Challenges and Opportunities in Using Fish Metrics for Reservoir Water Quality Evaluation

Abstract

The Water Framework Directive (WFD) was designed to protect the quality of all water resources. For reservoirs, the ecological potential classification assesses biological parameters, evaluating only the phytoplankton community. Thus, this study aimed to evaluate the effectiveness of using fish communities to determine water quality in reservoirs. A literature review was conducted to gather information on how fish community data were integrated into reservoir water quality assessment under the WFD. This work includes an exploratory case study of the Aguieira Reservoir (Portugal), evaluating the ichthyofauna community, along with physical, chemical, and biological assessment of the water. The results of the review show that fish abundance and composition (sensitive metrics) should be used to develop ecological indices for assessing water quality in reservoirs. However, the effects of anthropogenic pressures and invasive species are not included in the calculation of most proposed indices. The case study serves as an illustrative example and demonstrates low abundance and composition of the fish community with a high percentage of invasive species, revealing a poor water quality, regarding ichthyofauna biotic index results (F-IBIP). Nevertheless, including these metrics in the classification of ecological potential can help guide restoration strategies to mitigate the effects of anthropogenic pressures.

1. Introduction

The needs of human societies in response to population growth, combined with technological development, have led to systematic human intervention in natural ecosystems, aiming to improve living standards [1]. Numerous cross-river structures, such as weirs, dams, and small hydropower plants, have been built to meet human water needs for consumption, agricultural activities, and electricity production, as well as flood control and the protection and conservation of riverbanks and riverbeds [1]. However, the EU was faced with insufficient water protection legislation, which only included a few parameters for certain types of water bodies (e.g., standards for drinking water, bathing water, and fish and shellfish waters, including a list of harmful substances and limits for their discharge). This approach reduced environmental systems to individual parameters without an integrative way of assessing the real environmental quality state and aquatic health [2]. This resulted in the adoption, by several European countries, of the Water Framework Directive (WFD) (Directive 2000/60/EC, 2000), which provides a comprehensive framework for the protection and sustainable management of surface and groundwater. The WFD coordinates EU member states’ initiatives to attain the defined objectives and includes key metrics such as ecological assessment, integrated planning, and pollution elimination strategies, among others [3].

The WFD defines “Heavily Modified Water Bodies” as those altered physically by human activities and presents classification criteria for them, which include, in the case of reservoirs, parameters such as residence time, drainage area, and operational regime. For these types of waterbodies, the concept of “Ecological Potential” is used and refers to the difference between the current quality of the aquatic ecosystem and its maximum achievable quality after implementing mitigation measures and without significant adverse effects on human uses or the environment. The assessment of ecological or potential status considers biological, chemical, physical, and hydromorphological quality elements, requiring harmonized sampling and analysis procedures across member states to ensure comparable results. The overall status of surface water bodies is determined by the worst score among these factors. While the physical and chemical status is established by EU norms, each member state defines the ecological status/potential through the definition of local reference values [3].

Given that fish fauna is widely recognized as a key biological indicator of water quality, essential for assessing the ecological status of rivers, its importance is well established within the scientific community [3]. This is further underscored by its routine monitoring in several countries, including Germany, China [4], South Africa [5], and the United States [6]. Although the WFD only defines ichthyofauna as a biological evaluation element for water quality in European rivers, lakes, and transitional waters [7], the ecological roles and impacts of fish communities could also extend to reservoirs, since they can also influence the structure and functioning of these aquatic ecosystems. In addition, member states may choose not to include fish in the ecological status assessment of heavily modified and artificial water bodies (HMWBs and AWBs), depending on national typology and monitoring programs. In Portugal, for heavily modified water bodies such as reservoirs, the official WFD classification is currently based on phytoplankton as the biological quality element, as defined in national normative documents. The underuse of fish metrics in reservoirs is described as being related to a lack of standardized indices, caused by difficulties in collecting quality data (given sampling limitations in lentic habitats), defining the important variables to accurately explain biotic, abiotic, and anthropogenic factors, and determining the relative importance of abiotic and anthropogenic factors in explaining biotic factors [8].

