A Machine Learning Approach to Simulation of Mallard Movements DOI Creative Commons

Daniel Einarson,

Fredrik Frisk, Kamilla Klonowska

et al.

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(3), P. 1280 - 1280

Published: Feb. 3, 2024

Machine learning (ML) is increasingly used in diverse fields, including animal behavior research. However, its application to ambiguous data requires careful consideration avoid uncritical interpretations. This paper extends prior research on ringed mallards where sensors revealed their movements southern Sweden, particularly areas with small lakes. The primary focus distinguish the movement patterns of wild and farmed mallards. While well-known statistical methods can capture such differences, ML also provides opportunities simulate behaviors outside core study span. Building this, this applies techniques these movements, using previously collected data. It crucial note that unrefined lead incomplete or misleading outcomes. Challenges include disparities swimming flying records, mallards’ biased due feeding points, extended intervals between points. highlights challenges, while identifying discernible patterns, as well proposing approaches meet challenges. key contribution lies separating incompatible and, through different models, handle separately enhance reliability simulation models. approach ensures a more credible nuanced understanding mallard demonstrating importance critical analysis applications wildlife studies.

Language: Английский

A super SDM (species distribution model) ‘in the cloud’ for better habitat-association inference with a ‘big data’ application of the Great Gray Owl for Alaska DOI Creative Commons
Falk Huettmann, Philip Andrews,

Moriz Steiner

et al.

Scientific Reports, Journal Year: 2024, Volume and Issue: 14(1)

Published: March 27, 2024

The currently available distribution and range maps for the Great Grey Owl (GGOW; Strix nebulosa) are ambiguous, contradictory, imprecise, outdated, often hand-drawn thus not quantified, based on data or scientific. In this study, we present a proof of concept with biological application technical workflow progress latest global open access 'Big Data' sharing, Open-source methods R geographic information systems (OGIS QGIS) assessed six recent multi-evidence citizen-science sightings GGOW. This proposed can be applied quantified inference any species-habitat model such as typically species models (SDMs). Using Random Forest-an ensemble-type Machine Learning following Leo Breiman's approach from predictions-we Super SDM GGOWs in Alaska running Oracle Cloud Infrastructure (OCI). These SDMs were best publicly (410 occurrences + 1% new assessment sightings) over 100 environmental GIS habitat predictors ('Big Data'). compiled associated overcome first time limitations traditionally used PC laptops. It breaks ground has real-world implications conservation land management GGOW, Alaska, other worldwide 'new' baseline. As research field remains dynamic, have limits, ultimate final statement associations yet, but they summarize all topic testable fashion allowing fine-tuning improvements needed. At minimum, allow low-cost rapid great leap forward to more ecological inclusive at-hand. GGOWs, here aim correct perception towards inclusive, holistic, scientifically urban-adapted owl Anthropocene, rather than mysterious wilderness-inhabiting (aka 'Phantom North'). Such was never created bird before opens perspectives impact policy sustainability.

Language: Английский

Citations

12
Correcting forest aboveground biomass biases by incorporating independent canopy height retrieval with conventional machine learning models using GEDI and ICESat-2 data DOI Creative Commons
Biao Zhang, Zhichao Wang, Tiantian Ma

et al.

Ecological Informatics, Journal Year: 2025, Volume and Issue: unknown, P. 103045 - 103045

Published: Jan. 1, 2025

Language: Английский

Citations

2
Progress on the world’s primate hotspots and coldspots: modeling ensemble super SDMs in cloud-computers based on digital citizen-science big data and 200+ predictors for more sustainable conservation planning DOI Creative Commons
Moriz Steiner, Falk Huettmann

Ecological Processes, Journal Year: 2025, Volume and Issue: 14(1)

Published: May 16, 2025

Abstract Background Describing where distribution hotspots and coldspots are located is crucial for any science-based species management governance. Thus, here we created the world’s first Super Species Distribution Models (SDMs) including all described primate best-available predictor set. These SDMs conducted using an ensemble of modern Machine Learning algorithms, Maxent, TreeNet, RandomForest, CART, CART Boosting Bagging, MARS with utilization cloud supercomputers (as add-on option more powerful models). For global cold/hotspot models, obtained data from www.GBIF.org (approx. 420,000 raw occurrence records) utilized largest Open Access environmental set 201 layers. this analysis, occurrences have been merged into one multi-species (400+ species) pixel-based analysis. Results We present quantified hotspot prediction Central Northern South America, West Africa, East Southeast Asia, Southern Africa. The Antarctica, Arctic, most temperate regions, Oceania past Wallace line. additionally these modeled hotspots/coldspots discussed reasons a understanding non-human primates occur (or not). Conclusions This shows us focus future research conservation efforts should be, state-of-the-art digital indication tools reasoning. Those areas be considered highest priority, ideally following ‘no killing zones’ sustainable land stewardship approaches if to chance survival.

