Publications
NIBIOs employees contribute to several hundred scientific articles and research reports every year. You can browse or search in our collection which contains references and links to these publications as well as other research and dissemination activities. The collection is continously updated with new and historical material.
2025
Abstract
Efficient and objective measures of tree and stand structural complexity are essential to understanding the relationship between forest management, biodiversity, and ecosystem functioning, with laser scan-based structural complexity metrics playing an innovative role in monitoring and analysis.The objective is to develop an individual-tree crown complexity metric capable of distinguishing between structurally more or less complex trees while being scale-invariant. The second objective is to efficiently scale this metric to the stand level, enabling differentiation between forests of varying complexity and guiding precision silviculture.We developed a method to quantify crown complexity using individual-tree dense airborne LiDAR point clouds. Through optimization, we generate a 3D alpha shape crown model and calculate its volume and exposed surface area. This surface area is compared to that of a reference sphere with the same volume, as the sphere is the solid with the lowest surface-to-volume ratio, serving as a baseline for minimal complexity. This provides a scale-invariant measure of crown complexity. Summing this measure across all trees in a stand and applying a penalty for low vertical distribution yields a stand-level complexity metric that reflects structural heterogeneity.Applying our methodology to the FOR-instance dataset showed that the calculation of the 3D alpha shape crown model through optimization was successful, although sparse point clouds can present challenges. The crown complexity measure behaved as expected, ranking crowns according to their complexity, primarily determined by the roughness of the tree crowns, which increases the exposed surface area. When scaling the metric to the sample plot level, the measure effectively distinguishes between forests with structurally complex trees but low vertical stratification and those with less complex trees but high vertical stratification, identifying the latter as the more structurally complex forests.
Authors
Ryan BrightAbstract
Populærvitenskapelig Innlegg
Abstract
This paper explores the generation of “realistic” 3D representations of individual trees to enhance visualizations of forest simulation tool outcomes. By leveraging remote sensing data, we aim to capture individual tree features and characteristics accurately, linking them to dynamic simulations of forest structures and composition. Employing a deep learning approach, we train models on existing 3D scanned data to produce diverse and realistic visual representations of specific tree species. Our method addresses the limitations of existing synthetic tree generation techniques, which often overlook species-specific characteristics. Our approach emphasizes the generation of diverse tree forms, accounting for differences in trunk shape, canopy size, and branching structures. The resulting 3D data offers potential applications for realistic future forest visualizations and improved data augmentation in tree classification models, ultimately contributing to the creation of virtual forests that represent rich species diversity.
Authors
Thiago InagakiAbstract
No abstract has been registered
Authors
Lindner, Hanna NoemiAbstract
Plant Genetic Resources for Food and Agriculture (PGRFA) have declined drastically in recent decades, leaving the food system in a vulnerable state. In Norway, vegetable production relies heavily on imported seeds, which places the country in a fragile position regarding seed preparedness. To address this challenge, varieties adapted to Norway’s climatic and environmental conditions are needed, with landraces, characterized by their high genetic diversity, being particularly important. This case study applies a commoning approach to explore the role of KVANN, a Norwegian Community Seed Bank (CSB) together with the wider community of hobby gardeners, in facilitating access to seeds. The study combined 8 semi-structured interviews, two days participatory observation at a KVANN-seminar, and a questionnaire to hobby gardeners in Norway. Results indicate that although seed exchange is not the primary source of seeds and occurs only occasionally, these practices nonetheless impact management of plant genetic diversity, enable knowledge exchange and nurture trust in the community. By highlighting these dynamics, this case study contributes to the wider literature, revealing the important role hobby gardeners in Norway play in managing plant genetic diversity through seed and knowledge exchange. This is achieved by taking collective responsibility, collaboration with other actors in the seed system, and building trust among hobby gardeners. Ultimately, this thesis suggests that hobby gardeners, with KVANN as a central driver, can enhance access to diverse plant genetic resources and thereby contribute to seed-preparedness in Norway.
