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.
2026
Authors
Belachew Gizachew ZelekeAbstract
Tropical forests, despite their critical environmental and socio-economic roles, remain highly vulnerable to deforestation, forest degradation, and climate-related disturbances. There is a growing demand for robust and transparent forest monitoring systems, particularly under REDD+, the Paris Agreement’s Enhanced Transparency Framework (ETF), and emerging climate-finance mechanisms. Conventional approaches based on field inventories and traditional remote sensing are often constrained by limited or uneven field data, persistent cloud cover, complex forest conditions, and limited institutional and technical capacity. This review examines how artificial intelligence (AI) and machine learning (ML) are being integrated into remote sensing–based tropical forest monitoring to address these structural constraints. Using a semi-systematic synthesis of peer-reviewed studies, complemented by operational platforms and grey literature, the review assesses AI/ML approaches, remote sensing datasets, and applications relevant to national and large-scale monitoring. Evidence is synthesized across five analytical dimensions: AI/ML model families and workflows, multi-sensor datasets and training resources, operational monitoring platforms, application domains (including deforestation, degradation, and biomass/carbon estimation), and cross-cutting technical, institutional, and governance barriers. The review finds that AI/ML-enabled remote sensing, particularly those combining optical, radar, and LiDAR time series within cloud-based platforms, has substantially improved the automation, scalability, and speed of tropical forest monitoring. However, effective and equitable adoption remains constrained by limitations in training and validation data, dependence on proprietary platforms and data, uneven technical capacity, and unresolved governance and ethical challenges. Emerging solutions, including open and representative training datasets, platform-agnostic processing infrastructures, long-term capacity building, and inclusive data-governance frameworks, are identified as critical enablers of credible and nationally owned AI/ML-enabled forest-monitoring systems. The review highlights that AI/ML can play a transformative role in supporting climate mitigation, biodiversity conservation, and informed decision-making. This potential, however, depends on transparent data governance arrangements, long-term capacity building, and platform-agnostic infrastructures that support national ownership.
Authors
Karen Ane Frøyland Skjennum Jan Mulder Gijsbert Dirk Breedveld Thomas Hartnik Nicolas Estoppey Erlend Grenager SørmoAbstract
The present study investigates the long-term immobilization efficiency of biochar on target per- and polyfluoroalkyl substances (PFAS) and precursors in well-drained soils contaminated by aqueous film-forming foam (AFFF) (Ʃ27PFAS = 1624 ± 276 µg/kg) over 2 years. The total oxidizable precursor (TOP) assay revealed a large precursor reservoir in the soil. Fifteen outdoor field-scale columns were packed with contaminated soil (48 kg) without (control columns, triplicates) and with biochar amendments: Three sewage sludge-based biochars were homogeneously mixed into the soil at a 1% (w/w) dose in triplicate columns. One of the biochars was additionally applied as a barrier at the column base (1% w/w) in a separate set of columns. The best-performing biochar immobilized long-chain PFAS by 91.0 ± 35.0% and short-chain PFAS by 96.7 ± 32.9%, possibly due to a well-developed porosity. Compared to the control columns, the fluctuating PFAS leaching were negligible in columns amended with the best-performing biochar, but the immobilization efficiency of short-chain PFAS decreased after one year (from 97.8% to 74.2%). Applying biochar as a barrier was two times more effective than homogenous mixing, and the effect was most pronounced for long-chain PFAS. Our findings suggest that biochar may immobilize precursors, notably CF3-CF5 precursors, to the same extent or better than their typical target perfluoroalkyl acids transformation products. More research is, however, needed to confirm these trends. Going beyond simple lab experiments, this study suggests that biochar is a promising solution for PFAS remediation and brings the technology closer to field application.