The presence and distribution of piscivorous exotic species highlight broader ecological implications, since they are widely distributed across the major river basins, preferring lentic ecosystems [9]. Their introduction was primarily intentional, driven by commercial interests (e.g., sport fishing and gastronomy) and public health efforts (e.g., using predatory species to control insect larvae and fight malaria) but with significant impacts on the autochthonous ichthyofauna community. Additionally, other introduction pathways include natural movement through transboundary watercourses; improper release of species used for experimental or aquarium purposes; alteration of natural habitats, creating favorable conditions for these species; or even climate change through the modification of water temperature and change in hydrological patterns [10]. The exotic species (e.g., Cyprinus carpio—common carp, Micropterus salmoides—largemouth bass, Lepomis gibbosus—pumpkinseed, and Carassius auratus—goldfish) are often opportunistic, highly adaptable to habitat changes, and possess great dietary flexibility, preying on native species, especially their eggs and larvae [10]. Due to their greater adaptability, exotic species tend to dominate and outcompete native species, leading to a reduction in biodiversity, particularly in freshwater reservoirs. Reservoirs act as environmental stressors and create unpredictable dynamics for native fauna, which is often displaced by exotic species. For example, common carp (Cyprinus carpio), although not a predator, disrupts the structure and functioning of fish communities in lentic ecosystems, especially by reducing aquatic plant diversity [11]. Common carp tends to destroy vegetation and increase water turbidity, deteriorating the ecological quality of the ecosystem for native species (e.g., Barbus barbus—barbel, Pseudochondrostoma polylepis—Iberian nase, and Alosa sapidissima—shad) that require clean water and vegetation to survive. The largemouth bass (Micropterus salmoides) is a predatory carnivorous species that primarily consumes fish eggs, adults, amphibians, and crustaceans, posing a threat to indigenous species [12]. The goldfish (Carassius auratus) introduces fish diseases, feeds on native fish, their eggs, and larvae, reduces aquatic vegetation biomass, and resuspends sediments [13]. This resuspension increases nutrient concentration, which can lead to significant microalgae growth and consequent eutrophication processes.

In addition to invasive species, other factors have affected fish populations in aquatic ecosystems. The damming of rivers has caused decline or local extinction of migratory species, such as salmon (Salmo salar) or sturgeons (Acipenser spp.), the latter being one of the most vulnerable and threatened fish families [14]. This has led to the development of various fish passage systems (generically referred to as fishways) [1]. Hydraulic structures limit the movement of migratory and resident species upstream or downstream, reducing access to crucial areas for their life cycle [1]. Moreover, dam construction causes significant changes in the physical and chemical quality of the water body due to water retention, leading to alterations of the natural sediment transport (sediment accumulation upstream of the dam), and potential discharges from the reservoir during maintenance or cleaning. Additionally, temperature changes also occur in the river section affected by the dam, with the nature and extent of these variations depending on climatic conditions, river characteristics, and the type of structure [1].

Regarding fish communities, Portugal created a fish-based index of biotic integrity (F-IBIP) [7] designed to assess the biological quality of Portuguese rivers [7]. The F-IBIP, like other biotic integrity indices, consists of several metrics that attempt to reflect the basic structural and functional characteristics of fish communities in lotic ecosystems [7]. These metrics can increase or decrease depending on the intensity of anthropogenic disturbance and are grouped into two main categories: species richness and composition (e.g., number of native species and percentage of exotic individuals) and ecological factors (e.g., feeding or reproduction). Thus, to apply the F-IBIP, each sampling site must be classified according to its fish community grouping, from the six defined groups (salmonid from the northern region, salmonid–cyprinid transition from the northern region, medium-sized cyprinid from the northern region, small-sized cyprinid from the northern interior and southern regions, medium-sized cyprinid from the southern region, and cyprinid from the northern coastal region) [7]. First, fish species in a watercourse were characterized by their status (native or exotic) and, according to the ecological/functional guild concept, by the following ecological aspects: tolerance to degradation, feeding habits, reproductive habits, habitat use, and migratory behavior [7]. Based on this characterization, a broad set of metrics was defined for each grouping. The F-IBIP value was then obtained by calculating the arithmetic mean of the metrics in each fish community grouping, with individual metric values ranging, on a continuous scale, from 0 (zero) representing poor quality to 1 (one) representing excellent quality. At the final classification, the ecosystem quality was expressed in one of five quality classes (excellent, good, moderate, bad, and poor), with equal class ranges across all groupings [7].

