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
Tveter, Celine Sophie DollySammendrag
Mangrove forest is an important ecosystem, as it serves several roles both on a local and global scale. In addition to supporting biodiversity and local communities, mangroves act as major carbon sinks. Accurate estimation of above-ground biomass (AGB) is therefore important for assessing carbon storage and potential carbon emissions. However, field-based biomass estimation is resource intensive and difficult to conduct across large or inaccessible areas, making national-scale estimates challenging. This study aims to develop and compare machine-learning approaches for predicting and estimating mangrove forest above-ground biomass density (AGBD) using Global Ecosystem Dynamics Investigation (GEDI)-derived reference data and multi-sensor satellite predictors for coastal Tanzania. The objectives of this study is to: 1) develop and compare a set of machine-learning regression frameworks for AGBD estimation using robust validation strategies appropriate for spatial data, 2) quantify the performance of all the frameworks and the uncertainty of the best framework, and 3) produce a wall-to-wall map of mangrove AGBD for coastal Tanzania and compute the mean density and total biomass at national and regional level. Twelve frameworks were developed by combining the regression algorithms Random Forest, eXtreme Gradient Boosting, Support Vector Regression, and Ridge with the cross-validation strategies completely random, randomized stratification, and grouped stratification. The best-performing framework was further optimized and used for the final prediction. Random Forest with randomized stratification achieved the best overall performance among the tested frameworks, with test RMSE of 29.3 \si{\mega\gram\per\hectare} and R^2 of 0.36. After optimization, the metrics improved to 27.4 Mgha^-1 and 0.39 for RMSE and R^2, respectively. The final prediction estimated a national mangrove AGB of 4.05 million Mg for 2021, with a 95\% confidence interval of [3.93, 4.17] million Mg, corresponding to an average density of 30.98 Mgha^-1. The results showed the potential of combining multi-sensor remote sensing data and GEDI-derived biomass estimates for large-scale mangrove biomass estimation. However, uncertainties related to the GEDI reference data, signal saturation, and model heteroscedasticity, especially at high densities, should be considered when interpreting the results.
Forfattere
Nordfjellmark, LoviseSammendrag
Human land use change, including agricultural intensification, has resulted in global declines of pollinators, such as bumble bees. Increased agricultural intensification has reduced availability and distribution of floral resources and nesting habitats within landscapes, which may have cascading effects on bumble bee communities. This study explores how flower availability and landscape structure in the highly intensive agricultural landscapes of Southwest Norway affects local bumble bee diversity, abundance, and worker body size. I collected data on local bumble bee abundance and diversity using insect-traps, as well as floral resource abundance during three registrations periods during the summer of 2025. Intertegular distance (ITD) was used as a measure of bumble bee body size, since it is also an indicator of flight muscle volume, and therefore foraging capacity. I found no clear effect of total flower abundance, nor landscape heterogeneity on total bumble bee abundance, or on bumble bee worker body size. However, long foraging species were observed to be the most abundant, which suggests prior selection by the study landscape for species with long foraging range. I also found strong interspecific differences in intertegular distance among species, with Bombus s.str, the most abundant taxa, being significantly larger than the other analyzed species. This suggests that larger body size confers an overall advantage in homogenous landscapes.
Forfattere
Gramstad, LeneSammendrag
Agricultural intensification and other land use changes have had detrimental effects on pollinating insects. As such, identifying measures to slow or reverse their population declines is critical. Sown flower strips are one such mitigating measure. I evaluated the effects of annual and perennial flower strips on communities of bumble bees and butterflies by assessing abundance, richness, visitation rates, and movement of pollinators in Ås, Norway. Flower abundance was recorded through quadrat sampling complemented by site-level observations, while pollinator abundance and richness were recorded through repeated transect walks, during which bumble bees and butterflies were counted and identified, and butterflies were marked and recaptured. Annual flower strips were found to have the highest flower abundance and benefitted both pollinator groups, whereas perennial flower strips primarily benefitted bumble bees and had flower abundances more similar to grass dominated road verges. Bumble bee and butterfly abundances did not decrease with distance from the sown flower strips, suggesting high connectivity between the flower strips and a positive effect of landscape context on pollinator movement. The results suggest that annual flower strips conferred the greatest benefit to pollinators in the study system, and that perennial flower strips require improvements in floral resources to better support pollinator communities.
