Close-range remote sensing (CRRS) technologies are increasingly used in forestry, but there is a lack of awareness of the challenges, needs and expectations of both service providers and end users. We used a customised online questionnaire to interview professionals in the field, recruited through direct (existing networks) and indirect channels (social media). The main barriers we identified include the cost of equipment, the complexity of data processing workflows and insufficient access to specialised training. Our findings emphasise the need for interdisciplinary collaboration, the development of more intuitive and user-friendly tools and the expansion of specialised training programmes. The results of the questionnaire suggest that stronger partnerships between industry and academia should be encouraged to drive innovation and knowledge sharing. In addition, the development of standardised protocols for CRRS applications and the creation of accessible educational resources proved essential to support both novice and experienced users. Scientific conferences are the most important platform to gather all stakeholders in one place, and have underutilised potential to narrow the gap between theory and application. The recommendations we have made aim to facilitate the widespread adoption and efficient utilisation of CRRS technologies in practical forestry.

A key challenge in applying AIto forest monitoring is tree species identification, hindered by complex visual variations and limited data of tree canopies. Moreover, AI-assisted recognition frameworks that effectively integrate large language models with human knowledge bases in forestry have not yet been fully established. Against this background, this study introduces linguistic guided image diffusion model (LGINet), a framework integrating textually guided image generation with tree species detection to enhance forest applications through technological synergy. The framework comprises three key methodological components: (1) A module designed to leverage detailed textual descriptions specific to forest environments, incorporating elements such as species traits, spatial layout, and phenological features. It generates text embeddings that are semantically aligned with forest knowledge. (2) An innovative diffusion-based framework integrates an improved U-Net architecture with Markov chain theory and textual semantic embeddings, fusing noisy images with forest-specific linguistic semantics to generate highly realistic aerial tree images. (3) An optimized detection pipeline utilizing the generated images, based on an enhanced YOLOv11 architecture with a context-aware feature extractor and adaptive anchor scaling, enabling efficient, large-scale tree crown detection and species identification. Experimental evaluations validate the efficacy of the proposed framework, with the image generation module achieving a Structural Similarity Index (SSIM) of 0.94 and a Frechet Inception Distance (FID) score of 6.42, demonstrating exceptional fidelity in synthetic output quality. Furthermore, the detection pipeline attained a mean Average Precision (mAP50) of 0.868 in species identification tasks, consistently outperforming all baseline models across evaluation metrics. The integrated system improves capabilities of forest species detection by leveraging linguistic-driven synthesis to generate high-fidelity forest imagery and tree species identification.

Habitat fragmentation reduces core habitat and disrupts structural connectivity, ecosystem functioning and biodiversity persistence across human-modified landscapes. Fragmentation may be shaped not only by biophysical stressors but also by governance discontinuities created by administrative boundaries. We quantified four decades of forest change (1985–2023), assessed temporal trends in fragmentation and structural connectivity, and tested whether spatial variation in administrative boundary density and climate water deficit anomalies were associated with forest loss, fragmentation, and connectivity across northern Ethiopia at the contested intersection of Tigray, Amhara and Afar provinces. We mapped land cover for 1985, 1994, 2009, and 2023 using Landsat Collection 2 Level 2 surface reflectance and Random Forest classifiers in Google Earth Engine. Forest was extracted as a binary layer for Morphological Spatial Pattern Analysis. We quantified core, edge and connector classes and graph-based indices to assess structural connectivity across different dispersal thresholds. Impacts of boundary density and climatic water deficit anomalies were evaluated using mixed effect models. Total forest cover declined from 4503 km2 in 1985 to 1757 km2 in 2023, a net loss of 61

