Publikasjoner
NIBIOs ansatte publiserer flere hundre vitenskapelige artikler og forskningsrapporter hvert år. Her finner du referanser og lenker til publikasjoner og andre forsknings- og formidlingsaktiviteter. Samlingen oppdateres løpende med både nytt og historisk materiale. For mer informasjon om NIBIOs publikasjoner, besøk NIBIOs bibliotek.
2026
Forfattere
Adrian Straker Paul Magdon Marco Zullich Maximilian Freudenberg Christoph Kleinn Johannes Breidenbach Stefano Puliti Nils NölkeSammendrag
Deep learning-based tree species classification models applied to terrestrial laser scanning data achieve high accuracies, yet their classification decisions remain poorly understood. Addressing this gap, we present a novel explanation framework that systematically links salient regions of Finer-CAM saliency maps—which highlight image features contributing to the classification of a target species while suppressing features shared with similar species—to structural tree features in 2D side-view images of terrestrial laser scanning (TLS) point clouds. We applied this framework using YOLOv8-based tree species classification models trained on a subset of the FOR-species20K benchmark data set. Faithfulness analysis confirmed that Finer-CAM reliably identifies the most discriminative image regions for model predictions, validating its suitability for enhancing model interpretability. Analysis of 630 saliency maps indicate that the models primarily rely on image regions associated with tree crowns across most species, particularly for Silver Birch, European Beech, English Oak, and Norway Spruce, while stem regions contribute more strongly to the differentiation of European Ash, Scots Pine, and Douglas-fir. Image perturbation experiments further demonstrated that the visibility of tree structure—such as branching patterns—in the 2D side-view images enhances classification performance, indicating YOLOv8’s capabilities to leverage detailed structural information encoded in high-density TLS point clouds. Using a cross-validation approach our trained models achieved a mean overall accuracy of 96% (SD = 0.24%) on a test data set unseen during training. Beyond assessing the contribution of image regions associated with structural tree features to tree species differentiation, our methodological framework can facilitate a better understanding of model limitations, the detection of data set biases and shortcut learning, and contributes in building confidence in deep learning-based tree species classifiers.
Forfattere
Martyn Futter Katri Rankinen Joachim Audet Heleen de Wit Csilla Farkas Martin Forsius Jens Fölster Anastasija Isidorova Øyvind Kaste Dominika Krzeminska Katarina Kyllmar Emma Lannergård James Edward Sample Eva Skarbøvik Lars Sonesten Brian KronvangSammendrag
River mouth nitrogen: phosphorus: silicon (N:P:Si) stoichiometry can predict marine eutrophication potential. Furthermore, deviations from Redfield molar N: P:Si = 16:1:20 ratios can offer insight into limiting nutrients and risks for harmful algal blooms (HAB). Here, we present N:P:Si stoichiometry based on total P, dissolved inorganic N and dissolved Si for 88 coastal river mouths in Norway, Sweden and Finland between 2017 and 2024. Rivers ranged from ultra-oligotrophic to hyper-eutrophic. N and P concentrations increased from north to south, with no latitudinal Si gradient. Most rivers were either P or jointly NP depleted relative to Si, with no overall evidence of Si depletion. However, there was some evidence of seasonal Si depletion. We show how using multi-element approaches, e.g., the Index of Coastal Eutrophication Potential (ICEP) can help to guide management actions. Specifically, Using Si depletion relative to N and P concentrations as an indicator of HAB potential may offer a means to identify catchments where nutrient load reductions can have the largest benefit on marine ecosystem health. Such multi-element approaches can complement single nutrient assessments based on, e.g., the Carlson Trophic Status Index or Water Framework Directive status class assessments.
Sammendrag
In this paper, we propose a fishery control scheme that happens to be first best in the context of a Gordon‐Schaefer model with asymmetric information about effort/harvesting costs. It is widely acknowledged that in the absence of regulations, competition between vessels leads to a competitive game wherein the outcome is inefficient. We introduce a management scheme that regulates the fishing industry through an individual convex tax on effort or on catches. The shares of this tax can be traded in a market, thus solving problems related to asymmetric information. This can be achieved since heterogeneous fishing firms are individually stimulated to solve the same optimality problem as a social planner would. We apply this model to the Northwest Mediterranean demersal fishery as a case study.
Forfattere
Babak Minofar Simon Weldon Alice Budai Beata Gavurova Lukáš Trakal Daniel Rasse Anna Maroušková Josef MarousekSammendrag
Det er ikke registrert sammendrag
Sammendrag
I forbindelse med reguleringsarbeid knyttet til Ravnsborg gård gnr./bnr. 28/54 i Hvalstad i Asker kommune er det behov for jordbruksfaglig utredning. Jordloven §§ 1 og 9 slår fast at ved omdisponering av god matjord bør jordressursene bevares. Rapporten vurderer jordas egenskaper, hydrologiske forhold, planteskadegjørere samt andre forhold. Det ble foretatt feltarbeid og befaringer 17. og 19. november 2025. Tiltaksarealet har to jorder hvor kun det ene (jorde 1) ble undersøkt i denne rapporten. Matjordlaget er moldholdig siltig mellomleire. Jordsmonnet er leirjord med leiranrikningsjikt. Det ble besluttet at deler av jorde 1 er forstyrret og ikke lenger kan regnes som jordbruksareal. Topplaget fra forstyrrede arealer kan brukes til grøntanlegg internt i prosjektet. Eiendommen er ikke registrert i floghavreregisteret. Jorde 1 er fri for PCN. I reguleringsprosessen ble det besluttet at jordbruksarealene på tiltaksarealet ikke skal nedbygges, og driften opprettholdes.