Thus, in line with the points raised and considering the current issue of the scarcity of indices for fish communities, especially in reservoirs, this study has three objectives: (i) to conduct a review of the scientific literature on the use of fish communities as bioindicators in lentic ecosystems (reservoirs) under the WFD approach; (ii) to review the literature regarding the types of metrics and indices used to measure water quality in reservoirs, using fish communities as a biological element; and (iii) aligned with these findings, to provide an illustrative example of a case study in which the F-IBIP (an ichthyofauna index for rivers) was applied to test its replicability to reservoirs, in this case, the Aguieira Reservoir (center of Portugal). The case study is not designed to be a comprehensive ecological assessment but rather a practical scenario of previously found limitations.

2. Materials and Methods

2.1. Literature Review

To address the objectives (i) and (ii), an extensive literature review was conducted, following the guidelines suggested by Pullin and Stewart (2006) [15]. Various databases were consulted to ensure broad research coverage, including b-on, Web of Science, Scopus, and Google Scholar, focused on the European scientific literature. This methodology was employed to gather information on fish communities as bioindicators of water quality under the WFD approach and metrics used in water quality assessment and to understand the impact of fish exotic species (invasive or non-invasive) on water quality measurement. Considering the proposed objectives, several keywords were combined into different sets to obtain the widest possible search result (Table 1).

Table 1. Keywords combinations used for the literature search on the various databases and the number of articles found for each word group.

Keywords Sets	Nᵒ Articles
Lentic; AND Fish; AND Ecological Status; AND Water Quality	19
Fish; AND Reservoir; AND Ecological Status; AND Water Quality	50
Fish; AND Dam; AND Ecological Status; AND Water Quality	37
Fish; AND Ecological Status; AND Water Quality; AND Lentic; AND Reservoirs	10
Fish; AND Water Quality; OR Lentic; AND Reservoirs	408
WFD; AND Fish; OR Ecological Status; AND Water Quality	221
WFD; AND Fish; AND Exotic Species; OR Ecological Status	14
WFD; AND Fish; OR Ecological Status; AND Lentic	30
Fish; AND WFD; OR Ecological Status; AND Reservoir	93
Total Number of Articles	882

The search sequence was adapted for each database to handle different operators (e.g., using “AND”, “OR”, and excluding non-European countries). Articles resulting from each search across databases were evaluated for their relevance and focus on the research topics through three-level selection criteria applied to the article’s title, abstract, and full text. According to Pullin and Stewart (2006) [15], for an article to advance to the next selection level, it had to meet a set of pre-established criteria (Table 2). The inclusion criteria were deliberately defined to focus on European lentic freshwater ecosystems and studies published in English and Spanish to ensure methodological consistency and relevance to the Water Framework Directive context, thereby providing a coherent and regionally applicable synthesis within the scope of reservoir water quality assessment using fish communities. During data extraction, relevant information on characteristics and topics within the central research theme from selected articles was organized into tables using a questionnaire (Table 3). Given the thematic scope and the current scarcity of standardized, comparable data on fish-based indices in reservoir assessments under the WFD, a qualitative analysis was the most suitable approach to identify existing knowledge gaps, highlight methodological variability, and inform future directions for research and ecological monitoring.

Table 2. Inclusion/exclusion criteria.

Criteria	Inclusion	Exclusion
Year	1990 to 2024	<1990
Geographic location	Europe	Everywhere else
Language	English, Spanish	Every other language
Ecosystem	Lentic freshwater	Saltwater, lotic freshwater
Fish communities	Predominant	Not present
Water quality	Water quality assessed	Water quality not assessed

Table 3. The questionnaire elaborated to cover different topics regarding the central theme of the review.