BACKGROUND:The feeding behaviours of the malaria vector Anopheles arabiensis, and its competitive relationships with other sibling species within the Anopheles gambiae complex, remain largely unexplored within well conserved natural ecosystems, where its known preferred hosts are scarce or absent. METHODS:Potential aquatic habitats were surveyed for An. gambiae complex larvae across a gradient of natural ecosystem integrity in southern Tanzania, encompassing fully domesticated human settlements, a partially encroached Wildlife Management Area (WMA), and well conserved natural ecosystems within Nyerere National Park (NNP). Direct observations, tracks, spoor and other signs of human, livestock or wild animal activity around these water bodies were recorded as indirect indicators of potential blood source availability. FINDINGS:While only An. arabiensis was found in fully domesticated ecosystems, its non-vector sibling species An. quadriannulatus occurred in conserved areas and dominated the most intact natural ecosystems. Proportions of larvae identified as An. arabiensis were positively associated with human and/or cattle activity and negatively associated with distance inside NNP and away from human settlements. Proportions of An. quadriannulatus were positively associated with activities of impala, warthog and possibly bushpig, implicating them as likely preferred blood hosts. While abundant impala and lack of humans or cattle in intact acacia savannah within NNP apparently allowed it to dominate An. arabiensis, presence of warthog seemed to provide it with a foothold in miombo woodlands of the WMA, despite encroachment there by people and livestock. While this antelope and suid are essentially unrelated, both are non-migratory residents of small home ranges with perennial surface water, representing potential hosts for An. quadriannulatus that are widespread across extensive natural ecosystems all year round. Despite dominance of An. quadriannulatus in well-conserved areas, An. arabiensis was even found in absolutely intact natural environments > 40km inside NNP, suggesting it can survive on blood from one or more unidentified wild species. Such self-sustaining refuge populations of An. arabiensis inside conservation areas, supported by wild blood hosts that are fundamentally beyond the reach of insecticidal interventions targeted at humans or livestock, may confound efforts to eliminate this key malaria vector. However, they might also enable insecticide resistance management strategies that could restore the effectiveness of pyrethroids in particular. This new approach to indirectly identifying commonly utilized blood sources may also be applicable to an unprecedented diversity of zoophagic mosquitoes, enabling incrimination of possible bridge vector species capable of mediating pathogen spillover from wildlife reservoirs into livestock and/or human populations.

Abstract Habitat fragmentation disrupts metapopulation dynamics by altering environmental conditions and constraining demographic processes critical for persistence and recruitment. In the dry Afromontane forests of northern Ethiopia, we investigated how natural and anthropogenic drivers affect seedlings, saplings, and mature tree dynamics of Olea europaea subsp. cuspidata across 34 patches. We used dynamic occurrence models to quantify effects of patch area, altitude, browsing, and disturbance. Our results indicate that high disturbance reduces seedling occurrence probability lower disturbance sites has seedling in 30% of survey plots, high disturbance would bring this down to 10% (median = −1.322, 95% CI: −2.703 to −0.283). Disturbance makes seedling less likely to persist, while large patch size help seedling persists (median = −0.93, 9 5 % CrI −1.87 – −0.02). For mature individuals, disturbance was the only significant predictor of occurrence probability, suggesting greater resistance to environmental and spatial variability compared to earlier life stages. These findings emphasize that while mature trees display resilience, the successful regeneration of Olea europaea is constrained by disturbance, but current level of browsing is not a threat. Management strategies for conservation should prioritise reducing disturbance through community engagement and forest stewardship to enhance regeneration potential and ensure long-term population viability.

Rubber trees in coastal habitats are exposed to a high degree of wind stress. An algorithm-hardware synergetic methodology was developed for investigating and predicting rubber tree phenotyping excited by strong winds. The framework includes (1) a custom-designed industrial fan that recreates a variable airflow field at wind speeds of 15, 30 and 45 m/s coupled with a terrestrial laser scanner and bundled motion sensors to acquire point clouds and vibration data; (2) a graphic model that approximates tree canopies based on foliage clumps with phenotypic traits that are derived from point clouds captured while trees are subjected to aerodynamic drag; and (3) the wind characteristic parameters of forest canopies were calculated by a developed forest-specialized k-epsilon turbulence model combining the constructed tree models and grid-scale subdivision of the wind fluid field. (4) A digital twin model that incorporates detailed tree phenotypic traits and considers plant mechanical characteristics was established, depicting the related wind-induced actions of target trees under various wind influences. The results show that tree crowns with spreading forms are prone to yield larger pendulum amplitudes than compact crowns, but trees directly exposed to wind exhibit greater crown volume reductions than trees in sheltered areas. Within tree canopies, a one-fold increase in inlet wind speed intensified crown compression (approximately 17 % decrease in crown volume), generated 2.1-fold pressure gradients and increased turbulence kinetic energy by approximately 60 %. Moreover, the entire scenario of the adaptation of experimental trees to wind perturbations was visually restored using digital twin techniques, serving as an integral behaviour dataset for further data-driven decision-making. In summary, this paper presents a comprehensive methodology that can decipher the phenotypic manifestations of trees' reactions to wind hazards, with potential applications in phenotyping or envirotyping studies designed to evaluate the wind resistance properties of rubber trees.

