Advances in forecasting harmful algal blooms using machine learning models: A case study with Planktothrix rubescens in Lake Geneva

Derot, Jonathan; Yajima, Hiroshi; Jacquet, Stéphan

doi:10.1016/j.hal.2020.101906

Cited by 37 publications

(16 citation statements)

References 82 publications

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Our study deals with the issue of harmful blooms, more particularly about the attempt to propose some scenarios (even imperfect) about the occurrence of harmful cyanobacterial blooms in lakes (e.g. Anneville et al 2015, Gallina et al 2017, Derot et al 2020. This work makes sense when one knows that future climate scenarios project an increase of such proliferations both in terms of frequency and duration (Paerl & Huisman 2009).…”

Section: Discussionmentioning

confidence: 99%

“…determinism, toxin risk, predictive models, management), molecular pathways (including toxin production), abiotic and biotic interactions with cyanobacteria, role of toxin production, as well as about the variety of applications (food supply, socio-economic models). Our study deals with the issue of harmful blooms, more particularly about the attempt to propose some scenarios (even imperfect) about the occurrence of harmful cyanobacterial blooms in lakes (e.g.Anneville et al 2015, Gallina et al 2017, Derot et al 2020. This work makes sense when one knows that future climate scenarios project an increase of such proliferations both in terms of frequency and duration(Paerl & Huisman 2009).…”

mentioning

confidence: 99%

See 1 more Smart Citation

Die hard in Lake Bourget! The case of Planktothrix rubescens

Rimet

Girel³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Blooms of Planktothrix rubescens have been recorded for 15 years in Lake Bourget (France), from 1995 to 2009. Then, the presence of this filamentous and toxic cyanobacterium became anecdotic between 2010 and 2015 and it was clearly thought that such a proliferation was over. However, against all odds, blooms occurred again in 2016 and 2017 despite apparent very low phosphorus concentrations in surface waters of the lake. Aims of this study were thus to explain the reasons of this come back in order to propose scenarios likely to be helpful to stakeholders who need to know if such proliferations may occur again in the future. We show that phosphorus input, both from the main tributaries to the lake and possibly from the sediments, were likely the triggers of the new development of the cyanobacterium since a minimum autumn/winter inoculum of P. rubescens was detected the year before. Then, the bloom, that was observed deeper than previous years, was associated to a conjunction of factors already well-known to favour the development of this very competitive species (i.e. mild winter temperature, water column stability, available light at depth, surface water transparency, low predation, etc). Although many factors and processes could account for the occurrence and bloom of the cyanobacterium, not observed, measured or taken into account here, a plausible scenario could be proposed and may be useful to deciders. One thing remains unclear: where do the cyanobacterium hides when it is not observed during the routine monitoring survey and thus from which place it could initiate its development (nearshore, in the pelagic zone, from the sediment?), unless it is simply not sampled and observed due to methodological bias.

show abstract

Section: Discussionmentioning

confidence: 99%

mentioning

confidence: 99%

Die hard in Lake Bourget! The case of Planktothrix rubescens

Rimet

Girel³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

show abstract

“…However, HAB prediction may fail to estimate shellfish contamination, as HAB species greatly vary on toxin production, and shellfish accumulation/elimination dynamics vary among species. The role of other drivers, including climate and environmental variables, has been investigated as predictors of biotoxin contamination by multivariate time-series modeling based on a range of variables measured over time (e.g., [12][13][14]). The following review covers past and current directions in time-series forecasting of algal bloom events and shellfish contamination by marine biotoxin.…”

Section: Forecasting Decision Making Time-series Data Machine Learningmentioning

confidence: 99%

“…ML methods, on the other hand, are known to be effective in handling complex, nonlinear and noisy datasets, especially when underlying physical relationships are not fully understood [16]. Moreover, they do not require the a priori assumptions used by the process-based models [12]. Additionally, ML models can be applied to virtually any data, eliminating the need to specify assumptions regarding the underlying statistical distribution of the data, as in traditional statistical models [17].…”

Section: Forecasting Decision Making Time-series Data Machine Learningmentioning

confidence: 99%

“…Moreover, the model allowed the authors to determine the most influential variables for the forecast: chemical and biochemical oxygen demand, pH, and total nitrogen/total phosphorus. More recently, Derot et al (2020) [12] proposed an RF model to forecast HABs caused by the toxic cyanobacterium Planktothrix rubescens in Lake Geneva. The goal was to lay the foundation for a HAB forecasting model to help local stakeholders managing the lake.…”

Section: Random Forestmentioning

confidence: 99%

See 1 more Smart Citation

A Review of Recent Machine Learning Advances for Forecasting Harmful Algal Blooms and Shellfish Contamination

Cruz

Costa

Vinga

et al. 2021

JMSE

View full text Add to dashboard Cite

Harmful algal blooms (HABs) are among the most severe ecological marine problems worldwide. Under favorable climate and oceanographic conditions, toxin-producing microalgae species may proliferate, reach increasingly high cell concentrations in seawater, accumulate in shellfish, and threaten the health of seafood consumers. There is an urgent need for the development of effective tools to help shellfish farmers to cope and anticipate HAB events and shellfish contamination, which frequently leads to significant negative economic impacts. Statistical and machine learning forecasting tools have been developed in an attempt to better inform the shellfish industry to limit damages, improve mitigation measures and reduce production losses. This study presents a synoptic review covering the trends in machine learning methods for predicting HABs and shellfish biotoxin contamination, with a particular focus on autoregressive models, support vector machines, random forest, probabilistic graphical models, and artificial neural networks (ANN). Most efforts have been attempted to forecast HABs based on models of increased complexity over the years, coupled with increased multi-source data availability, with ANN architectures in the forefront to model these events. The purpose of this review is to help defining machine learning-based strategies to support shellfish industry to manage their harvesting/production, and decision making by governmental agencies with environmental responsibilities.

show abstract