Authors
Mikarlsen, NoraAbstract
Pest control is a central part of modern strawberry farming. Spider mites are one of the most common pests in strawberries, and can cause significant reduction in yield. In order to properly manage and control spider mite populations, early detection is crucial. This thesis sets out to detect two-spotted spider mites (TSSM) in strawberries using hyperspectral imaging (HSI). A variety of methods have explored including visual inspection of the spectrum and its derivatives, as well as the use of vegetation indices (VIs). In addition, this thesis also explores machine learning (ML) and deep learning (DL) for early detection of TSSM. The mean spectrum from the images was used for classification in combination with Linear Discriminant Analysis (LDA) and Random Forest Classifier. Two separate Random Forest models were trained, one that distinguished between control, drought, and mite-infested strawberry plants, and one five-class with three different infestation levels, in addition to control and drought group. The three-class model achieved an F1-score of 0.86, while the five-class model had an F1-score of 0.845. The images themselves were used for classification by a ResNet18 model. The model was trained for each imaging day separately, and achieved accuracies in the range of 0.7-0.9 and F1-scores between 0.709-0.903. The work presented in this thesis highlights the capabilities of HSI in combination with ML and DL for early detection of TSSM in strawberries.
Authors
Hannah Rivedal Todd N. Temple Robert J. Starchvick Joseph P. Gallagher Dustin W. Herb Jason Crisp Nicole Anderson Tatiana Benedetti Amy B. PeetzAbstract
Oregon’s grass seed industry specialises in producing forage grasses including annual ryegrass (ARG, Lolium multiflorum), a host for the seed gall nematode (SGN, Anguina funesta). SGN causes yieldlimiting seed galls and are strictly regulated in international trade. From 2019 to 2020, over 500 metric tons of Oregon ARG seed were rejected from international ports due to SGN detection. A 2022 field survey of 22 ARG fields in the Willamette Valley of Oregon resulted in SGN detection in 50% of the fields throughout the growing season. Several approaches managing SGN are under evaluation. Previous reports indicate that there may be genetic resistance to SGN in other Lolium species. Therefore, a breeding population of 240 public accessions of L. multiflorum have been seeded with two seed galls and planted in the field. Seed were harvested to evaluate for galls in July 2025 and to identify potential resistant families for future study. To date, no nematicides are labelled for the control of SGN. Varied fluopyram timings and rates, as well as an untreated control, are being evaluated in the field with and without growth regulation for SGN control. Seed yield and galled seed data was collected showing limited differences between treatments. Cultural control methods are also being considered, including seed cleaning and utilizing high energy pulses on seed galls. Preliminary data suggests that these could be viable treatments to reduce SGN inoculum. Successful control options for the SGN in ARG seed production are important to reduce the spread of this nematode globally and maintain healthy forage production.
Authors
Seth Dorman Darrin Walenta Navneet Kaur Hannah Rivedal Nicole AndersonAbstract
Successful mitigation of agricultural insect pests depends on integrated pest management practices incorporating multiple techniques for effective population management below economically damaging levels. Pest surveillance remains the cornerstone of IPM programs, enabling appropriately timed management action. Phenological models using weather data are commonly used as decision support tools to predict the timing of ‘when’ economically important life stages are expected to occur. Furthermore, geostatistical models that consider landscape-level variation in environmental drivers of pest densities may inform ‘where’ outbreaks are expected to occur along spatial gradients of abiotic and biotic risk factors. Thus, predictors of pest populations can be leveraged to generate spatiotemporal risk assessments. Migratory and overwintering moth species in the family Noctuidae (order Lepidoptera) comprise a complex of serious pests that threaten the profitability of grasses grown for seed in the Willamette Valley, Oregon, USA, as well as seed production globally. Noctuid pests, including black cutworm, true armyworm, and winter cutworm, inflict crop damage by direct feeding on plant crowns and roots in the larval developmental stage. Management action with foliar insecticides is most effective when larvae are immature (early instar stage) due to insecticide susceptibility and pest behaviour. For noctuid pests (and other priority pests) in grass seed systems, real-time phenological models can provide field practitioners with information to better allocate pest monitoring and management resources to reduce input costs.