ABSTRACTLand use change threatens global biodiversity and compromises ecosystem functions, including pollination and food production. Reduced taxonomic α‐diversity is often reported under land use change, yet the impacts could be different at larger spatial scales (i.e., γ‐diversity), either due to reduced β‐diversity amplifying diversity loss or increased β‐diversity dampening diversity loss. Additionally, studies often focus on taxonomic diversity, while other important biodiversity components, including phylogenetic diversity, can exhibit differential responses. Here, we evaluated how agricultural and urban land use alters the taxonomic and phylogenetic α‐, β‐, and γ‐diversity of an important pollinator taxon—bees. Using a multicontinental dataset of 3117 bee assemblages from 157 studies, we found that taxonomic α‐diversity was reduced by 16%–18% in both agricultural and urban habitats relative to natural habitats. Phylogenetic α‐diversity was decreased by 11%–12% in agricultural and urban habitats. Compared with natural habitats, taxonomic and phylogenetic β‐diversity increased by 11% and 6% in urban habitats, respectively, but exhibited no systematic change in agricultural habitats. We detected a 22% decline in taxonomic γ‐diversity and a 17% decline in phylogenetic γ‐diversity in agricultural habitats, but γ‐diversity of urban habitats was not significantly different from natural habitats. These findings highlight the threat of agricultural expansions to large‐scale bee diversity due to systematic γ‐diversity decline. In addition, while both urbanization and agriculture lead to consistent declines in α‐diversity, their impacts on β‐ or γ‐diversity vary, highlighting the need to study the effects of land use change at multiple scales.

Context: Deer (Cervidae) populations are increasing in many global regions, leading to concerns about their impacts on temperate forests. Advancing evidence-based management requires a detailed understanding of the dietary habits of deer and how these are shaped by resource availability. Methodology: We studied the diet of fallow deer (Dama dama) in North Wales (United Kingdom), using faecal DNA metabarcoding. Samples were collected monthly from three woodlands during 2019-2021. Tree surveys and seasonal ground flora surveys were conducted in these woodlands and seven additional woodlands. Preference analyses were used to assess the consumption of plant taxa relative to their availability. Results: The fallow deer consumed high proportions of bramble (Rubus fruticosus agg.) across the seasons, especially in the winter months. Diet diversity was significantly lower in winter compared to the other seasons, suggesting that the deer were bulk foraging on a widely available, predictable resource to conserve energy during winter. Grasses did not form a major component of the diet in any season. The preference analysis showed that spatially clustered woody taxa (e.g. Betula sp., Corylus sp. and Fraxinus sp.) occurred less often than expected in the diet, while widespread woody species occurred in the diet more often than expected (e.g. Rosa sp., Prunus sp. and Quercus sp.). Practical implication: The expansion of deer populations in the United Kingdom has occurred alongside the recovery and maturation of degraded or planted forests since the middle of the 20th century. Despite reduced light availability in these closed-canopy forests and increased herbivory pressure, bramble has remained a dominant understory plant compared to other less herbivory-tolerant plant species. Perhaps as a consequence, bramble has become the winter survival resource for this fallow deer population, remaining a prominent dietary component throughout the year. With increasing disturbance from extreme weather and tree diseases leading to a more open canopy structure, bramble cover is set to increase in European forests, which could support further expansion of deer populations. As we work to expand tree cover and enhance forest resilience and biodiversity, we should seek to understand the dynamic interactions of increasing deer populations with rapidly changing treescapes.