Q.1	Which articles address the objectives proposed by the Water Framework Directive for lentic systems?
Q.2	What indices and metrics were used in the studies of the selected articles?
Q.3	What types of study areas were defined and examined in the selected articles?
Q.4	What type of methodology was used in the selected articles?
Q.5	Which articles address the issue of exotic species concerning water quality?

2.2. Case Study: Aguieira Reservoir

The Aguieira Reservoir was chosen as a case study since it is the only Portuguese reservoir included in the WFD inter-calibration exercise for defining elements to be assessed in these water bodies [16]. This choice allows the applicability of a riverine fish-based index (F-IBIP) to a lentic, heavily modified system to be tested, in line with the literature indicating the need for validating and adapting indices across water body types [17,18,19]. The Aguieira Reservoir is located on the Mondego River at the border between the municipalities of Penacova and Mortágua in the district of Coimbra, with an approximately 3100 km² watershed draining area. The primary uses of the reservoir include water supply (for populations, industry, and agriculture) and hydroelectric power generation, while secondary uses include fishing, swimming, and recreational rowing and sailing [16]. These diverse uses imply multiple anthropogenic pressures, including hydromorphological alterations, eutrophication risks, and biological invasions, making it an appropriate site for testing fish-based bioassessment approaches.

The fish community in the Aguieira Reservoir was sampled during a single standardized campaign in the spring of 2019 at three sites (Ag1, Ag3, and Ag4; Figure 1) chosen to represent different sections of the reservoir (littoral vs. pelagic areas, upstream vs. downstream), following the recommendations of the European Committee for Standardization for sampling fish in lakes and reservoirs [17,18]. Sampling was conducted using multi-mesh, facilitating comparisons with other studies [17,18]. The gillnets are composed of 12 different mesh sizes ranging from 5 mm to 55 mm knot to knot following a geometric series, ensuring capture of both small and large-bodied species across functional guilds. The nets were deployed from a boat under the fish fauna sampling and analysis protocol established by INAG [20], as well as the European standard for fish sampling in lakes and reservoirs [21]. The nets were set in the late afternoon, placed perpendicular to the shore to optimize representativeness and maximize fish capture, and retrieved the following morning after approximately 12–14 h of soaking time. This combined diurnal–nocturnal approach captures diel activity of several species, including nocturnal feeders, such as pikeperch (Sander lucioperca), as recommended by Mueller et al. (2017) [22]. All captured species were placed in buckets with water for identification and some biometric measurements (weight and length). Individuals were also assigned to ecological/functional guilds (trophic, reproductive, habitat use, and pollution tolerance), following the framework used for the F-IBIP [7,23]. Dead specimens were recorded, and care was taken to return live fish to the water as quickly as possible to minimize handling stress.

Figure 1. A representative map of Aguieira Reservoir with the sampling site’s location.

The Fish-based Index of Biotic Integrity for Portugal (F-IBIP) [7,23] was applied to the dataset using the official web application [7]. This index combines metrics related to species richness, native vs. exotic species composition, and functional guilds, returning an overall score between 0 (poor quality) and 1 (high quality). It should be noted that this case study was designed as an exploratory application of an F-IBIP to a lentic, heavily modified reservoir rather than as a comprehensive ecological survey. Although originally designed for lotic systems, the index was applied here as a proxy to explore its sensitivity and potential for adaptation to lentic systems, consistent with approaches used in other European studies (e.g., Blabolil et al. (2016) [17], Launois et al. (2011) [19]).

To contextualize the ecological quality results, a set of water physical and chemical parameters was determined according to the criteria for classifying the status of surface water bodies in reservoirs established by INAG in 2009 [24]. In situ measurements, including dissolved oxygen (O₂: mg/L and %) and pH using a multiparameter probe (Multi 3630 IDS SET F), were conducted. A water sample was also collected for later laboratory determination of phosphorus (mg P/L) and nitrate (mg NO₃/L) [24]. Additionally, the phytoplankton community was also evaluated regarding the concentration of chlorophyll a (μg/L), total biovolume (mm³/L), cyanobacteria biovolume (mm³/L), and the algae general index. These data provide insight into trophic status, nutrient enrichment, and other stressors that influence fish communities, enabling cross-comparison with literature-reported relationships between water quality and fish-based metrics [17,25,26]. This multi-parametric approach ensures that the case study is not only descriptive but also analytically comparable with similar studies across Europe, enhancing its relevance for testing the integration of fish-based metrics into WFD reservoir assessments.