Forfattere
Lund, ØyvindSammendrag
Coastal vegetation provides important ecosystem services such as carbon sequestration, wildlife habitat, and protection from erosion, and mapping the different vegetation types is vital for their protection. There has been little research on classification and mapping of the coastal vegetation in Western Africa, and especially in Ghana. This study classifies the coastal vegetation of Ghana using both high-resolution UAV imagery and Sentinel-2 imagery and compares three machine learning models for the task: Random Forest, Support Vector Machine and Extreme Gradient Boosting. The classification was done using an Object-Based Image Analysis (OBIA) workflow, where the imagery was segmented, labeled and classified using spectral, textural, and geometric features. The UAV classification was also used as an additional source of training data for the Sentinel-2 classification. XGBoost achieved the highest scores for the UAV classification with a macro F1-score of 0.858, though all three classifiers performed comparably, while SVM performed best for the Sentinel-2 imagery with a macro F1-score of 0.817. The integration of UAV training data worked well for classes that were detectable at 10 m resolution, but the limited spatial coverage of the UAV imagery meant it could not provide enough samples for the smaller classes. The final output is a coast-wide vegetation map of Ghana, where the total mangrove area was estimated at 102.84 km2, which is lower than in previous studies. The main sources of error were the lack of field data, limited spectral and spatial coverage of the UAV imagery, and the broad class definitions. Despite this, the study shows that combining UAV and Sentinel-2 imagery in an OBIA workflow can produce a consistent coast-wide classification of coastal vegetation. It also highlights the need for probability-based area estimation in future work to produce more reliable and comparable area estimates.
Forfattere
Frøydis GillundSammendrag
Innlegg på fagdag om matproduksjon, beredskap og fremtidens landbruk på Gibostad
Forfattere
Anna Amberntsson Dina Moxness Konglevoll Annika Eliassen Mari Mohn Paulsen Anne Lise Brantsæter Marianne Hope Abel Lene Frost AndersenSammendrag
Abstract Purpose: Alternatives to animal-sourced foods (Alt-ASFs) are processed plant-based products that mimic animal-sourced foods. This study aimed to describe the consumption of Alt-ASFs in the Norwegian dietary survey (Norkost 4) and compare population characteristics and nutritional intake in regular consumers. Further, to compare the nutritional quality of Alt-ASFs consumed and describe the distribution of the Nutri-Score and NewTools-score - a proposed revised version of Nutri-Score, among these products. Methods: Dietary data were collected from 1,815 adults in Norkost 4 using 2*24-h recalls and a food frequency questionnaire. Alt-ASFs were identified and compared with corresponding ASFs. Nutritional quality was assessed and products were scored using Nutri-Score and NewTools-score. Results: Overall, the consumption of Alt-ASF was low (mean: 0.3 energy %, SD: 1.6). 8% ( n = 147) reported consuming one or more Alt-ASFs regularly, defined as ≥ 2 times/week, with dairy alternatives being most frequently consumed. Regular consumers were typically younger, female, and more likely to report a vegetarian diet. Most Alt-ASFs contained lower levels of the assessed micronutrients except for meat and fortified milk, whose micronutrient content was similar or higher than ASFs. There was a wide variation in nutritional scores for Alt-ASFs. Unfortified Alt-ASFs typically had a low micronutrient content but could still achieve favourable Nutri-Scores if low in salt, sugar and saturated fat. Conclusion: Our findings indicate that only a minor part of Norwegian adults regularly consumes Alt-ASFs and consumption is low. Shifting from ASFs to Alt-ASFs may have some nutritional benefits and some disadvantages, largely dependent on the chosen products.
Sammendrag
Abstract Algal cultures are widely used to study algae–bacteria interactions, yet the relative roles of algal taxonomy, culturing environment, and historical contingency in shaping bacterial communities of long-maintained monoalgal cultures remain unclear. In a two-stage experimental design, we used 16 S rRNA metabarcoding to first profile the bacterial communities of 14 filamentous freshwater algal strains from three orders (Chaetophorales, Oedogoniales, Zygnematales) maintained in two distinct culture collections, and then, in a second experiment to test community mixing using a co-culture and re-isolation design with two algal strains ( Stigeoclonium sp. and Oedogonium sp.). The 14-strain profiling showed that bacterial communities were broadly similar at higher taxonomic ranks, with many shared orders within Pseudomonadota, whereas composition at the ASV level was largely algal-strain-specific. Ordination of phylogeny-aware UniFrac distances showed that culture collection of origin explained more variation in community composition than algal order, an effect reduced but not eliminated after standardizing the culture medium. The co-culture and re-isolation experiment showed that the two algal strains exchanged many bacterial community members, yet algal strain identity remained detectable after re-isolation. These results support a framework where culture history results in strain-specific ASV-level composition within recurrent higher-rank structure, and where algal-strain-specific effects on bacterial community composition become evident when cultures are exposed to a shared bacterial pool. Co-culture followed by algal re-isolation offers a practical approach to examine the persistence of algal-culture bacterial communities under mixing and may help identify candidate bacterial partners with potentially strong host-specific interactions without requiring axenic hosts.
