SMSamu Mäntyniemi

Samu Mäntyniemi

*This is an algorithmically generated author record based on publicly available information.

Current Organizations / Institutes

The organizations/institutes from the scholar's latest article

Previous Organizations / Institutes

The organizations / institutes from the scholar's other articles in the past ten years


Samu Mäntyniemi has a total of 130 co-authors. Here are the top co-authors by number of shared publications:


Latest Publications

  • Open Access

    Genetic Risks and Population Fragmentation: The Threats to Finnish Wolves?

    by Laura Kvist, Johanna Sallmén, Inari Helle, Helena Johansson, Ilpo Kojola, Samu Mäntyniemi, Mia Valtonen, Jouni Aspi

    Understanding the dynamics of population size, survival, and genetic diversity is critical for the management of the Finnish wolf population, particularly in the context of inbreeding and population subdivision. We used Population Viability Analysis (PVA) simulations to assess the effects of inbreeding and population division on the viability of the Finnish wolf population over a 100-year period. Our models included scenarios with different numbers of lethal equivalents and varying migration rates between the Eastern and Western Finnish subpopulations, as well as between Finland and Russia. Results showed that population subdivision significantly decreased the overall population size, survival probability, and genetic diversity, with the eastern subpopulation particularly vulnerable. The addition of migration from Russia improved the survival of the Eastern subpopulation but did not affect the Western subpopulation significantly. Genetic diversity was highest in the three-population model (including Western Finland, Eastern Finland and Russia separately), although it still declined with an increase in inbreeding. The simulations highlighted that the current level of genetic diversity in the Eastern Finnish population is higher than in the Western population, but the Eastern population performed worse in terms of survival and genetic diversity, likely due to a smaller population size. Our findings underscore the importance of enhancing migration between the Finnish subpopulations and between Finland and Russia to maintain population viability and minimize the long-term detrimental effects of genetic drift and inbreeding.

  • Open Access

    Integrating Long-Term Ecological Research and Adaptive Management – Theoretical Solutions and Practical Challenges

    by Samu Mäntyniemi

    Effective natural resource management in the face of ecological complexity and uncertainty requires a seamless integration of long-term ecological research (LTER) and adaptive management. Bayesian decision analysis (BDA) provides a robust theoretical foundation for addressing this need, offering a structured framework for decision-making under uncertainty. At its core, BDA integrates three critical components: a causal model that represents the underlying dynamics of the ecological system, a set of alternative management decisions, and a valuation of future states of nature to assess the outcomes of these decisions. a causal model that represents the underlying dynamics of the ecological system, a set of alternative management decisions, and a valuation of future states of nature to assess the outcomes of these decisions. Central to BDA is the concept of the value of information (VOI), which quantifies the benefits of reducing uncertainty in the causal model through additional research or monitoring. VOI provides a mechanism to prioritize data collection efforts that are expected to yield the greatest improvements in decision-making. Similarly, the value of control (VOC) evaluates the potential benefits of influencing system dynamics through management actions. VOC highlights trade-offs between current decisions and future flexibility, aiding in the identification of strategies that maximize long-term ecological and socio-economic objectives. These concepts underscore the dynamic interplay between learning (via LTER) and acting (via adaptive management). The causal model serves as a formalized representation of ecological knowledge, encapsulating the relationships between key system variables, external drivers, and management interventions. By accounting for uncertainties within these relationships, the model facilitates predictions of system responses to various management actions. Alternative decisions, representing possible management pathways, are evaluated based on their capacity to achieve desired outcomes. Valuation of future states of nature, incorporating ecological, economic, and cultural dimensions, quantifies the relative benefits and trade-offs of different management scenarios. Despite its theoretical strengths, implementing Bayesian decision analysis in practice presents significant challenges. One major hurdle lies in defining an accurate and comprehensive causal model of the ecological system. The complexity of ecological interactions, combined with data limitations, often impedes the development of models that are both realistic and computationally tractable. Furthermore, the ability to value future states of nature and apply VOI and VOC concepts remains a formidable challenge. Valuation requires integrating diverse perspectives, accounting for uncertain and delayed outcomes, and navigating trade-offs between competing objectives. To address these constraints, Management Strategy Evaluation (MSE) has emerged as a valuable approximation of full Bayesian decision analysis. MSE uses simulation-based approaches to test alternative management strategies against a range of plausible scenarios, accounting for uncertainty in system dynamics and stakeholder objectives. By generating insights into the robustness and trade-offs of different strategies, MSE provides a pragmatic framework for decision-making when full BDA is infeasible. Developing a real-time BDA model for decision analysis closely parallels the concept of creating a Digital Twin of the ecological system. A Digital Twin is a dynamic, real-time representation of a physical system, continuously updated with data and capable of simulating system behavior under various scenarios. In ecological management, a real-time BDA model serves as a Digital Twin by integrating live data from monitoring systems, updating causal models to reflect current conditions, and providing immediate insights for decision-making. This real-time capability enhances the relevance and accuracy of adaptive management, particularly in rapidly changing environments. Integrating LTER with adaptive management enhances both MSE and BDA. Long-term data refine system understanding, reduce uncertainty, and identify thresholds, improving VOI and VOC assessments. Adaptive management facilitates iterative learning, allowing for real-time updates to causal models, valuation frameworks, and control strategies as new information becomes available. Realizing this potential requires collaboration across disciplines, stakeholder engagement, and investments in computational tools that support decision analysis and Digital Twin implementations. In this presentation I explore the interplay between theory and practice in integrating LTER and adaptive management, emphasizing how Bayesian decision analysis, VOI, VOC, MSE, and Digital Twins provide a pathway to informed, flexible, and resilient decision-making in ecological systems. It also highlights practical limitations and identifies key areas for advancing methodologies. Bridging the gap between theory and practice is imperative for achieving sustainable and adaptive outcomes.