Authors
Jing Zhou Nicole Anderson Marshall Q. Garrett Logan P. Snell Thomas G. ChastainAbstract
Seed moisture content (SMC) is the most reliable indicator of optimal harvest timing in many seed crops, including grass seed. Measuring SMC as grass seed crops approach maturity is recommended to determine optimal harvest timing. Currently, to measure SMC, seeds need to be stripped from heads by hand, weighed, dried until all the moisture has been lost, then re-weighed, and SMC manually calculated. Collecting SMC information in this manner is arduous, time consuming, and prone to error. Consequently, this procedure has resulted in inadequate SMC testing or failure to test in a timely fashion. In addition, SMC is an important factor in the storage of harvested seed, which typically needs to be stored under 12% SMC, to ensure high seed quality. The goal of this project is to develop a portable sensing device - The Grady Sensor - that allows for rapid and accurate SMC measurement of grass seed. The sensor employs near-infrared (NIR) spectroscopy principles that water molecules absorb specific NIR wavelengths. By analysing the light reflected from the seed surface, the sensor predicts SMC based on the intensity of the reflected light at moisture-sensitive wavelengths. Over two years, multiple sensor prototypes have been developed, and their performance has been validated through field tests in Oregon and New Zealand. Sensor readings were compared to laboratory oven gravimetric SMC values of samples collected from major grass seed species, including tall fescue, annual ryegrass, perennial ryegrass, orchardgrass, creeping red fescue, creeping bentgrass, Chewings fescue, and Kentucky bluegrass. The sensor readings demonstrated a significant linear relationship with the oven SMC. Mean absolute errors of sensor SMC predictions were within 1.2 to 4.6% across all grass species. The results indicate that the prototype is a reliable replacement for the traditional oven drying method.
Authors
Navneet Kaur Seth J. Dorman Christy Tanner Nicole AndersonAbstract
Tychius picirostris Fabricius (Coleoptera: Curculionidae), the clover seed weevil (CSW), is a major pest of white clover (Trifolium repens L.) seed crops in Oregon, USA, where larval feeding during seed development reduces yield and quality, causing substantial economic losses. For decades, CSW management has relied on bifenthrin (IRAC Group 3A), but since 2017 growers have reported reduced efficacy, prompting concern of resistance. Laboratory bioassays in 2022-2023 confirmed very high resistance to bifenthrin (RR50 = 178-726) and moderate resistance to malathion (Group 1B; RR50 = 7.8-32.8), underscoring the need for alternative chemistries and insecticide resistance management (IRM) guidelines. From 2022 to 2024, on-farm insecticide efficacy trials were conducted in commercial white clover seed fields in western Oregon. Early-season (pre-bloom or PB) application using contact insecticides (malathion, isocycloseram [Group 30], indoxacarb [Group 22]) targeted adults, either alone or in sequence with mid-season (full bloom or FB) application of systemic insecticides (chlorantraniliprole and cyantraniliprole [Group 28]) targeting larvae. Adult abundance was monitored with 20-sweep net samples, and larval densities were estimated from 30 inflorescences per plot extracted with Berlese funnels. Across sites and years, isocycloseram consistently suppressed adult populations, cyantraniliprole reduced larval densities, while indoxacarb showed variable performance. Although seed yields did not differ significantly among treatments, yet efficacy data supported product registration in Oregon and highlighted the value of chemical rotation plans for resistance management. Based on these findings, we recommend discontinuing bifenthrin and adopting an integrated resistance management (IRM) program that applies contact insecticides during spring adult migration when ≥2 weevils per sweep are detected and systemic insecticides during full bloom when ≥3 per 30 inflorescences are observed.