3. Results

3.1. Fish Communities as Bioindicators in Lentic Ecosystems—Literature Review

A total of 558 articles (after removing duplicates) between 1990 and 2024 that met the initially defined criteria were obtained in the bibliographic research process. Title screening led to the exclusion of 382 articles (68% of the initial articles, Table 4). Most exclusions were due to deviations from the central topic, covering topics such as genetic characteristics, macroinvertebrates, lotic ecosystems, etc. According to exclusion criteria, based on title information, a total of 176 articles proceeded to abstract screening (second analysis level).

Table 4. Number and percentage of excluded articles through the different levels of analysis.

No of Articles Analyzed in Each Phase	558	176	44
	Title	Abstract	Complete Text
No of the excluded articles	382	132	32
Excluded articles (%)	68	75	73

Abstract analysis resulted in the exclusion of 132 articles (75% of those moved to this level, Table 4). These articles were excluded as they did not focus on fish communities in European reservoirs or water quality assessment. Another group of articles was excluded because the studies were not conducted in Europe or were written in languages other than English or Spanish. After this exclusion step, 44 articles remained for full-text review.

At the final analysis level, 32 articles were excluded (73% of the 44 articles under review at this level, Table 4) due to insufficient or inadequate focus on topics related to the study objectives. Moreover, this exclusion was conducted regarding the inadequate approach to fish communities, water quality assessments based on assumptions or experiments (biomanipulation) rather than using indices based on fish communities, and pollutant interference with water quality, which, in turn, would affect fish communities.

After the selection process, 12 scientific articles were fully analyzed, using a pre-prepared questionnaire designed to cover various topics within the central research theme (see Table 3).

3.1.1. Q.1. Which Articles Address the Objectives Proposed by the Water Framework Directive for Lentic Systems?

Regarding the 12 selected scientific articles, 11 were aligned with the WFD and with the proposed objectives [17,18,19,22,25,26,27,28,29,30,31]. Santos et al. (2017) [32] present an independent study that does not rely on the WFD’s objectives. Instead of creating an index for water quality classification using fish communities, the authors addressed the impacts on fish communities after dam construction. This study presents an integrated assessment of two dam impacts on aquatic fauna in the Sabor River Reservoir (Portugal). According to the authors, the changes in water quality caused by dam construction were significant for both reservoirs, especially the downstream reservoir. Santos et al. (2017) [32] also showed that the consequences of deteriorated water quality for aquatic fauna were severe, marked by sharp declines in native fish species and the invasion of exotic species.

Table 5 summarizes how each article addresses the objectives proposed by the WFD (Q.1), including the second question about indices and metrics used in the studies (Q.2) and the types of study areas that were examined (Q.3).

Table 5. Table summary of articles description focused on Q.1—Which articles address the objectives proposed by the Water Framework Directive for lentic systems?; Q.2—What indices and metrics were used in the studies of the selected articles?; and Q.3—What types of study areas were defined and examined in the selected articles?