Authors
Ulrika Jansson Asplund Damian Petkovic Karlsen Anne Krag Brysting Rune Halvorsen Håvard Kauserud O. Janne Kjønaas Johan AsplundAbstract
No abstract has been registered
Authors
Ming Yu Sebastian Kepfer-Rojas Yamina Micaela Rosas Teresa Gómez de la Bárcena Inger Kappel Schmidt Per Gundersen Ludovica D'Imperio Carsten W. Mueller Lars VesterdalAbstract
Afforestation of agricultural land is widely promoted as a nature-based solution to enhance carbon (C) sequestration and mitigate atmospheric CO2 levels. However, the temporal dynamics of soil organic carbon (SOC) after afforestation, particularly in mineral soils, remain uncertain due to the complex interaction of biogeochemical processes and their spatial variability. We investigated changes in SOC sequestration over five decades of afforestation on former cropland by extending the chronosequence approach with three repeated soil inventories in oak (Quercus robur L.) and Norway spruce (Picea abies (L.) Karst.) stands. Aboveground biomass C stocks were also quantified to evaluate the contribution of SOC to post-agricultural ecosystem C stocks. Forest floor C stocks increased rapidly in the early years and stabilized after approximately three decades, with consistently higher accumulation under Norway spruce than oak. In contrast, mineral SOC stocks in 0-25 cm depth increased with forest age by 0.18 ± 0.06 Mg ha−1 yr−1 under oak and 0.44 ± 0.07 Mg ha−1 yr−1 under Norway spruce. These contrasting trends in forest floor and mineral soil indicated a shift in C source-sink strength over time and between species. After 50 years of afforestation, total ecosystem C stocks in afforested stands reached up to 75% of those in a 200-year-old forest, with most new C stored in biomass (84-86%), followed by mineral soil (10-11%) and forest floor (4-5%). Despite higher sequestration of new C in Norway spruce stands, the relative distribution across ecosystem compartments was similar between tree species.
Authors
L. Duncanson P. M. Montesano A. Neuenschwander A. Zarringhalam N. Thomas D. M. Minor M. A. Wulder J. C. White E. Guenther T. Feng V. Leitold S. Hancock J. Armston Stefano Puliti A. I. Mandel S. Shah C. Silva M. Purslow J. Bruening Johannes Breidenbach Erik Næsset Svetlana Saarela N. Hunka J. R. Kellner S. P. Healey D. Schepaschenko J. Wallerman C. S. R. Neigh N. Carvalhais R. DubayahAbstract
Forest aboveground biomass Density (AGBD) maps provide important constraints on climate and carbon cycle models and enable the long-term monitoring of carbon stocks. NASA's latest spaceborne lidar instruments provide unprecedented observations of forest structure that are used to compile spatially continuous, locally trained maps of circa 2020 AGB density and stock. To address a geographical limitation of the International Space Station deployed GEDI instrument, herein we map high northern latitude forests with a fusion of NASA's ICESat-2 mission, Harmonized Landsat Sentinel-2 (HLS), and Copernicus GLO-30 topographic data. We report a domain-wide estimate for boreal forests of 72.96 +/− 0.61 Pg AGB. Combining these maps with those of tropical and temperate forests from GEDI products provides a global estimate of 2020 biomass stocks of 593.49 +/− 11.47 Pg AGB. We analyze biomass means and totals across the boreal domain in different land cover types, slope classes, and ecoregions. We compare these products to national estimates of AGB stocks for high latitude countries, finding good general agreement between national reports and EO-based estimates, particularly for countries with robust National Forest Inventories (NFIs) and boreal forests, while underestimation of high AGBD in tall, dense forests remains a challenge for AGBD mapping. Satellite products can be used to support wide area estimation of aboveground biomass, and the maps described here provide insights into carbon stocks, patterns, and dynamics at scales relevant to forest management. These open access data products serve as a baseline to assess future changes associated with drought, fire, insects, deforestation, and degradation.
