Luke

Silmäile

 
Hakuohjeet

Jukuri

Tervetuloa käyttämään Jukuria, Luonnonvarakeskuksen (Luke) avointa julkaisuarkistoa. Jukurissa on tiedot Luken julkaisutuotannosta. Osa julkaisuista on vapaasti ladattavissa. Luken muodostaneiden tutkimuslaitosten aikaisemmasta julkaisutuotannosta osan tiedot ovat järjestelmässä jo nyt ja kattavuus paranee jatkuvasti.

Kokoelmat

Viimeksi tallennetut

Tailored biosecurity training for veterinarians and farmers: bridging knowledge and practice gaps
Mehmedi, Blerta; Niemi, Jarkko; Saegerman, Claude; De Meneghi, Daniele; Iatrou, Anna Maria; Yildiz, Ramazan; Chantziaras, Ilias; Allepuz, Alberto; Toppari, Ina; Batikas, Georgios; Viltrop, Arvo; Niine, Tarmo
Frontiers in veterinary science (Frontiers Media S.A., 2025)
Biosecurity is fundamental to animal health, public health, and the economic resilience of livestock systems; however, farm-level adoption remains uneven across regions. Knowledge gaps, language and financial constraints, and limited communication competence among veterinary advisers impede implementation, especially on small- and medium-scale farms. Behavior change-oriented interventions, such as Motivational Interviewing (which deploys multiple specific behavior change techniques as defined in BCTTv1), offer promise but are seldom embedded in veterinary curricula. This study proposes a concept and key elements for biosecurity training. It highlights a modular, evidence-based training framework developed under the COST Action CA20103 “BETTER” (2021–2025), aimed at improving biosecurity understanding and implementation by veterinarians and farmers. The initiative convened European experts to co-design a flexible curriculum that addresses both technical and behavioral challenges using participatory methods and interdisciplinary expertise. The resulting framework consists of five progressive modules: (1) Introduction, (2) Behavior Change and Communication, (3) Disease Transmission & Risk Assessment, (4) Emergency Response & Clinical Biosecurity, and (5) On-Farm Practices. These modules are designed to be combined in a “pick-and-choose” format to match local needs, target audiences and resources. Delivery blends online micro-lessons, participatory workshops, peer networks, and low-cost on-farm demonstrations, while materials are culturally and linguistically adapted and framed in terms of clear economic benefits. Continuous feedback loops encourage iterative refinement and habit formation during the learning process. The proposed training framework seeks to transform biosecurity from a prescriptive doctrine into a farmer-owned daily routine by integrating technical content with behavioral science and context-specific delivery.
Comparing machine learning algorithms for simultaneous prediction of tree diameter distribution percentiles
Ciceu, Albert; Aksoy, Hasan; Badea, Ovidiu; Bullock, Bronson; Ezenwenyi, Jacinta Ukamaka; Gorgoso-Varela, Jose Javier; Leca, Ştefan; Ledermann, Thomas; Mäkinen, Harri; Ogana, Friday N.; Yang, Sheng-I; Mehtätalo, Lauri
Ecological informatics (Elsevier, 2025)
Accurate predictions of tree diameter distributions are important for assessing forest structure, quantifying biodiversity, and estimating carbon sequestration. Percentile-based approaches are among the most effective methods for reconstructing diameter distributions from stand-level variables. In this study, we compared three modelling approaches, generalised least squares (GLS), Multi-Output Random Forest (MORF), and a multi-output deep learning-based model (MODL), across nine datasets representing different forest types and management regimes, aiming to predict simultaneously six diameter distribution percentiles. Our results show that MODL consistently outperformed both GLS and MORF in predictive accuracy across all nine training subsets and five out of nine test subsets, demonstrating strong generalisation across diverse forest types. MODL was particularly effective in achieving high accuracy while preserving the standard deviation of the response variables. While GLS performed slightly better in predicting the 100th percentile, MODL showed superior performance at the lower percentiles in most datasets. Interestingly, although MORF was generally the least accurate, it was the only method that consistently maintained the monotonicity of the predicted percentiles, a desirable property not inherently ensured by GLS or MODL, especially in the case of narrow diameter distributions. These findings underscore the strong potential of deep learning models for predicting diameter distribution percentiles and position MODL as a promising alternative to traditional parametric approaches.