Context : Rapid expansion of deer (Cervidae) populations is a concern for forest ecosystems. Despite extensive reviews on how deer affect forests, variation in effects across deer species has received less attention. A lack of focus on species‐specific effects may lead to oversights and failure to achieve desired management outcomes. Methodology : We used a systematic approach to compile data on the extent to which the effects of seven deer species on woody vegetation have been studied. We focused on the six deer species present in Britain and Ireland, and elk ( Cervus canadensis ). Results : A total of 455 studies were included from across the globe. Red deer ( Cervus elaphus ) ( n = 163) and elk ( n = 158) were the most studied species, while Reeve's muntjac ( Muntiacus reevesi ) ( n = 18) and Chinese water deer ( Hydropotes inermis ) ( n = 5) were the least researched. Fifty‐four per cent of studies ( n = 245) used fenced exclosures to assess deer impacts. Research mainly focused on defoliation via browsing and grazing ( n = 424), while debarking ( n = 44), defecation ( n = 8) and trampling ( n = 5) were less frequently studied. Vegetation density ( n = 235), height ( n = 189) and diversity ( n = 135) were the most common metrics used, while fewer studies focused on vegetation mortality ( n = 74), structural variability ( n = 28) and condition ( n = 15). Practical implication : While previous studies have often focused on the probability or severity of deer damage to woody vegetation, we identified key knowledge gaps on the ecological influence of such damage, with a species‐specific focus. Researchers should treat deer species as distinct entities and appreciate the differences in their body size, sociality, physiology and behaviour when studying their ecological effects. Where multiple deer species co‐occur, identifying relative local species abundance and differences among species foraging behaviours will help to determine how their interactions—whether additive, synergistic or antagonistic—affect ecosystem processes and vegetation dynamics.

Light Detection and Ranging (LiDAR) has emerged as an important data source for monitoring forest resources. Terrestrial laser scanning (TLS) and Mobile laser scanning (MLS) have already shown high potential in further advancing forest inventory development. By enabling the retrieval of new forest attributes in addition to traditional ones, these technologies could drive forest inventories into a new paradigm by introducing innovative approaches to measuring and monitoring forests. The debate on the possible implementation of TLS and MLS in forest inventories, particularly in national forest inventories (NFIs), continues in both the scientific community and the public institutions. To date, few studies have evaluated the application of TLS and MLS technologies in large-scale forest inventories or assessed their practical operational limits. In this practice-oriented paper, we first detail TLS and MLS data acquisition and processing for tree attribute estimation, assessing their maturity and main limitations. We then explore three European case studies-from the French, Finnish, and Swiss National Forest Inventories (NFIs)-where these technologies have been tested. Based on these experiences, we identify the main constraints and challenges for operational implementation. Lastly, we discuss the prospects for TLS and MLS within the NFI context and the requirements for their successful adoption. We conclude that TLS and MLS should be viewed not as a replacement for, but as a complement to and enhancement of, traditional NFI practices. Emphasis should be placed on the new opportunities these technologies offer, rather than on direct comparisons with conventional methods.

Deep learning, which has exhibited considerable potential and effectiveness in forest resource assessment, is vital for comprehending and managing forest resources and ecosystems. However, extensive assessment of forest resources is highly challenging due to the complex and varied nature of forest types sourced from diverse remote sensing platforms, which include images, point clouds, and fusion data. To facilitate further study, we systematically review the current status, applications and prospects of deep learning technologies for different types of forest remote sensing data. After considering more than two hundred forest-related papers published over the past decade, we introduce sensors and devices for forest data acquisition, classify deep learning methods based on their data processing methods and operational principles, and categorize diverse instances of these methods with various forest applications. Moreover, we summarize available datasets related primarily to forest data and examine the global geographic distribution of the relevant literature. Comprehensive insights into the advantages and limitations of each method are described, offering a forward-looking perspective on the trend of applying deep learning technology to forest research. In this paper, we aim to provide an overview of the current trends and challenges of deep learning techniques applied to forest research, creating a comprehensive picture for use as a reference by both academia and industry professionals.

Download This Paper Open PDF in Browser Add Paper to My Library Share: Permalink Using these links will ensure access to this page indefinitely Copy URL Copy DOI

Determining the spatial patterning of tree species can provide inferences on underlying ecological processes. Gonystylus brunnescens is a South-east Asian subcanopy forest tree. To determine the spatial patterns of this species, we recorded the distribution of all individuals in a 0.4 ha sampling plot in eastern Borneo. We found that the pattern deviated from random and was well-described by the Matérn cluster model; clusters had a radius of approximately 4.2 m and contained an average of six seedlings each. This supports the hypothesis of animal-dispersed seeds and, due to a clear lack of association of juveniles with adults, may be due to scatter-hoarding of seeds by small mammal seed dispersers.