Ref.	Q.1. WFD alignment	Q.2. Indices/Metrics Used	Q.3. Types of study areas
Irz et al. (2006) [28]	Compare fish communities to determine if lakes and rivers can serve as a reference for reservoirs. Similar species between lakes and reservoirs (e.g., Esox lucius, Perca fluviatilis, Rutilus rutilus, and Tinca tinca)—subset of those in rivers. Community comparison suggests an intermediate position of reservoirs between lake and river systems. Consider artificial ecosystems in monitoring and assessment programs.	% occurrence rates.	21 lakes, 50 reservoirs, and 549 rivers in France
Navarro et al. (2009) [25]	Classified reservoirs by geological and morphometric characteristics. Quality comparisons between different reservoir typologies and identified intrinsic differences—use of Ecological Quality Ratio (EQR). Most reservoirs met the WFD quality criteria, but reservoirs with low EQR values highlighted issues, such as spills, droughts, or untreated water.	Chlorophyll a concentration, cyanobacterial chlorophyll concentration, % of fish anomalies, CPUE (catch per unit effort) of carp in reservoir littoral zones, CPUE of carp in the limnetic zone, % of carp in littoral and limnetic zones, Secchi disk depth, % hypolimnetic oxygen, and total phosphorus. Danish EQR reference system, selecting the metrics highlighted above.	Reservoirs in northeastern Spain
Launois et al. (2011) [19]	Few fish-based biotic integrity indices (FBIs) have been developed only for lakes. The challenge was to develop an FBI tool for French lakes. Fish surveys, natural lake parameters, anthropogenic pressures at the catchment scale, and local pressures were assessed.	Fish categorized by trophic, reproductive, and tolerance associations (richness and evenness were added). These metrics were modeled against pressures and natural environmental variables. Final indices: Natural lakes included intolerant species, intolerant species biomass, and % of intolerant species biomass; artificial lakes were % of lithophilic species biomass; omnivorous species biomass; herbivorous species count; % piscivorous species; strictly lithophilic species %; and biomass per unit effort (BPUE).	24 natural lakes and 67 reservoirs in France
Argillier et al. (2012) [26]	Use of the European Committee for Standardization method for sampling fish. Lakes characterized by environmental factors and eutrophication proxies. Following a site-specific approach, with validation, a fish index (EU index) was developed.	Out of 90 potential metrics, only 3 metrics (total CPUE, total BPUE, and CPUE of omnivorous fish species) were selected to be incorporated into the EU index for reservoir ecological quality assessment.	445 lakes in twelve European countries
Paulovis et al. (2012) [29]	Spatial analysis to assess ecological status. The Kis-Balaton Water Protection System (KBWPS) was designed to function as a natural retention zone to protect Lake Balaton’s water quality. Use of different fish metrics to assess the ecological status of the KBWPS and Lake Balaton. Preliminary approach to investigate which fish attributes and metrics are more sensitive for use in future studies for biotic index calculations.	Total number of fish species and exotic fish species, % of exotic fish, % biomass of exotics, % of tolerant species, % biomass of tolerant species, total number of intolerant species, omnivore proportion, % omnivore biomass, invertivore proportion, % invertivore biomass, piscivore proportion, % piscivore biomass, herbivore proportion, % herbivore biomass percentage, planktivorous proportion, % planktivorous biomass, and total fish count.	Three geographically connected sites: two reservoirs of the KBWPS and the western basin of Lake Balaton—Hungary
Blabolil et al. (2016) [17]	New fish-community-based methodology to assess water quality in Central and Western European reservoirs, which is potentially transferable to other European ecoregions. After collecting fish data from the reservoirs, a final index was composed, accurately identifying reservoirs that were lightly, moderately, and heavily affected by eutrophication.	New fish index—CWE-FBI with three fish metrics—abundance (total CPUE and total BPUE), composition (Shannon diversity index, with fish classified by reproductive associations, families, trophic groups, feeding habitat, and tolerance; family/guild metrics expressed as proportions of total CPUE and BPUE), and age (indirectly evaluated through the average fish weight). The final metrics used for CWE-FBI were total BPUE, piscivore CPUE, and planktivorous CPUE.	Reservoirs in France and the Czech Republic