Authors
José Luís Trevizan Chiomento Nícolas Alberton Mandelli Ana Júlia Schroeder Ely Matheus Welter Henrique Didó Maria Antônia Lacourt Oliveira Bruna Jirkowski Francisco Wilson Reichert Junior Mateus Possebon Bortoluzzi Rafael Rieder Thomas dos Santos Trentin Anita SønstebyAbstract
This study investigated the effect of arbuscular mycorrhizal fungi (AMF), Trichoderma harzianum, and their combinations on yield and quality of strawberry fruits in a soilless cultivation system. Six treatments were applied, control (no biostimulants), T. harzianum (TH), Claroideoglomus etunicatum (CE), a multispecies mycorrhizal community (IP CS), TH + CE, and TH + IP CS, arranged in a randomized block design with four replicates. While total fruit yield was not significantly affected, the application of T. harzianum, either alone or in combination with AMF, enhanced cumulative fruit production. The use of C. etunicatum improved sugar content and the sugar/acid ratio by 28% and 31%, respectively, compared to the control. Biostimulant treatments also increased total phytochemical content, particularly with the multispecies inoculant IP CS (increased anthocyanin content by 39% compared to the control) and the combinations TH + CE (flavonoid content 41% higher than the control) and TH + IP CS (flavonoid content 39% higher than the control). Multivariate analysis grouped the treatments into two groups, with the control (no biostimulants) forming a distinct group. In conclusion, biostimulation of ‘San Andreas’ strawberry plants improved fruit quality without significantly increasing yield. The combined use of AMF and T. harzianum is proposed as a sustainable strategy for enhancing fruit quality in soilless strawberry cultivation systems.
Authors
Milica Fotirić Akšić Mirjana Pešić Ilinka Pećinar Mihajlo Jakanovski Danijel Milinčić Aleksandar Kostić Marko Kitanović Uroš Gašić Dragana Dabić Zagorac Dušanka Milojković Opsenica Mekjell MelandAbstract
This study included the nectar of nine standard apple (Malus × domestica) cultivars (‘Red Aroma’, ‘Discovery’, ‘Summerred’, ‘Rubinstep’, ‘Elstar’, ‘Asfari’, ‘Eden’, ‘Fryd’, and ‘Katja’) and two crab apple (Malus sylvestris) cultivars (‘Dolgo’ and ‘Professor Sprenger’). The aim was to determine the diversity of chemical compounds in the floral nectar of the two different apple species and their cultivars. Chemical analysis identified five sugars, two sugar alcohols, two organic acids, forty phenolic compounds, and five phenylamides. The crab apples ‘Dolgo’ and ‘Professor Sprenger’, along with the commercial cultivar ‘Rubinstep’, had the highest levels of all three main sugars (glucose, sucrose, and fructose). The cultivar’s ‘Katja’ nectar had the highest level of total phenolic content (60.7 mg/100 g GAE), the nectar sample from ‘Dolgo’ exhibited the greatest ability to neutralise DPPH radicals (83.4 mg/100 g TE), and the ‘Dolgo’ (100.6 mg/100 g TE FW) and ‘Katja’ (72.1 mg/100 g TE FW) nectars proved to be the best reducing agents. Floral nectar from ‘Eden’ and ‘Fryd’ showed very high levels of isorhamnetin, 49.04 mg/kg and 50.83 mg/kg, respectively, while nectar from ‘Katja’ had the highest level of gentisic acid at 39.06 mg/kg. Besides being vital for insects, apple floral nectar is a significant reservoir of phenolic compounds and can be considered a “superfood” for the human diet.