  • Open Access

    A Bayesian multi-state model with data augmentation for estimating population size and effect of inbreeding on survival

    by Diego Rondon, Samu Mäntyniemi, Jouni Aspi, Laura Kvist, Mikko J. Sillanpää


  • Open Access

    Effect of the trap net emptying method on release mortality of Atlantic salmon estimated by a Bayesian system model

    by Timo J. Ruokonen, Henni Pulkkinen, Samu Mäntyniemi, Jaakko Erkinaro, Petri Suuronen


  • Open Access

    Assessment of the residential Finnish wolf population combines DNA captures, citizen observations and mortality data using a Bayesian state-space model

    by Samu Mäntyniemi, Inari Helle, Ilpo Kojola

    Assessment of the Finnish wolf population relies on multiple sources of information. This paper describes how Bayesian inference is used to pool the information contained in different data sets (point observations, non-invasive genetics, known mortalities) for the estimation of the number of territories occupied by family packs and pairs. The output of the assessment model is a joint probability distribution, which describes current knowledge about the number of wolves within each territory. The joint distribution can be used to derive probability distributions for the total number of wolves in all territories and for the pack status within each territory. Most of the data set comprises of both voluntary-provided point observations and DNA samples provided by volunteers and research personnel. The new method reduces the role of expert judgement in the assessment process, providing increased transparency and repeatability.


  • Open Access

    Assessment of the residential Finnish wolf population combines DNA captures, citizen observations and mortality data using a Bayesian state-space model

    by Samu Mäntyniemi, Inari Helle, Ilpo Kojola

    Assessment of the Finnish wolf population relies on multiple sources of information. This paper describes how Bayesian inference is used to pool the information contained in different data sets (point observations, non-invasive genetics, known mortalities) for the estimation of the number of territories occupied by family packs and pairs. The output of the assessment model is a joint probability distribution, which describes current knowledge about the number of wolves within each territory. The joint distribution can be used to derive probability distributions for the total number of wolves in all territories and for the pack status within each territory. Most of the data set comprises of both voluntary-provided point observations and DNA samples provided by volunteers and research personnel. The new method reduces the role of expert judgement in the assessment process, providing increased transparency and repeatability.

  • Bayesian arrival model for Atlantic salmon smolt counts powered by environmental covariates and expert knowledge

    by Henni Pulkkinen, Panu Orell, Jaakko Erkinaro, Samu Mäntyniemi

    Annual run size and timing of Atlantic salmon (Salmo salar) smolt migration was estimated using Bayesian model framework and data from 6 years of a video monitoring survey. The model has a modular structure. It separates subprocesses of departing, traveling, and observing, of which the first two together define the arrival distribution. The subprocesses utilize biological background and expert knowledge about the migratory behavior of smolts and about the probability to observe them from the video footage under varying environmental conditions. Daily mean temperature and discharge were used as environmental covariates. The model framework does not require assuming a simple distributional shape for the arrival dynamics and thus also allows for multimodal arrival distributions. Results indicate that 20%–43% of smolts passed the Utsjoki monitoring site unobserved during the years of study. Predictive studies were made to estimate daily run size in cases with missing counts either at the beginning or in the middle of the run, indicating good predictive performance.


  • Open Access

    Incorporating stakeholders' values into environmental decision support: A Bayesian Belief Network approach

    by Mirka Laurila-Pant, Samu Mäntyniemi, Riikka Venesjärvi, Annukka Lehikoinen