Data-driven modelling for assessing trophic status in marine ecosystems using machine learning approaches

Assessing eutrophication in coastal and transitional waters is of utmost importance, yet existing Trophic Status Index (TSI) models face challenges like multicollinearity, data redundancy, inappropriate aggregation methods, and complex classification schemes. To tackle these issues, we developed a novel tool that harnesses machine learning (ML) and artificial intelligence (AI), enhancing the reliability and accuracy of trophic status assessments. Our research introduces an improved data-driven methodology specifically tailored for transitional and coastal (TrC) waters, with a focus on Cork Harbour, Ireland, as a case study. Our innovative approach, named the Assessment Trophic Status Index (ATSI) model, comprises three main components: the selection of pertinent water quality indicators, the computation of ATSI scores, and the implementation of a new classification scheme. To optimize input data and minimize redundancy, we employed ML techniques, including advanced deep learning methods. Specifically, we developed a CHL prediction model utilizing ten algorithms, among which XGBoost demonstrated exceptional performance, showcasing minimal errors during both training (RMSE = 0.0, MSE = 0.0, MAE = 0.01) and testing (RMSE = 0.0, MSE = 0.0, MAE = 0.01) phases. Utilizing a novel linear rescaling interpolation function, we calculated ATSI scores and evaluated the model's sensitivity and efficiency across diverse application domains, employing metrics such as R2, the Nash-Sutcliffe efficiency (NSE), and the model efficiency factor (MEF). The results consistently revealed heightened sensitivity and efficiency across all application domains. Additionally, we introduced a brand new classification scheme for ranking the trophic status of transitional and coastal waters. To assess spatial sensitivity, we applied the ATSI model to four distinct waterbodies in Ireland, comparing trophic assessment outcomes with the Assessment of Trophic Status of Estuaries and Bays in Ireland (ATSEBI) System. Remarkably, significant disparities between the ATSI and ATSEBI System were evident in all domains, except for Mulroy Bay. Overall, our research significantly enhances the accuracy of trophic status assessments in marine ecosystems. The ATSI model, combined with cutting-edge ML techniques and our new classification scheme, represents a promising avenue for evaluating and monitoring trophic conditions in TrC waters. The study also demonstrated the effectiveness of ATSI in assessing trophic status across various waterbodies, including lakes, rivers, and more. These findings make substantial contributions to the field of marine ecosystem management and conservation.

Impact of extreme events on the transformation of hydrological characteristics of Asia's largest brackish water system, Chilika Lake

The earth is experiencing the impact of climate change due to global warming. Lake ecosystems are no exception and are expected to cope with the consequences of extreme climatic events (hereafter ECE), such as storms, floods, and droughts. These events have significant potential to alter the hydrological characteristics (HC) influencing the physical, chemical, and biological behavior of lake ecosystems. Considering such ecosystem's high-value services and benefits, it is the need of the hour to monitor and evaluate the impact of ECE on lake ecosystems. The second-largest brackish water system in the world, Chilika Lake, situated at the shore of the Bay of Bengal (BoB), has encountered a total of 1306 tropical cyclonic storms in the last 131 years. Since most tropical cyclones lead to heavy floods, this could be devastating for the ecosystem and its services. Hence, in order to bridge the knowledge gap, the present study was carried out to understand its impact, based on the available field data of more than two decades (1999 to 2020) and historical records of ECE and HC since 1840 and 1915 respectively from the literature. The study revealed that the ECE attributed to short-term changes in HC which were reflected through an immediate change in trophic state index (TSI, indicator of lake health) and trophic switchover (net autotrophic to heterotrophic) between net sink and source of carbon dioxide (CO₂) in specific regions. This study showed that both the ECE as well as a human intervention (opening of the new mouth) had an integrated role in the maintenance of HC within the lake as indicated by the variability of salinity level which is the lifeblood of the Chilika. Major ECE factors which controlled the salinity in Chilika were freshwater input through cyclone-induced flash flooding and seawater exchange through varying mouth conditions, i.e., opening of the new mouth, shifting, and widening of existing mouths due to cyclone impacts. The impact of the cyclone-induced flash flood was sustained for a couple of months to years depending on the magnitudes. As evidenced from the historical data available for ECEs, respective mouth variability, and salinity regime, ECE was found to maintain the salinity regime of the lake in the long run. Since the hydrological characteristics are found to be maintained through ECE as well as human intervention, the Chilika Lake recorded a substantial increase in fishery, seagrasses, Irrawaddy dolphins, migratory birds, and reduction in weed infestation. This study highlights the importance of historical data collection through a continuous systematic lake monitoring program which would enable understanding the ecosystem functioning and behavior with ECE-induced changing environmental conditions which is also a key component for formulating a sustainable management action plan for lake ecosystems around the globe.