Blabolil et al. (2017) [18]	Use of fish communities and a dataset covering various pressures and environmental gradients. Fish data correlated as a proxy for eutrophication and as indicators for developing the ecological quality index. Several metrics were used to create the final index based on fish communities (CZ-FBI, Czech Fish-Based Index), meeting the criteria required by the WFD. Finally, the index was compared with a general fish index previously developed for Central and Western Europe, and, as it encompasses additional pressures (e.g., acidification, pollution by oil products and drug metabolites, lake recreational uses, etc.), it was helpful for a comprehensive analysis of the ecosystem’s health.	Ptotal as a proxy for assessing eutrophication (relationship evaluation between each metric). Metrics: total fish biomass, total fish abundance, relative biomass of Abramis brama, relative abundance of Gymnocephalus cernua, relative biomass of Perca fluviatilis, relative biomass of Scardinius erythrophthalmus, relative biomass of Salmonidae, and presence of young fish from six species (Abramis brama, Rutilus rutilus, Alburnus alburnus, Gymnocephalus cernua, Perca fluviatilis, and Sander lucioperca) to calculate the CZ-FBI index with RQE correction	Reservoirs in Czech Republic
Mueller et al. (2017) [22]	Efficiency and representativeness of various fishing methods at different times of the day. Recommended sampling at different times of day. Daylight sampling may overlook important temporal variations in fish behavior and community structure, although nighttime sampling poses increased costs, safety risks, and potential legal barriers in some countries. To address limitations and standardize sampling efficiency, the authors advocate integration of night sampling, not only in scientific research but also within the legal frameworks of the WFD and other governmental fish monitoring programs.	CPUE, species richness, evenness, and Shannon index were compared among fish population sampling methods, backwaters, transects, and times of day.	Three backwaters (lentic systems) of the Inn River in Germany
Bobori et al. (2018) [27]	The role of benthic macroinvertebrates and fish communities in assessing the ecological quality of an artificial water body over two years (2013–2015). The anthropogenic pressures on the watershed and the benefits of water quality improvement are discussed in the context of WFD implementation, concerning the introduction of sustainable management plans.	Monitoring physico-chemical water parameters, benthic macroinvertebrates, and fish communities. Various indices (the Greek Lake Benthic Invertebrate Index—GLBiL, the Greek Lake Fish Index—GLFI, and the Hellenic Evaluation System 2—HESY2) were calculated to assess and classify water quality. Only the GLFI was used: the numerical contribution of invasive species and the contribution of omnivorous species to total biomass.	Karla and Kalamaki reservoirs in Greece
Česonienė et al. (2020) [31]	Examined the ecological status, focusing on how morphometric factors (lake size, depth, and water exchange rate) influence water quality and ecological health. Larger and deeper water bodies have better conditions for water purification, slower sediment accumulation, and overall healthier ecological status. Overall, depth, area, and water exchange rates were significant in determining the ecological status, with larger, deeper, and well-circulating lakes showing healthier ecological metrics.	The Lithuanian Fish Index (LFI) measures ichthyofauna diversity and community structure, the presence of species sensitive to pollution, and changes in trophic conditions. Higher depths were linked to lower Ptotal, Ntotal, and specific macrophyte values, indicating better ecological quality. Larger areas correlated with lower Ptotal and Ntotal levels and higher ecological status indicators in fish, macroinvertebrate, and macrophyte communities. Rapid water exchange improved conditions (reduced Ptotal, Ntotal, and chlorophyll), but slower exchange benefited macrophyte health.	29 lakes and 10 ponds in Lithuania
Pieckiel et al. (2024) [30]	Pilot ichthyological index to evaluate the ecological potential of reservoirs in Central and Eastern Europe under the guidelines of the WFD. Data from various fishing methods across reservoirs with diverse hydrological and morphological features were collected, aiming to correlate fish species presence with eutrophication levels. Identify typical fish communities, biodiversity levels, suitable fishing gear, and pressure-sensitive species. Results provide information for ecological potential classification in reservoirs, focusing on undisturbed fish communities and appropriate reference states. Key findings include a strong correlation between pressure indicators and the Cyprinidae/Percidae ratio.	Carlson’s Trophic State Index (TSI) pressure indicator by quantifying eutrophication through Ptotal, Secchi depth, and chlorophyll. Pearson correlation coefficient correlates the fish composition (CPUE as abundance (NPUE), and biomass (BPUE)) and the TSI index. A Cyprinidae/Percidae ratio (C/P Index) was developed to compare the relative abundance of Cyprinidae (often eutrophication-tolerant species) to Percidae (sensitive to eutrophic conditions). A strong negative correlation between Percidae abundance and TSI, reinforcing the effectiveness of this metric in indicating ecological degradation.	11 reservoirs with varying hydromorphological parameters in Poland