Language: Английский

Citations

1
Machine learning applied to species occurrence and interactions: the missing link in biodiversity assessment and modelling of Antarctic plankton distribution DOI Creative Commons
Marco Grillo, Stefano Schiaparelli, Tiziana Durazzano

et al.

Ecological Processes, Journal Year: 2024, Volume and Issue: 13(1)

Published: July 25, 2024

Abstract Background Plankton is the essential ecological category that occupies lower levels of aquatic trophic networks, representing a good indicator environmental change. However, most studies deal with distribution single species or taxa and do not take into account complex biological interactions real world rule processes. Results This study focused on analyzing Antarctic marine phytoplankton, mesozooplankton, microzooplankton, examining their co-existences. Field data yielded 1053 interaction values, 762 coexistence 15 zero values. Six phytoplankton assemblages six copepod were selected based abundance roles. Using 23 descriptors, we modelled to accurately represent occurrences. Sampling was conducted during 2016–2017 Italian National Programme (PNRA) ‘P-ROSE’ project in East Ross Sea. Machine learning techniques applied occurrence generate 48 predictive maps (SDMs), producing 3D for entire Sea area. These models quantitatively predicted occurrences each assemblage, providing crucial insights potential variations biotic interactions, significant implications management conservation resources. The Receiver Operating Characteristic (ROC) results indicated highest model efficiency, Cyanophyta (74%) among Paralabidocera antarctica (83%) communities. SDMs revealed distinct spatial heterogeneity area, an average Relative Index Occurrence values 0.28 (min: 0; max: 0.65) 0.39 0.71) copepods. Conclusion this are science-based one world’s pristine ecosystems addressing climate-induced alterations interactions. Our emphasizes importance considering planktonic studies, employing open access machine measurable repeatable modelling, informed strategies face

Language: Английский

Citations

4
Ensemble-based forecasting of wildfire potentials using relative index in Gangwon Province, South Korea DOI Creative Commons
Sang Yeob Kim, Changhyun Jun, Wooyoung Na

et al.

Ecological Informatics, Journal Year: 2025, Volume and Issue: unknown, P. 103021 - 103021

Published: Jan. 1, 2025

Language: Английский

Citations

0
Using correlative science, open access big data and ensemble machine learning to track contamination signals in the wild: A first landscape-scale prediction for the Himalayan vulture (Gyps himalayensis) associated with diclofenac in Asia DOI Creative Commons
Dikpal Krishna Karmacharya, Ganesh Puri, Ganga Ram Regmi

et al.

Ecological Informatics, Journal Year: 2025, Volume and Issue: unknown, P. 103243 - 103243

Published: June 1, 2025

Language: Английский

Citations

0
Declining planetary health as a driver of camera-trap studies: Insights from the web of science database DOI Creative Commons
Thakur Dhakal, Tae-Su Kim,

Seong‐Hyeon Kim

et al.

Ecological Informatics, Journal Year: 2024, Volume and Issue: 83, P. 102801 - 102801

Published: Aug. 28, 2024

Planetary health is crucial to human well-being, ecosystem sustainability, and biodiversity preservation. In this context, camera traps are an effective remote sensing tool for monitoring biodiversity. Given the rising importance of understanding patterns trends, study examines possible factors influencing camera-trap studies provides bibliometric insights from 2377 publications indexed in Web Science (WoS). To explore potential drivers research growth, we used a logistic model based on specific variables, including global gross domestic product, temperature planetary measure declining living planet index, population growth. The index was identified as statistically significant driver growth (p-value <0.01), suggesting that curiosity regarding other beings influences studies. Through analysis, observed predominantly conducted United States, followed by England Australia, with notable upward trend over recent years. These align sustainable development goal 15 (Life Land) primarily classified under ecology category WoS. Further, have visualized network co-occurrence authors authors' affilation regions, keywords, keywords plus documents. Overall, assesses ecological conservation informatics reference scholars, policymakers, decision-makers.

Language: Английский

Citations

1
A Machine Learning Approach to Simulation of Mallard Movements DOI Creative Commons

Daniel Einarson,

Fredrik Frisk, Kamilla Klonowska

et al.

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(3), P. 1280 - 1280

Published: Feb. 3, 2024

Machine learning (ML) is increasingly used in diverse fields, including animal behavior research. However, its application to ambiguous data requires careful consideration avoid uncritical interpretations. This paper extends prior research on ringed mallards where sensors revealed their movements southern Sweden, particularly areas with small lakes. The primary focus distinguish the movement patterns of wild and farmed mallards. While well-known statistical methods can capture such differences, ML also provides opportunities simulate behaviors outside core study span. Building this, this applies techniques these movements, using previously collected data. It crucial note that unrefined lead incomplete or misleading outcomes. Challenges include disparities swimming flying records, mallards’ biased due feeding points, extended intervals between points. highlights challenges, while identifying discernible patterns, as well proposing approaches meet challenges. key contribution lies separating incompatible and, through different models, handle separately enhance reliability simulation models. approach ensures a more credible nuanced understanding mallard demonstrating importance critical analysis applications wildlife studies.

Language: Английский

Citations

0