EUVIMEDEUVIMED entdecken
Open Access

Environmental constraints on freshwater red algae in Estonian streams: Insights from long‐term observations and machine‐learning models

Kai Piirsoo iD, Sirje Vilbaste iD, Peeter Pall, Kalle Olli iD

Journal of Phycology · 2026-08 · 62 (4) · 1169-1184

DOI: 10.1111/jpy.70188 ↗

Kurz gefasst

Fachartikel in der Zeitschrift Journal of Phycology, online erschienen am 24.06.2026. Verfasst von Kai Piirsoo und 3 weiteren Autor:innen. Beteiligte Einrichtung: Hydrobiology and Fisheries, Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, Tartu, Estonia. Hauptthemen laut MEDLINE: Machine Learning, Rivers, Rhodophyta. Die Arbeit wurde bisher 1-mal zitiert (Crossref) und nennt 83 Literaturangaben. Der Volltext ist frei in PubMed Central verfügbar. Gefördert von: Eesti Teadusagentuur, European Commission.

AbstractSchlagwörterBibliografische AngabenZitierenVerwandte Studien

Abstract (Originalsprache)

The division Rhodophyta (red algae) includes an inconspicuous and often overlooked component of benthic freshwater biota. In many cases, freshwater red algae are considered indicators of good ecological status, yet their environmental preferences remain poorly understood. Species richness is low globally, and the distribution of many taxa remains poorly known. Here, we report the occurrence of 12 taxa of benthic red algae in 272 Estonian streams, based on 734 sampling sites and 1263 sampling occasions. The aim of this study is to identify the key environmental drivers of freshwater red algae occurrence and to evaluate their potential as indicators of water quality. Red algae were recorded in 20% of the samples. The influence of 37 environmental variables on the occurrence of red algae was assessed using machine‐learning algorithms (random forest and gradient boosting trees). Hydrological and hydrochemical variables were by far the most important variable groups, followed by bottom substrate, watershed land cover and land use, and geological bedrock type. Among the hydrochemical variables, biological oxygen demand was the single most important predictor, indicating a high sensitivity of freshwater red algae to labile organic pollution. In contrast, the effect sizes of mineral and total nutrients were lower and more nuanced, suggesting that freshwater red algae are not particularly sensitive indicators of eutrophication. Some division‐level responses may have been obscured by contrasting species‐specific patterns. Species‐level effects were often difficult to verify using machine‐learning approaches due to low occurrence frequencies.

Schlagwörter

MeSH-Schlagwörter (MEDLINE)

Machine LearningRiversRhodophytaEstoniaFresh Water

Stichwörter der Autor:innen

Batrachospermaceaefreshwater red algaehabitat preferencesmachine learningriver ecosystemsspecies distribution

Autor:innen und Institutionen

  • Hydrobiology and Fisheries, Institute of Agricultural and Environmental Sciences, Estonian University of Life Sciences, Tartu, Estonia.
  • Institute of Ecology and Earth Sciences, University of Tartu, Tartu, Estonia.

ORCID-Profile: Kai Piirsoo, Sirje Vilbaste, Kalle Olli

Förderung

  • Eesti Teadusagentuur
  • European Commission

Zitieren

APA

Piirsoo, K., Vilbaste, S., Pall, P., & Olli, K. (2026). Environmental constraints on freshwater red algae in Estonian streams: Insights from long‐term observations and machine‐learning models. Journal of Phycology, 62(4), 1169-1184. https://doi.org/10.1111/jpy.70188

Vancouver

Piirsoo K, Vilbaste S, Pall P, Olli K. Environmental constraints on freshwater red algae in Estonian streams: Insights from long‐term observations and machine‐learning models. Journal of Phycology. 2026;62(4):1169-1184. doi:10.1111/jpy.70188

BibTeX
@article{Piirsoo2026,
  author = {Piirsoo, Kai and Vilbaste, Sirje and Pall, Peeter and Olli, Kalle},
  title = {Environmental constraints on freshwater red algae in Estonian streams: Insights from long‐term observations and machine‐learning models},
  journal = {Journal of Phycology},
  year = {2026},
  volume = {62},
  number = {4},
  pages = {1169--1184},
  doi = {10.1111/jpy.70188}
}
Kostenlos starten →