Characterizing Trophic State in Tropical/Subtropical Reservoirs: Deviations among Indexes in the Lower Latitudes

Trophic state indexes (TSI) guide management strategies regarding eutrophication control worldwide. Such indexes usually consider chlorophyll-a (Chl-a), total phosphorus (TP), and Secchi disk depth (SDD) as independent variables for estimating aquatic productivity and the degree of impairment. TSIs for each of these components are frequently averaged to produce a single TSI value associated with a trophic state classification (e.g., oligotrophic, mesotrophic, or eutrophic). The potential divergence among equations and classification systems originally developed for temperate lakes or tropical/subtropical reservoirs might be particularly relevant in the tropics, where there is a lack of data and the use of equations originally developed for temperate systems may be inappropriate. We calculated two widely used TSIs for temperate lakes (TSI_temp) or tropical reservoirs (TSI_trop) and explored the deviations among TSI components in Brazilian reservoirs. When applied to our tropical/subtropical reservoirs, the TSI_temp provided a conservative approach, with lower limits anticipating increasing trophic state classification. TSI components for Chl-a and SDD significantly deviated for both sets of equations, and these discrepancies were related to turbidity, water temperature, and cyanobacterial biomass. For TSI_temp, but not for TSI_trop, TSI values in relation to Chl-a and TP were also significantly different. All such deviations have important management implications especially when Chl-a, TP, and SDD are averaged in a single TSI, representing loss of information and less useful trophic state classifications. Our results demonstrate that tropical water bodies may respond to drivers of eutrophication differently than temperate systems, highlighting the need for more data to better inform management of these understudied ecosystems. As managers collect data from more tropical water bodies, regional models may offer even better understanding of factors influencing trophic state.

Ecological health and water quality of village ponds in the subtropics limiting their use for water supply and groundwater recharge

Ponds are a typical feature of many villages in the subtropics, and have been widely used as important sources of water for agriculture, aquaculture and groundwater recharge, as well as enhancing village resilience to floods and drought. Currently many village ponds are in a very poor state and in dire need of rejuvenation. This paper assesses the current water quality status and ecological health of twelve sub-tropical village ponds, situated in western Uttar Pradesh, India. This assessment is used to evaluate their wastewater treatment needs in relation to potential village uses of the water. Physico-chemical (Secchi depth, Total phosphorus and Total nitrogen) and biological (Phytoplankton chlorophyll-a) indicators highlight hypertrophic conditions in all the ponds. The study indicates that the status of village ponds requires significant investments in wastewater treatment to restore their use for many purposes, including aquaculture, although some may still be acceptable for irrigation purposes, as long as pathogenic bacteria are not abundant. We propose increased implementation of decentralised systems for wastewater treatment, such as septic tanks and constructed wetlands, to reduce the organic and nutrient loads entering village ponds and allow their use for a wider range of purposes.

Trophic state modeling for shallow freshwater reservoir: a new approach

Lakes and reservoirs around the world are facing a substantial water quality degradation problem that poses significant environmental, social, and economic impacts. Reservoir productivity is influenced primarily by the climatic, morphometric, and hydro-edaphic features. High nutrient loadings in the reservoir from agriculture runoff often provide ideal conditions for algal blooms, leading to eutrophication. Reservoir and lake management to prevent or reduce eutrophication, therefore, has become the need of the hour. The traditional approach of trophic state monitoring by rigorous field surveys and eutrophication modeling has been revised in the present study by developing a new trophic state index (TSI)-based model for tropical shallow freshwater reservoirs. The new model has been constructed based on Carlson's Limnology and Oceanography, 22, 361-369, (1977) guidelines by establishing an empirical relationship between trophic parameters including total phosphorus (TP), Secchi disk depth (SDD), and chlorophyll (Chl-a). After comparing the new model with various earlier models for its applicability and validation with actual field conditions, it was found to be most precise over previous TSI models. Temporal and spatial fluctuations in the water quality of the Tiru reservoir were primarily attributed to the changing climatic conditions during the study period. Seasonal monsoon with less frequency, heavy nutrient loading from agriculture runoff, and increased turbidity due to a high level of sediment inflow during monsoon raised the TSI (SDD) values of the Tiru reservoir to place it in the hyper-eutrophic class. Average TSI values during winter for SDD, Chl-a, and TP were indicative of the meso-eutrophic to eutrophic state. Saturation of nutrients due to low water level during summer season caused the poly-eutrophic condition for TSI (SDD)- and TSI (TP)-based estimates and eutrophic condition as per TSI (Chl-a) estimates. However, seasonal deviations of the TSI values based on the relationship between TSI (Chl-a) and TSI (SD) indicated a predominance of smaller particles (non-algal turbidity) during all seasons. Even though TP present in the Tiru reservoir is controlling the algal production, it is also affected by low-light conditions due to non-algal turbidity. The recommendation from this study is that the TSI method for estimating the health of the water bodies is the efficient, cost-effective, and time-saving approach. The model developed during the study would help managers and policy makers to take necessary steps to reduce eutrophication levels in the reservoir and would be helpful for researchers in developing new concepts and protocols, mainly focusing on shallow freshwater reservoirs.