Forfattere
Eirik Haugstvedt Henriksen Simo Njabulo Maduna Snorre B. Hagen Guttorm ChristensenSammendrag
Pink salmon ( Oncorhynchus gorbuscha ) has rapidly expanded throughout northern Europe, raising concern that continued warming may enable establishment even in high Arctic ecosystems. Svalbard has experienced increasing occurrences of pink salmon in recent years, yet it remains unclear whether these individuals reflect successful local reproduction or straying from mainland populations. We combined field surveys and population genomics to assess the origin and establishment potential of pink salmon on Svalbard. Five rivers around Isfjorden were surveyed in 2024 following observations of unusually high pink salmon abundance and suspected spawning in 2023. Despite extensive sampling of spawning substrates, no eggs or fry were detected. Because most Svalbard rivers freeze to the bed or dry out during winter, current environmental conditions are likely insufficient to support successful egg incubation and emergence . Genomic analyses of 24 adults captured in 2023, integrated with a 3RAD SNP reference dataset (6174 SNPs), revealed that Svalbard individuals retained genetic diversity within the observed range of the mainland populations but formed no distinct genetic cluster. Instead, they exhibited admixture and greatest affinity to sampling populations in the Varanger Fjord region, indicating recurrent long-distance dispersal from established invasion hotspots. Together, the absence of detected recruitment and the lack of a distinct Svalbard genetic cluster indicate that pink salmon in Svalbard currently represent a demographic and genetic sink sustained by recurrent immigration from mainland populations rather than a self-sustaining population. However, the genetic diversity of incoming strays suggests that, if future environmental change weakens current barriers to egg incubation and juvenile emergence, establishment may become increasingly possible.
Sammendrag
Plastic mulch films are widely used in agriculture, often in combination with pesticide application. While biodegradable plastic mulch films (BDM) are designed for soil incorporation, their influence on pesticide fate remains poorly understood. We investigated the influence of BDM on the degradation kinetics of four widely used pesticides (boscalid, pyraclostrobin, fluopyram and pyraflufen ethyl) in two agricultural soils under controlled conditions. Pesticides were applied at agronomically relevant doses either directly to soil or adsorbed onto a commercially available PBAT-starch BDM prior to soil incorporation, and pesticides and their relevant metabolites were monitored for 112 days. Adsorption to BDM significantly slowed pesticide degradation for all compounds, increasing degradation half-life DT50 values by factors of 2-26 compared with direct soil application. While degradation pathways were not fundamentally altered, the presence of mulch delayed metabolite formation and prolonged pesticide persistence. Rapid degradation of pyraflufen ethyl led to the accumulation of pyraflufen-ethyl metabolite E 3, a highly persistent multi halogenated metabolite that remains analytically and regulatorily overlooked. BDM itself showed limited mass loss (12 %) over the experimental period, suggesting that pesticide-plastic interactions can persist over environmentally relevant timescales. Overall, this study identifies plastic mulch residues as overlooked modifiers of pesticide persistence and metabolite accumulation in soil, suggesting that plastic-pesticide interactions warrant further consideration in environmental exposure and risk assessment frameworks.
Forfattere
Frode VeggelandSammendrag
Det er ikke registrert sammendrag
Forfattere
Frode VeggelandSammendrag
Det er ikke registrert sammendrag
Sammendrag
Arealbrukssektoren (LULUCF) omfatter arealbruk og arealbruksendringer, med tilhørende utslipp og opptak av CO₂, CH₄ og N₂O. Framskrivningene bygger på Norges rapportering til FNs klimakonvensjon (NID 2026) og i hovedsak på observerte data for referanseperioden 2011–2024. De representerer en «business as usual»-utvikling basert på vedtatt politikk. Utviklingen levende biomasse, død ved, strø, og mineraljord i skog er framskrevet med SiTree og Yasso20 under klimascenarioet RCP 4.5. Netto opptak har gått ned siden toppårene 2008–2010. I referanseperioden var gjennomsnittlig netto opptak 14,5 mill. tonn CO₂-ekv. per år. Framskrivningene viser videre nedgang til om lag 2,2 mill. tonn per år i 2030–2050, før opptaket øker til 7,0 mill. tonn per år i 2090–2100. Utviklingen styres i stor grad av netto opptak i skog. Dyrka mark, utbygd areal, beite samt vann og myr er netto utslippskilder, og forventes å forbli det. De største utslippene kommer fra drenert organisk jord og avskoging, særlig til utbygd areal. Under EUs LULUCF-regelverk forventes et samlet underskudd mot netto-null-målet på om lag 11,1 mill. tonn CO₂-ekv. for 2021–2025. En evaluering av maskinlæringsmodeller for vekst og mortalitet tyder på at dagens metode kan overvurdere biomasseoppbyggingen i skog. En følsomhetsanalyse viser at 20 % økt hogst kan gjøre sektoren til en netto utslippskilde innen ca. fem år. Usikkerheten øker betydelig mot slutten av tidsserien.