Planetaarinen hyvinvointi : vinkkikortit
Kämäräinen, Helena; Lipponen, Maija; Vehmasto, Elina (Luonnonvarakeskus, 2024)
Kun luonto voi hyvin, myös ihminen voi hyvin. Planetaarinen hyvinvointi tarkoittaa elämäntapaa, jossa ihminen ottaa luonnon huomioon. Jos emme pidä huolta luonnosta, elämästä maapallolla tulee vaikeaa. Jo nyt meillä on monia ongelmia. Ilmastonmuutos etenee ja luonnon monimuotoisuus vähenee. Luonto ja ihmisten arki muuttuvat. Monet eläimet ja kasvit ovat vaarassa kadota. Jokainen voi kuitenkin tehdä luontotekoja, eli tekoja, jotka ovat hyviä luonnolle ja ihmiselle. Luontoteko auttaa ihmisen omaa hyvinvointia, lähiympäristöä ja koko maapalloa. Nämä kortit kertovat • mitä on planetaarinen hyvinvointi • miten voit vahvistaa suhdettasi luontoon • miten liikut ja syöt terveellisesti • miten voit suunnitella ja seurata omia luontotekojasi.
Modeling Trophic Dynamics in Lake Võrtsjärv: Energy Flow and Species Interactions
Tirronen, Maria; Kuparinen, Anna
Ecology and evolution : 7 (Wiley-Blackwell, 2025)
Understanding ecosystem dynamics is essential for ecological research and resource management. Bioenergetic or allometric trophic network models are effective in elucidating these interactions. However, aligning them accurately with empirical data remains challenging. Our present study contributes to such efforts by developing a trophic network model to describe population dynamics at Lake Võrtsjärv, Estonia, with a focus on predator–prey relationships and energy considerations. We calibrate this model to empirical biomass time series data using numerical optimization methods, a process previously applied to bionergetic models with considerably fewer guilds and/or parameters. Our approach emphasizes aligning the model closely with empirical time series and yields 77%–81% similarity between the modeled average dynamics and recorded biomasses. Despite relatively high similarity, the models we tested for noise—those assuming observation noise, as well as those incorporating environmental noise through stochastic differential equations—could not describe the annual variation of biomasses realistically. Overall, our tentative results demonstrate both the potential and the challenges involved in calibrating bioenergetic models to empirical data from large food webs.
Sensitivity of soil organic carbon stabilization indicators to 24 years of land-use change across soil depth
Kanari, Eva; Karhu, Kristiina; Salonen, Anna-Reetta; Lemola, Riitta; Soinne, Helena; Barré, Pierre; Baudin, François; Mizohata, Kenichiro; Oinonen, Markku; Gil, Jenie; Kohl, Lukas; Pennanen, Taina; Liang, Chao; Heinonsalo, Jussi
Geoderma (Elsevier, 2025)
Soil organic carbon (SOC) and its dynamics are sensitive to changes in land management while assessments of SOC dynamics rely on different indicators of SOC stabilization and are often restricted to topsoil. Here, we evaluated six indicators of SOC stabilization along a 70 cm soil profile under long-term land-use change. Using an agronomic experiment including an unmanaged meadow and two cropland treatments, we quantified isotopic signatures (14C and δ13C), size (mineral-associated OC; MaOC) and thermal (Rock-Eval® and PARTYSOC-derived centennially stable C; CS) fractions, and biochemical composition (amino sugar-derived microbial necromass C; MNC, and glomalin-related soil proteins; GRSP). Isotopic signatures and thermal analysis indicated older SOC (+∼5000 years), a decreasing influence of fresh C (+0.7 ‰ δ13C) and higher proportion of CS (by 75 %) with depth. In the cropland compared to the meadow, mean SOC age increased by ∼ 250 years, δ13C was enriched by 0.75 ‰ and CS was 27 % higher. The proportion of MaOC reflected a slight increase in SOC stabilization with depth (6 %) but decreased in the cropland compared to the meadow (−5%). The proportions of the two biochemical indicators to total OC decreased with depth (−67 % for MNC and − 78 % for GRSP), following the same trend as bulk SOC, while the proportion of MNC decreased (−15 %) and GRSP increased but with very high uncertainties (37 ± 20 %) in the cropland compared to the meadow. Our results suggest that different indicators likely represent SOC stabilization at different scales, and their validity should be assessed across soil layers.