Sammendrag
Abstract Single-date UAV vegetation indices are widely used to assess potato crop performance and estimate yield. However, whether their biological interpretation remains consistent across cultivars and growth stages is rarely examined. This study investigated how the relationships between UAV-derived multispectral and thermal canopy indicators and potato yield changed during crop development under contrasting irrigation in two cultivars with contrasting maturity classes (Anouk and Celandine). UAV surveys were conducted in July and August, and canopy metrics were related to final yield and tuber traits. The relationship between NDVI and yield differed markedly between cultivars and growth stages. In the early-maturing cultivar Anouk, July NDVI showed a strong positive correlation with yield ( r = 0.87, p = 0.002), whereas August NDVI was not informative. In the late-maturing cultivar Celandine, the relationship reversed from a positive correlation in July ( r = 0.56) to a strong negative correlation in August ( r = −0.80, p = 0.009), indicating that greater late-season canopy greenness was associated with lower yield. Exploratory LOOCV Random Forest models achieved the highest yield prediction accuracy using July features for Anouk ( R 2 = 0.72) and August features for Celandine ( R 2 = 0.47), while multi-temporal features provided similar predictive performance. Greater temporal declines in NDVI were associated with higher yield in both cultivars, consistent with the change in canopy greenness acting as a phenological indicator of the transition from vegetative growth to tuber bulking. These results demonstrate that the interpretation of UAV-derived vegetation indices is strongly dependent on cultivar maturity and growth stage. A single UAV acquisition may therefore provide misleading assessments of crop performance when canopy phenology is not considered. Rather than serving as fixed proxies for yield, canopy vegetation indices should be interpreted within cultivar-specific and temporal contexts when used for potato monitoring and agronomic decision-making.
Sammendrag
Understanding how soil porosity varies across environmental gradients is critical for interpreting land–surface processes, including water storage, flow pathways, gas exchange, and carbon cycling. While organic carbon (OC) and soil texture are important influences on soil structure, their combined influence on porosity at landscape to continental scales remains incompletely resolved. Here, we compile topsoil datasets from Great Britain, Wales, the European LUCAS survey, and the USDA-NRCS NCSS database (n = 6335) to quantify relationships between porosity, OC, and clay content across diverse environmental conditions. Focusing on minimally disturbed soils, we evaluate individual and combined controls using generalized additive models. We develop and test simple empirical descriptors of OC:clay ratio, including a signed metric (NonCOCC) representing deviation from a reference ratio of 1 g OC to 10 g clay, positioning soils along a continuum from clay excess to OC excess. Across all regions, OC has the strongest relationship to porosity, explaining substantially more variation than OC:clay ratio metrics. NonCOCC provided a modest improvement in model performance when combined with OC and clay. Across regional datasets, the steepest portions of the fitted NonCOCC partial responses generally occurred at low positive values, suggesting a region of elevated porosity sensitivity. In a subset with aggregate-size data, this range of NonCOCC values also corresponds to a shift towards larger aggregates. This may suggest a structural pathway linking organic enrichment, aggregation, and porosity, but requires testing using data with measured OC partitions. The findings confirm importance of OC for predicting porosity and identify added value of simple metrics (NonCOCC) from widely available soil observations.
Forfattere
J. A. Woolliams M. Janes H. Tumas J. J. Ilska Paul McLean J. E. Cottrell J. Mackay G. GorjancSammendrag
Abstract Genomics is advancing estimation of breeding values for selective breeding. Traditional pedigree information used in the estimation of breeding values can still be competitive to genomic information when markers do not adequately capture genetic variance or genomic datasets are not sufficiently large. Using the UK Sitka spruce breeding population as a study system, we assessed the accuracy of estimated breeding values using pedigree or genomic information. Nine phenotypic traits associated with growth and wood quality were measured across 1295 individuals in 50 full-sib families replicated across two plantation sites. Pedigree-based estimation was compared to genome-based estimation using identity by state (IBS) or identity by descent (IBD) information from 4123 SNPs. Accuracy of estimated breeding values was assessed using model-based standard errors, which assumes the fitted model and parameters are correct, and more objective 12-fold cross-validation by withholding some phenotypes from the model fitting. Accuracies obtained directly from the model were very similar across different types of information, but cross-validation showed genomic information gave highest accuracies for all but one of the traits. No bias was observed for any of the trait- information combinations. Results demonstrate that the estimation of breeding values with genomic information in this study system is already competitive with the pedigree information, despite evidence that the marker panel did not capture the full genetic variance. Furthermore, accuracies with genomic information have the potential to significantly improve, unlike pedigree, by increasing the size of the training set and marker density.