Aquatic Ecosystems: Biodiversity and Conservation

1. Introduction to the Special Issue

The structure and functioning of marine and inland water ecosystems are highly dependent on living organisms. The biodiversity of any particular habitat emerges from the complex relationships among the organisms themselves and their interaction with physical and chemical factors. Biodiversity reflects changes over long-term periods. Estimates on the actual number of living species on Earth are still controversial; however, it is agreed that much life and many ecological interactions are yet to be discovered. Furthermore, time is short and with the present worldwide political and economic volatility, conditions are very much unfavorable for preserving biodiversity. Currently, a quarter of the freshwater fauna is estimated to be at risk of extinction [1], while their aquatic habitats often experience a severe state of eutrophication, which is frequently dominated by pervasive cyanobacteria blooms [2,3]. Historically, freshwater ecosystems have been neglected by the official systems of protected areas [4,5], even in megadiverse continental countries [6]. In oceans and coastal waters, the widespread plastic contamination [7,8] and the extinction risk of certain taxonomic groups, such as sharks and rays [9,10,11], are also dramatic.

Biodiversity assessment is essential for different scientific approaches, as well as for conservation strategies and ecosystem management. This Special Issue collected papers that address the distinct spatial and temporal scales, as well as the past and present processes, that affect the conservation of the biodiversity of aquatic ecosystems.

2. Overview of the Articles Published in This Special Issue

Lim et al. tested conceptual models of the effect of salinity on macroinvertebrates in twelve estuaries in Southwestern Australia. They found that few species were associated with low salinities, which differed from conceptual models, while species richness was the greatest at mid-level salinities, declining at high salinities. Because hypersaline conditions are predicted to increase with global climate change, this research may help predict how benthic macroinvertebrate diversity in Mediterranean climates could decline in the future.

Spieles et al. examined the effect of a variety of land use-related parameters on macroinvertebrate diversity and biotic integrity in thirty low-order streams in Central Ohio. They found that the percentage of coverage of developed land at the watershed scale was the strongest predictor of both macroinvertebrate diversity and biotic integrity, while high-intensity development had a greater negative correlation with these measures than low-intensity development or agriculture. Taken together, these results indicate that impervious land cover at the scales of watershed and the Active River Area are important in affecting stream aquatic macroinvertebrate communities.

Demiray et al. used the transformer deep learning model (among other models) to determine its effectiveness in predicting the presence of harmful algal blooms (HABs) in Lake Erie. The target variable of chlorophyll a, as a surrogate for phytoplankton biomass, was used with a variety of physical, chemical, and biological water quality measurements taken from the western basin of Lake Erie as influencing parameters. Furthermore, using SHapley Additive exPlanations (SHAP) values as an explainable artificial intelligence (XAI) tool, it was determined that Particulate Organic Carbon (POC), Particulate Organic Nitrogen (PON), and Total Phosphorus (TP) were critical factors influencing HAB prediction in Lake Erie. This study demonstrates that predictive HAB models can be used successfully and may be applied to other systems in the future.

Simović et al. used a Convolutional Neural Network (CNN), which is a type of deep learning model, to identify aquatic insect species from three orders (Ephermeroptera (mayflies), Plecoptera (stone flies), and Trichoptera (caddisflies)) from an unbalanced data set. This data set contained between 10 and 80 individuals from each species. The authors found very little decrease in classification accuracy, as long as at least 30 individuals of the species were used. Strongly imbalanced data sets only decreased the classification accuracy by 2%, while moderately imbalanced data sets did not significantly affect classification accuracy. This is important to know because there are often wildly different numbers of individual taxa collected in aquatic macroinvertebrate biomonitoring.

Chung et al. examined riverine phytoplankton (also known as potamoplankton) in the Yeongsan River, following the construction of two weirs on this major river in South Korea. They examined Bacillariophyceae (diatoms), Chlorophyceae (green algae), Cyanophyceae (bluegreen algae), and other phytoplankton from 2019 to 2023, investigating how phytoplankton community structure and riverine physicochemical factors were related. They used several multivariate statistical techniques (e.g., canonical correspondence analysis (CCA) and principal components analysis (PCA)) to elucidate these relationships. PCA showed significant correlations with other phytoplankton and chlorophyll a (Chl-a) in spring, as well as with Cyanophyceae and water temperature in summer. CCA showed that the diatom Stephanodiscus sp. was associated with nitrogen-based nutrients, while the blue–green algae Microcystis sp. and Dolichospermum sp. were associated with water temperature and phosphate phosphorus (PO₄-P). This is important because Stephanodiscus can cause the clogging of waterworks and both Microcystis and Dolichospermum are toxin-producing genera.

Ergović et al. applied the Chironomid Pupal Exuvial Technique (CPET) to 28 man-made lakes in the Pannonian Lowlands and Dinaric Western Balkan Ecoregion of Croatia. This biomonitoring technique uses shed chironomid pupal skin (exuvia) to assess water quality. Of the 5698 chironomid pupal skins collected, the authors identified members of 141 taxa (including 97 species) belonging to five subfamilies. Physicochemical factors that influenced chironomid community assemblages differed from the Pannonian Lowlands (organic carbon (TOC) and orthophosphates (PO₄³⁻)) and the Dinaric Western Balkan (conductivity) ecoregions. The authors attribute these differing results to greater anthropogenic impacts in the Pannonian Lowlands. This study shows that in the future, the CPET can be applied to biomonitoring in lakes in Croatia and elsewhere.