Seasonal dynamics of phytoplankton in response to environmental variables in contrasting coastal ecosystems

Seasonal distribution of phytoplankton community and size structure was assessed in three different tropical ecosystems of the western Bay of Bengal viz. estuary (Mahanadi), lagoon (Chilika), and coastal waters (off Gopalpur) in response to ambient hydrobiology. Salinity regimes differentiated the study regions as contrasting ecosystems irrespective of seasons (pre-monsoon, monsoon, post-monsoon). Taxonomic account revealed a total no of 175, 65, and 101 phytoplankton species in the estuary, lagoon, and coastal waters, respectively. Prevalence of marine, brackish, and fresh water types in the coastal waters, lagoon, and estuary, respectively, characterized the contrasting nature of the study regions in hosting the phytoplankton community. In general, phytoplankton abundance was observed in increasing order of coastal waters > estuary > lagoon during post-monsoon and pre-monsoon, while lagoon > coastal waters > estuary during monsoon. Bacillariophyta dominated the phytoplankton community in the estuary and coastal waters during all the seasons. In contrast, the lagoon exhibited a diverse array of phytoplankton group such as cyanophyta, dinophyta, and bacillariophyta during monsoon, post-monsoon, and pre-monsoon, respectively. Over the seasons, microphytoplankton emerged as the dominant phytoplankton size class in the coastal waters. Diversely, nanophytoplankton contributed to major fraction of chlorophyll-a concentration in the estuary and lagoon. Interestingly, pre-monsoon dinophyta bloom (causative species: Noctiluca scintillans with cell density 9 × 10⁴ cells·l^-1) and monsoon bacillariophyta bloom (causative species: Asterionellopsis glacialis 5.02 × 10⁴ cells·l^-1) resulted decline in species diversity. Multivariate statistical analysis deciphered salinity as a major environmental player in determining the distribution, diversity, and composition of phytoplankton communities in the three contrasting environments. Trophic state indices signified the lagoon and estuary as hypereutrophic during all season. The coastal water was marked as highly eutrophic through trophic state index during monsoon and pre-monsoon.

Phytoplankton biomass and composition in a well-flushed, sub-tropical estuary: The contrasting effects of hydrology, nutrient loads and allochthonous influences

The primary objective of this study was to examine trends in phytoplankton biomass and species composition under varying nutrient load and hydrologic regimes in the Guana Tolomato Matanzas estuary (GTM), a well-flushed sub-tropical estuary located on the northeast coast of Florida. The GTM contains both regions of significant human influence and pristine areas with only modest development, providing a test case for comparing and contrasting phytoplankton community dynamics under varying degrees of nutrient load. Water temperature, salinity, Secchi disk depth, nutrient concentrations and chlorophyll concentrations were determined on a monthly basis from 2002 to 2012 at three representative sampling sites in the GTM. In addition, microscopic analyses of phytoplankton assemblages were carried out monthly for a five year period from 2005 through 2009 at all three sites. Results of this study indicate that phytoplankton biomass and composition in the GTM are strongly influenced by hydrologic factors, such as water residence times and tidal exchanges of coastal waters, which in turn are affected by shifts in climatic conditions, most prominently rainfall levels. These influences are exemplified by the observation that the region of the GTM with the longest water residence times but lowest nutrient loads exhibited the highest phytoplankton peaks of autochthonous origin. The incursion of a coastal bloom of the toxic dinoflagellate Karenia brevis into the GTM in 2007 demonstrates the potential importance of allochthonous influences on the ecosystem.