The gap fraction (GF) of vegetative canopies is an important property related to the contained bulk of reproductive elements and woody facets within the tree crown volume. This work was developed from the perspectives of porous media theory and computer graphics techniques, considering the vegetative elements in the canopy as a solid matrix and treating the gaps between them as pores to guide volume-based GFvol calculations. Woody components and individual leaves were extracted from terrestrial laser scanning data. The concept of equivalent leaf thickness describing the degrees of leaf curling and drooping was proposed to construct hexagonal prisms properly enclosing the scanned points of each leaf, and cylinder models were adopted to fit each branch segment, enabling the calculation of the equivalent leaf and branch volumes within the crown. Finally, the volume-based GFvol of the tree crown following the definition of the void fraction in porous media theory was calculated as one minus the ratio of the total plant leaf and branch volume to the canopy volume. This approach was tested on five tree species and a forest plot with variable canopy architecture, yielding an estimated maximum volume-based GFvol of 0.985 for a small crepe myrtle and a minimal volume-based GFvol of 0.953 for a sakura tree. The 3D morphology of each compositional element in the tree canopy was geometrically defined and the canopy was considered a porous structure to conduct GFvol calculations based on multidisciplinary theory.

There is great potential for the use of terrestrial laser scanning (TLS) to quantify aspects of habitat structure in the study of animal ecology and behaviour. Viewsheds-the area visible from a given position-influence an animal's perception of risk and ability to respond to potential danger. The management and conservation of large herbivores and their habitats can benefit greatly from understanding how vegetation structure shapes viewsheds and influences animal activity patterns and foraging behaviour. This study aimed to identify how woodland understory structure influenced horizontal viewsheds at deer eye height. Mobile TLS was used in August 2020 to quantify horizontal visibility-in the form of Viewshed Coefficients (VC)-and understory leaf area index (LAI) of 71 circular sample plots (15-m radius) across 10 woodland sites in North Wales (UK) where fallow deer (Dama dama) are present. The plots were also surveyed in summer for woody plant size structure, stem density and bramble (Rubus fruticosus agg.). Eight plots were re-scanned twice in winter to compare seasonal VC values and assess scan consistency. Sample plots with higher densities of small stems had significantly reduced VC 1 m from the ground. Other stem size classes, mean percentage bramble cover and understory LAI did not significantly affect VC. There was no difference in VC between summer and winter scans, or between repeated winter scans. The density of small stems influenced viewsheds at deer eye height and may alter behavioural responses to perceived risk. This study demonstrates how TLS technology can be applied to address questions in large herbivore ecology and conservation.

Forested environments feature a highly complex radiation regime, and solar radiation is hindered from penetrating into the forest by the 3D canopy structure; hence, canopy shortwave radiation varies spatiotemporally, seasonally, and meteorologically, making the radiant flux challenging to both measure and model. Here, we developed a synergetic method using airborne LiDAR data and computer graphics to model the forest canopy and calculate the radiant fluxes of three forest plots (conifer, broadleaf, and mixed). Directional incident solar beams were emitted according to the solar altitude and azimuth angles, and the forest canopy surface was decomposed into triangular elements. A ray tracing algorithm was utilized to simulate the propagation of reflected and transmitted beams within the forest canopy. Our method accurately modeled the solar radiant fluxes and demonstrated good agreement ( R 2 ≥ 0.82 ) with the plot-scale results of hemispherical photo-based HPEval software and pyranometer measurements. The maximum incident radiant flux appeared in the conifer plot at noon on June 15 due to the largest solar altitude angle (81.21°) and dense clustering of tree crowns; the conifer plot also received the maximum reflected radiant flux (10.91-324.65 kW) due to the higher reflectance of coniferous trees and the better absorption of reflected solar beams. However, the broadleaf plot received more transmitted radiant flux (37.7-226.71 kW) for the trees in the shaded area due to the larger transmittance of broadleaf species. Our method can directly simulate the detailed plot-scale distribution of canopy radiation and is valuable for researching light-dependent biophysiological processes.