Zhang et al. examined the fungal biodiversity and ecological function of fungal groups in the Tangchi Hot Springs in China. Illumina MiSeq high-throughput sequencing technology paired with bioinformatic techniques led the researchers to find that the dominant fungal phylum was Ascomycota, followed by Basidiomycota, Chytridiomycota, and Olpidiomycota, while the dominant fungal genera were Rhizophydium, Aureobasidium, Rhodotorula, and Sclerotinia. FUNGuild functional analysis showed that the dominant guilds in the hot springs were plant pathogens, followed by a variety of undefined sapotrophs. This research is important because most microbial investigations of hot springs have previously focused on bacteria, while this study adds to the knowledge regarding the fungal communities in these extreme ecosystems.

Li et al. examined the functional traits and niche characteristics of plant guilds in two ecosystems in the Xiangxi River basin in China. The reservoir water level fluctuation zones (RWLFZs) and the natural riparian zones (NRZs) were studied under different flooding regimes. The authors found 78 plant species in these zones; the dominant species were annuals and their percentage increased from 65.79% in the NRZs to 67.34% in the RWLFZs. Most of these annuals in the RWLFZs adopted the R adaptation strategy (sensu Grime). Furthermore, the Simpson dominance index of the RWLFZs was significantly higher than that of the NRZ. Finally, highly adaptable and widely distributed species with larger niche breadths and high-importance values usually had a greater niche overlap value in the RWLFZs than in the NRZs, which has implications both for competition after flooding events and the ecological restoration of these ecosystems.

Urbanski and Nogueira compared the effects of nutrient pollution on the fish fauna of Tietê River basin in São Paulo, Brazil, during dry and wet conditions. They measured physicochemical parameters and fish assemblages in three sites on the Tietê River and three sites on its tributaries. The Tietê River sites were classified from hypereutrophic to supereutrophic, while the tributaries were typically classified as mesotrophic during wet conditions and variable during dry conditions. Dissolved oxygen was less than 2 mg/L at all three sites in the Tietê River during wet conditions and >2 mg/L during dry conditions. With respect to fish, the richness per sample in the tributaries (11 to 14 spp.) was greater compared to that of the Tietê River proper (3 to 4 spp.). Siluriformes with accessory breathing dominated the Tietê proper, while the highly tolerant and detritivorous Prochilodus lineatus was the only species found at all sites. The authors conclude that eutrophication has created a chemical barrier to fish dispersal in this system, which needs to be addressed in order to restore the fish assemblage of the Tietê River.

Song et al. investigated the water quality and the macroinvertebrate community of the Ulungu River basin in northwest China between May and August 2022, as well as in October 2023. Along with sampling physicochemical parameters, the researchers collected 6101 macrobenthic organisms belonging to 3 phyla, 7 classes, 14 orders, 57 families, and 117 genera, with >85% of species being arthropods. Using these data, they calculated the biological monitoring working party (BMWP), family-level biotic index (FBI), and Shannon–Wiener Index. The mean values for the three indices (Shannon–Wiener, FBI, and BMWP) were greater for tributaries (light pollution, poor, and good, respectively) than the Ulungu River proper (moderate, poor, and general, respectively), varying according to sampling date.

Hu et al. examined the evolutionary mechanisms and morphological differences in geographically distinct populations of Gymnodiptychus dybowskii, which is a member of the carp family, in the Turks and Manas rivers in Xinjiang, China, between 2020 and 2021. The researchers performed morphometric analyses on 158 fish and found that the populations of the Turks and Manas rivers were different with one-way ANOVAs, showing 22 highly significant and 1 significant difference among the 35 morphological traits examined. The authors suggest that these differences are sufficient to make the fish in the two rivers different subspecies.

Hutorowicz used hydroacoustic sampling techniques in 2008 and 2021 to assess fish in Lake Dejguny (Poland). The target strength distribution was determined, along with temperature and oxygen concentrations, at 2 m depth intervals. A decline in the target strength in deep waters (>24 m) was associated with low oxygen levels (<2.5 mg/L). Finally, the researcher developed an index based on the number of large fish as a percentage of all acoustically recorded fish echoes; this was a modification of the Large Fish Index (LAFI), which compared the ratio of large fish to small fish.

Burandt et al. studied phytocenoses (plant communities) in riparian wetlands from 2017 to 2019 in the floodplains of northeastern Poland. They found that both the hydrological regime and geomorphological conditions controlled both the structure and pattern of the plant communities. Using linear discriminant analysis (LDA), four habitat types (wet, semi-wet, semi-dry, and dry zones) were identified. Indicator species analysis (ISA) showed that of the species with high water requirements, reed canary grass (Phalaris arundinacea) and common reed (Phragmites australis) dominated. Finally, flood pulses were found to be important in structuring these riparian wetland plant communities.

3. Conclusions

Researchers from eight different countries (Australia, Brazil, China, Croatia, Poland, Serbia, South Korea, and the United States of America) shared the results of their scientific investigations carried out on streams, rivers, lakes, artificial reservoirs, and estuary ecosystems. Ecological processes considered distinct spatial and temporal scales—local, local to regional, and regional geographic scales—as well as seasonal and long-term time series. The studied biota included phytoplankton/harmful algal blooms, plant guilds/riparian vegetation, and benthic macroinvertebrates, especially aquatic insects, fungal groups, and ichthyofauna. The main stressors affecting biotic integrity were global climate change, the intensification of land use, eutrophication, excessive pollution, and hydrological regime disturbance. Biotic communities’ responses were evaluated through distinct methodological strategies, such as deep learning models, bioinformatics, DNA genomics, and morphological taxonomy. The authors also revisited integrated data sets produced by water quality and biomonitoring programs, suggesting innovative approaches. This Special Issue provides a comprehensive view of the predictors of the biodiversity changes in freshwater and estuarine environments from five continents. The effective inclusion of aquatic ecosystems in the conservation and restoration agendas is crucial, while knowledge about their extraordinary biodiversity helps the healthy reconnection of people with the natural world.