Ecology and Evolution, Journal Year: 2024, Volume and Issue: 14(8)
Published: Aug. 1, 2024
Species' ranges are shifting rapidly with climate change, altering the composition of biological communities and interactions within among species. Hybridization is species that may change markedly yet it understudied relative to others. We used non-invasive genetic detections build a maximum entropy distribution model investigate factors delimit present future American marten (
Language: Английский
Ecological Informatics, Journal Year: 2025, Volume and Issue: unknown, P. 103006 - 103006
Published: Jan. 1, 2025
Language: Английский
Citations
1
Ecological Informatics, Journal Year: 2024, Volume and Issue: 81, P. 102604 - 102604
Published: April 21, 2024
Correcting sampling bias in species distribution models (SDMs) is challenging. The difficulty lies accurately identifying and quantifying the scarcity of samples, which greatly impedes implementation correction. Current methods often adjust presence or background points within geographic environmental spaces to correct probability estimation SDMs. However, these may lead information loss, rely on subjective assumptions, separate geography environment when correcting for bias. This study proposes a novel easily implementable method termed "aggregation background." selects data based aggregation degree feature space, thereby approximating representation correction samples. We compared this new with other prevalent existing literature by analyzing ecological authenticity. Under varying biases sample sizes, filtering achieved more accurate predictions target group methods. Notably, size was small (≤70), superior that obtained using method. These findings underscore effectiveness improving limited available data, without relying assumptions about Our provides approach complex unknown
Language: Английский
Citations
7
Journal of Biogeography, Journal Year: 2024, Volume and Issue: 51(9), P. 1783 - 1794
Published: April 25, 2024
Abstract Aim The continuous development of statistical tools applied to ecology has contributed great advances for modelling species' niches and distributions from opportunistic observations. However, as these observations are subject biases caused by spatial variation in sampling effort, ecological niche models (ENMs) also frequently biased. Among several bias correction methods that have been proposed, filtering—imposing a minimum distance between occurrences—is widely used, yet lacks clear guidelines choosing the filtering distance. Here, we aimed explore impact distances on performance ENMs. Location Europe. Taxon Virtual species. Methods We ENMs two virtual species with contrasting levels specialisation, across spectrum conditions, types sample sizes. Results Models specialist had average lower than those generalist Using biased reduced model performance, especially when was strong, size large. In many cases, failed improve or even it. did find an improvement modelled large strongly datasets. there no optimal distance, this linearly positively associated Moreover, because initial strong filtered dataset became very small, resulting only low accuracy. Main Conclusions Our results suggest is dealing ENMs, never improves enough draw accurate predictions. therefore recommend be employed cautiously, data available, bearing mind its effectiveness remains highly uncertain.
Language: Английский
Citations
6
Nature Communications, Journal Year: 2024, Volume and Issue: 15(1)
Published: Dec. 19, 2024
Machine learning-based geospatial applications offer unique opportunities for environmental monitoring due to domains and scales adaptability computational efficiency. However, the specificity of data introduces biases in straightforward implementations. We identify a streamlined pipeline enhance model accuracy, addressing issues like imbalanced data, spatial autocorrelation, prediction errors, nuances generalization uncertainty estimation. examine tools techniques overcoming these obstacles provide insights into future AI developments. A big picture field is completed from advances processing general, including demands industry-related solutions relevant outcomes applied sciences. In this scoping review, authors explore challenges implementing data-driven models—namely machine learning deep algorithms—in research.
Language: Английский
Citations
5
Journal of Ecology, Journal Year: 2025, Volume and Issue: unknown
Published: Feb. 13, 2025
Abstract Environmental conditions are dynamic, and plants respond to those dynamics on multiple time scales. Disequilibrium occurs when a response more slowly than the driving environmental changes. We review evidence regarding disequilibrium in plant distributions, including their responses paleoclimate changes, recent climate change new species introductions. There is strong that distributions often some with conditions. This poses challenge projecting future using distribution models (SDMs). Classically, SDMs assume set of occurrences an unbiased sample suitable However, environment may have higher‐than‐expected occurrence probabilities (e.g. due extinction debts) or lower‐than‐expected dispersal limitation) different areas. If unaccounted for, this will lead biased estimates suitability. methods for avoiding such biases SDMs, ranging from simple thinning dataset complex dynamic process‐based models. Such require large data inputs, natural history knowledge technical expertise, so implementing them can be challenging. Despite this, we advocate increased use, since provide best potential account model training then represent occupancy as ranges shift. Synthesis . Occurrence records climate. trained produce species' niche unless addressed modelling. A range tools, spanning wide gradient complexity realism, resolve bias.
Language: Английский
Citations
0
Journal of Zoology, Journal Year: 2024, Volume and Issue: 322(4), P. 291 - 308
Published: March 13, 2024
Abstract People have unique sets of direct sensory interactions with wild species, which change through their days, weeks, seasons, and lifetimes. Despite having important influences on health well‐being attitudes towards nature, these personalized ecologies remain surprisingly little studied are poorly understood. However, much can be inferred about by considering them from first principles (largely macroecological), alongside insights research into the design effectiveness biodiversity monitoring programmes, knowledge how animals respond to people, studies human biology demography. Here I review three major drivers, opportunity, capability motivation, shape people's ecologies. Second, then explore implications mechanisms for more passively actively practical improvements made in Particularly light declines richness that being experienced world (the so‐called ‘extinction experience’), significant consequences, marked improvement many experiences nature may key future biodiversity.
Language: Английский
Citations
2
Ecological Modelling, Journal Year: 2024, Volume and Issue: 493, P. 110749 - 110749
Published: May 11, 2024
Language: Английский
Citations
2
Methods in Ecology and Evolution, Journal Year: 2024, Volume and Issue: unknown
Published: Oct. 27, 2024
Abstract Recent advances in DNA barcoding have immeasurably advanced global biodiversity research the last two decades. However, inherent limitations barcode sequences, such as hybridization, introgression or incomplete lineage sorting can lead to misidentifications when relying solely on sequences. Here, we propose a new Niche‐model‐Based Species Identification (NBSI) method based idea that species distribution information is potential complement identifications. NBSI performs membership inference by incorporating niche modelling predictions and traditional Systematic tests across diverse scenarios show significant improvements identification success rates under newly proposed framework, where largest increase from 4.7% (95% CI: 3.51%–6.25%) 94.8% 93.19%–96.06%). Additionally, obvious were observed using potentially ambiguous sequences whose genetic nearest neighbours belongs another more than species, which occurs commonly with represented single short barcodes. These results support our assertion environmental factors/variables are valuable complements sequence data for avoiding inferred alone. The framework currently implemented R package, ‘NicheBarcoding’, open source GNU General Public Licence freely available https://CRAN.R‐project.org/package=NicheBarcoding .
Language: Английский
Citations
1
Ecology and Evolution, Journal Year: 2024, Volume and Issue: 14(8)
Published: Aug. 1, 2024
Species' ranges are shifting rapidly with climate change, altering the composition of biological communities and interactions within among species. Hybridization is species that may change markedly yet it understudied relative to others. We used non-invasive genetic detections build a maximum entropy distribution model investigate factors delimit present future American marten (
Language: Английский
Citations
0