Accurate segmentation of individual tree crowns (ITCs) from airborne light detection and ranging (LiDAR) data remains a challenge for forest inventories. Although many ITC segmentation methods have been developed to derive tree crown information from airborne LiDAR data, these algorithms contain uncertainty in processing false treetops because of foliage clumps and lateral branches, overlapping canopies without clear valley-shape areas, and sub-canopy crowns with neighbouring trees that obscure their shapes from an aerial perspective. Here, we propose an approach to crown segmentation using computer vision theories applied in different forest types. First, a dual Gaussian filter was designed with automated adaptive parameter assignment and a screening strategy for false treetops. This preserved the geometric characteristics of sub-canopy trees while eliminating false treetops. Second, anisotropic water expansion controlled by the energy function was applied for accurate crown segmentation. This utilized gradient information from the digital surface model and explored the morphological structures of tree crown boundaries as analogous to the maximal valley height difference from surrounding treetops. We demonstrate the generality of our approach in the subtropical forests within China. Our approach enhanced the detection rate of treetops and ITC segmentation relative to the marker-controlled watershed method, especially in complicated intersections of multiple crowns. A high performance was demonstrated for three pure Eucalyptus plots (a treetop detection rate r = 0.95 and crown width estimation R-2 = 0.90 for canopy trees; r = 0.85 and R-2 = 0.88 for sub-canopy trees) and three plots dominated by Chinese fir (r = 0.95 and R-2 = 0.87 for canopy trees; r = 0.79 and R-2 = 0.83 for sub-canopy trees). Finally, in a relatively complex forest park containing a wide range of tree species and sizes, a high performance was also achieved (r = 0.93 and R-2 = 0.85 for canopy trees; r = 0.70 and R-2 = 0.80 for sub-canopy trees). Our method demonstrates that methods inspired by the computer vision theory can improve on existing approaches, providing the potential for accurate crown segmentation even in mixed forests with complex structures

Benthic fauna form spatial patterns which are the result of both biotic and abiotic processes, which can be quantified with a range of landscape ecology descriptors. Fine- to medium-scale spatial patterns (<1–10 m) have seldom been quantified in deep-sea habitats, but can provide fundamental ecological insights into species’ niches and interactions. Cold-water coral reefs formed by Desmophyllum pertusum (syn. Lophelia pertusa) and Madrepora oculata are traditionally mapped and surveyed with multibeam echosounders and video transects, which limit the ability to achieve the resolution and/or coverage to undertake fine-scale, centimetric quantification of spatial patterns. However, photomosaics constructed from imagery collected with remotely operated vehicles (ROVs) are becoming a prevalent research tool and can reveal novel information at the scale of individual coral colonies. A survey using a downward facing camera mounted on a ROV traversed the Piddington Mound (Belgica Mound Province, NE Atlantic) in a lawnmower pattern in order to create 3D reconstructions of the reef with Structure-from-Motion techniques. Three high resolution orthorectified photomosaics and digital elevation models (DEM) >200 m2 were created and all organisms were geotagged in order to illustrate their point pattern. The pair correlation function was used to establish whether organisms demonstrated a clustered pattern (CP) at various scales. We further applied a point pattern modelling approach to identify four potential point patterns: complete spatial randomness (CSR), an inhomogeneous pattern influenced by environmental drivers, random clustered point pattern indicating biologically driven clustering and an inhomogeneous clustered point pattern driven by a combination of environmental drivers and biological effects. Reef framework presence and structural complexity determined inhabitant distribution with most organisms showing a departure from CSR. These CPs are likely caused by an affinity to local environmental drivers, growth patterns and restricted dispersion reproductive strategies within the habitat across a range of fine to medium scales. These data provide novel and detailed insights into fine-scale habitat heterogeneity, showing that non-random distributions are apparent and detectable at these fine scales in deep-sea habitats.

Anthropogenic litter (solid manufactured waste) is an understudied but pervasive element of river systems worldwide. Its physical structure generally differs from natural substrates, such as gravel and cobbles (hereafter rocks ). Consequently, anthropogenic litter could influence ecological communities in urban rivers by providing novel habitats. This study compares the macroinvertebrates recorded on anthropogenic litter with those on rocks to test whether the different substrates support distinct communities. Macroinvertebrates were collected from individual rocks and anthropogenic litter, predominantly plastic, metal, and glass, in three U.K. rivers. Macroinvertebrate communities on anthropogenic litter were consistently more diverse than those found on rocks, reflecting its greater surface complexity, but the density of macroinvertebrates was similar among substrates. The community composition also varied between substrates, with five taxa only recorded on anthropogenic litter. Community differences largely reflected greater abundances of common taxa on anthropogenic litter, which were relatively insensitive to environmental quality. Plastic and fabric anthropogenic litter communities were the most dissimilar to those on rocks, probably due to their flexibility, which could replicate the physical structure of aquatic macrophytes. Our findings indicate that anthropogenic litter supports a distinct and diverse community of macroinvertebrates in urban rivers, which are otherwise relatively homogenous in habitat structure. Removal of anthropogenic litter from urban rivers may not be beneficial for local biodiversity. Understanding the functional habitats provided by anthropogenic litter could help better manage urban rivers to replace habitat lost through urbanisation.