão, Brazil