A Simple Habitat Model Is Quietly Reshaping How Scientists Map Human-Wildlife Conflict

As human populations expand and natural habitats shrink, encounters between people and wildlife are becoming more frequent, more intense, and more politically charged. Conservation scientists have long needed tools that can translate these tensions into maps and numbers that decision-makers can actually use. A new systematic review published in the journal BioRisk examines one of the most popular of those tools—the InVEST Habitat Quality model—and reveals both its remarkable rise and its surprising blind spot when it comes to the very human-wildlife interactions it could help resolve.

The InVEST Habitat Quality model is part of a suite of open-source software developed by the Natural Capital Project, a partnership involving Stanford University, the University of Minnesota, The Nature Conservancy, and WWF. Its core purpose is to map and assess ecosystem services, helping decision-makers understand how changes in ecosystems affect both human well-being and biodiversity. The Habitat Quality module estimates the capacity of a landscape to sustain viable wildlife populations using three relatively simple inputs: a land use and land cover map, a table of threats such as urban development or agriculture, and a sensitivity table describing how vulnerable each habitat type is to those threats. Because it relies on readily available data rather than expensive field surveys, it has become attractive to researchers and planners working in data-scarce regions and under tight budgets.

The review’s authors, led by researchers synthesising the rapidly expanding literature, searched the Scopus and Web of Science databases for peer-reviewed papers that genuinely applied the model. Their initial search returned 75 publications, but after screening for documented applications and removing papers in languages the team could not evaluate, 61 studies published between 2016 and 2025 made the final cut. Each paper was systematically coded using a structured template covering six components: general publication characteristics, case study details, the model’s application, its practical and policy context, ecosystem services addressed, and the technical specifics of inputs and outputs. Five reviewers worked through the dataset, and their entries were merged and checked for consistency before analysis.

The numbers tell a striking story of accelerating adoption. Only one paper applied the model in 2016, but annual output climbed to 13 papers by 2025, with roughly 82 percent of all studies published after 2019. The publications appeared across 36 different journals, with Sustainability leading at six papers, followed by Ecological Indicators, Land, and Remote Sensing with five each. Geographically, the evidence base is heavily concentrated: China alone accounts for 49.2 percent of all case studies, with Italy and Iran trailing far behind. Nearly all applications operated at the local or regional scale, with only a handful of studies tackling national or subnational assessments.

Three major methodological trends emerged from the analysis. First, the model is overwhelmingly used to resolve spatial discrepancies between nature conservation and other human activities, such as urban development or watershed management, identifying priority areas for protection, restoration, and development. Second, although the model is often considered best suited for assessing biodiversity in general, a parallel trend of species-specific applications has taken hold—18 of the 61 papers focused on particular species or groups, including tigers, snub-nosed monkeys, elephants, crocodile lizards, and pancake tortoises. Third, and perhaps most telling, the model is rarely used alone. Only 25 percent of studies applied it as a standalone tool; the rest combined it with other models, approaches, or both, averaging 2.5 complementary methods per study.

This integration pattern reflects both the model’s flexibility and its inherent limitations. Landscape ecology models such as FRAGSTATS and the Minimum Cumulative Resistance model were the most common companions, helping capture spatial structure, fragmentation, and connectivity that the Habitat Quality model alone cannot address. Statistical models, particularly the species distribution tool MaxEnt, allow habitat suitability to be grounded in empirical occurrence data rather than expert judgement. GIS-based spatial analysis approaches were the most frequently used category of complementary methods overall. Notably, combinations with other InVEST modules—such as Carbon Storage or Sediment Delivery Ratio—appeared in only a handful of papers, all in general rather than species-specific applications.

Yet the review also uncovered significant methodological inconsistencies that complicate comparisons across studies. The number of threats modelled ranged from two to 16, with six being the most common. Urban areas were considered a threat in 90 percent of studies, but in over half of those they were treated as a single undifferentiated category, while others split them into as many as seven subcategories such as residential versus rural areas or population density. Because threat weights and decay distances are assigned by expert judgement rather than derived from empirical impact data, this inconsistency in threat granularity directly affects how much of the observed variation in results reflects genuine ecological differences versus differences in how threats were operationalised. Similarly, the number of land cover classes used ranged from three to 45, with species-specific studies averaging 15.8 classes compared to 9.8 in general applications, increasing the number of expert-based judgements required and the potential for subjective bias.

The most consequential finding concerns human-wildlife interactions. Only 22 of the 61 reviewed papers had any direct or indirect relation to this theme, and just ten established a direct connection—all of them species-specific studies. Most examined how a particular human activity, such as deforestation or traffic, affects the habitat quality of a given species, or evaluated protected areas and wildlife corridors. In these applications, the model typically generates a spatially explicit baseline of habitat suitability against which a threat or intervention is assessed, rather than modelling the interaction itself—conflict frequency, encounter risk, or bidirectional impacts between wildlife and people. The authors argue this reveals an important gap: even in studies most aligned with human-wildlife themes, the model’s parameters are calibrated to represent generic land-use pressures rather than the specific dynamics of conflict or coexistence.

The review’s conclusions point toward two complementary research directions. First, systematically testing and validating combined modelling frameworks that integrate the Habitat Quality model with species distribution models and landscape ecology approaches would help bridge the gap between general ecosystem service assessments and species-specific ecological processes, reducing reliance on unvalidated expert parameterisation. Second, future studies should move beyond using the model solely as a habitat baseline and explore closer integration with methods explicitly designed to capture human-wildlife dynamics, such as spatial conflict-risk modelling or agent-based approaches. The authors also note that a systematic assessment of how often applications validate their outputs against independent field observations or species records was beyond the scope of the current review but represents an important priority for future syntheses.

For a field racing to keep pace with intensifying human-wildlife conflict, the message is clear: the InVEST Habitat Quality model has earned its popularity as an accessible, spatially explicit workhorse for habitat assessment, but it functions best as a core spatial framework that requires additional modelling layers to capture ecological complexity. Whether it can evolve from a general-purpose mapping tool into a genuine instrument for predicting and managing human-wildlife conflict remains one of the most pressing open questions in applied conservation science.

Subject of Research: A literature review of methodological trends and model integration in applications of the InVEST Habitat Quality model, with implications for human-wildlife interaction research.

Article Title: Applications of the InVEST Habitat Quality model: A literature review of methodological trends and model integration, with implications for human-wildlife interactions

Article References: Nedkov, S., Parvanova, K., Yakovleva, S., Vasileva, I., & Paunova, M. (2026). Applications of the InVEST Habitat Quality model: A literature review of methodological trends and model integration, with implications for human-wildlife interactions. BioRisk, 24, 237-262. https://doi.org/10.3897/biorisk.24.197390

Image Credits: AI Generated

DOI: 10.3897/biorisk.24.197390

Keywords: InVEST Habitat Quality model, ecosystem services, habitat quality, human-wildlife interactions, land use land cover, nature conservation, species distribution models, MaxEnt, FRAGSTATS, spatial modelling, biodiversity conservation, literature review