Authors
Chen‐Jui Yang Tomas Funda Rajiv Chaudhary Zhi‐Qiang Chen Maximiliano Estravis Barcala Harry WuAbstract
Genetic differentiation among populations often varies significantly across the genome due to factors such as selection and recombination, resulting in a heterogeneous genomic landscape. However, variation in low‐differentiation regions—genomic valleys—remains poorly understood. Moreover, most insights into plant genomic landscapes come from flowering plants, while comparable genome‐wide studies in other taxa, such as conifers, remain limited. We analyzed whole‐genome sequencing data from 100 individuals of three pine species— Pinus banksiana , Pinus contorta , and Pinus nigra . We found substantial genome‐wide variation in recombination rates, with intergenic regions exhibiting higher recombination than genic regions, and rates decreasing with increasing distance from genes. Recombination rate was negatively correlated with gene length, driven primarily by intron length, suggesting that long introns in conifers may promote the retention of exceptionally long genes by maintaining low recombination in these regions. Genomic scans further revealed that genomic valleys are maintained through either balancing, background, or parallel selection. Additionally, multiple forms of selection were strongly associated with local recombination rate variation, highlighting the significant role of recombination in shaping patterns of genomic differentiation. Our findings provide new insight into the evolution and maintenance of extremely long genes in conifers. Moreover, the results indicate that allopatric selection in regions of low recombination is a major force structuring genomic variation in these species.
Authors
Radosav Cerović Milica Fotirić Akšić Milena Ðorđević Marko Kitanović Anđelija Obradović Mekjell MelandAbstract
This study examines the reproductive biology of five widely cultivated apple cultivars in Norway (‘Discovery’, ‘Rubinstep’, ‘Red Aroma’, ‘Elstar’, and ‘Asfari’), when crossed with the main pollenizers (‘Summerred’, ‘Discovery’, ‘Katja’, ‘Rubinstep’, ‘Red Aroma’, ‘Fryd’, and ‘Eden’, and two crab apples ‘Professor Sprenger’ and ‘Dolgo’), as well as under self-pollination and open pollination. The experiment was conducted over two seasons (2022–2023) in Hardanger, a region in Western Norway. Flowering time and overlap, in vitro pollen germination, pollen tube growth within the styles and ovary, embryo sac viability, fertilization success, and fruit set were analyzed as key reproductive parameters. Under broadly comparable climatic conditions across both seasons, the results showed that both mother cultivar and the pollenizer strongly influenced progamic processes and fruit set. Pollen tube growth through the pistil was generally faster and more successful in 2022 for all pollination combinations, resulting in a higher fruit set. The only exception was ‘Elstar’, which exhibited a higher fruit set in 2023. If a single optimal pollenizer were to be selected for each apple cultivar in Western Norway, it would be ‘Red Aroma’ for ‘Discovery’ and ‘Rubinstep’; ‘Rubinstep’ for ‘Red Aroma’ and ‘Elstar’; and ‘Professor Sprenger’ for ‘Asfari’. Based on pollen tube growth in vivo and the fruit set, cultivars ‘Discovery’, ‘Rubinstep’, ‘Red Aroma’, ‘Elstar’, and ‘Asfari’ showed self-incompatibility.
Authors
Rajiv Chaudhary Maximiliano Estravis Barcala Irena Fundova Tomáš Funda Zhi-qiang Chen Harry X. WuAbstract
Background Drought intensity and frequency are increasing under global warming in the boreal forests, and breeding for drought resistance will facilitate adaptation of new planting material to changing climate conditions. We used a tree-ring dataset of 559 individuals to study Scots pine genetic variation and the efficiency of genomic selection of drought-response traits (drought resistance, recovery and resilience), for the first time. From genotyping-by-sequencing (GBS), 31,101 SNPs were generated and used for the study. Results Significant genetic variation was detected for drought-response and other growth, wood-anatomy and wood density traits. Heritability estimates for wood-anatomical traits were higher than those for drought-response and growth traits. Genetic correlations between drought-response and wood-anatomical traits were generally high but mostly nonsignificant. In contrast, drought resistance and recovery showed positive and significant correlations with basal area increment and height. We found that the predictive ability and accuracy for drought-response traits were lower than those for wood-anatomical traits, and were comparable between GBLUP and ABLUP. Greater genetic gain per year can be achieved through genomic selection relative to pedigree-based selection if the generation interval is reduced. Conclusions The positive genetic correlation between drought-response and growth traits will enable simultaneous selection for improved growth and increased drought resistant trees in Scots pine breeding through either pedigreed-based and genomic selection.