3.1.2. Q.4. What Type of Methodology Was Used in the Selected Articles?

The methodologies outlined in the selected articles provide a basis for determining the ecological potential across different study areas (Table 6). Almost all studies describe gillnets and/or electrofishing as the methodology for fish species collection (Table 6). In addition to “fish sampling,” Bobori et al. (2018) [27] also used “macroinvertebrate sampling” to assess the ecological quality of water in the Karla reservoir. Other articles showed that physical and chemical parameters were also used to assist in evaluating ecological water quality [18,22,25,27,32]. Mueller et al. (2017) [22] showed the effectiveness of various fish species collection methodologies (e.g., baited fish traps, multi-mesh gillnets, fyke nets, seine nets, lift nets with bait, line fishing, longline fishing, snorkeling, and electrofishing) (Table 6). Electric fishing proved to be the most effective method, revealing higher species richness, species features representation, and CPUE, followed by seine nets. For specific species like dace (Leuciscus leuciscus), Eurasian ruffe (Gymnocephalus cernua), common bream (Abramis brama), and silver bream (Blicca bjoerkna), seine netting was more effective. Some species were captured more consistently at night, dusk, or dawn than during the day. Electric fishing along 30 m transects at different times of the day can be recommended for monitoring shallow standing water fish communities, enabling maximum comparability with adjacent river habitats. Seine netting is an alternative if habitat accessibility is restricted or electric fishing is prohibited.

Table 6. Parameters and methodologies used for fish and macroinvertebrate collection in articles that evaluated physical and chemical water analyses in their studies.

Parameters and/or Methodologies		References
Physical and Chemical Parameters
Temperature (°C)		Argillier et al., 2012 [26]; Mueller et al., 2017 [22]; Santos et al., 2017 [32]; Bobori et al., 2018 [27]; Pieckiel et al., 2024 [30]
pH		Mueller et al., 2017 [22]; Santos et al., 2017 [32]; Bobori et al., 2018 [27]
Conductivity (μS/cm)
Dissolved Oxygen (mg/L)		Santos et al., 2017 [32]; Bobori et al., 2018 [27]; Pieckiel et al., 2024 [30]
Biochemical Oxygen Demand over 5 days (mg/L)		Santos et al., 2017 [32]; Bobori et al., 2018 [27]
Chemical Oxygen Demand (mg/L)		Bobori et al., 2018 [27]
Secchi (cm)		Navarro et al., 2009 [25]; Argillier et al., 2012 [26]; Mueller et al., 2017 [22]; Santos et al. 2017 [32]; Bobori et al., 2018 [27]; Pieckiel et al., 2024 [30]
Turbidity (NTU)		Mueller et al. 2017 [22]; Santos et al., 2017 [32]; Bobori et al., 2018 [27]
Total Phosphorus (mg/L)		Navarro et al. 2009 [25]; Argillier et al., 2012 [26]; Blabolil et al., 2016 [17]; Santos et al. 2017 [32]; Bobori et al., 2018 [27]; Česonienė et al., 2020 [31]
Total Dissolved Phosphorus (mg/L)		Blabolil et al.,2017 [32]; Bobori et al., 2018 [27]
Total Nitrogen (mg/L)		Bobori et al., 2018 [27]; Česonienė et al., 2020 [31]
NO3-N (mg/L)		Santos et al., 2017 [32]; Bobori et al., 2018 [27]
NO2-N (mg/L)		Bobori et al., 2018 [27]
NH4-N (mg/L)
[Chlorophyll a] (mg/m3)		Navarro et al., 2009 [25]; Santos et al., 2017 [32]; Bobori et al., 2018 [27]; Česonienė et al., 2020 [31]
Cyanophyte Chlorophyll (mg/m3)		Navarro et al., 2009 [25]; Bobori et al., 2018 [27]
Macroinvertebrates
Ekman–Birge Dredge		Bobori et al., 2018 [27]
D-shaped Net (250 mm × 230 mm)
Fish Communities
Gillnet	Benthonic	Irz et al., 2006 [28]; Navarro et al., 2009 [25]; Launois et al., 2011 [19]; Argillier et al., 2012 [26]; Blabolil et al., 2016 [17]; 2017 [18]; Mueller et al., 2017 [22]; Santos et al., 2017 [32]; Bobori et al., 2018 [27]; Česonienė et al., 2020 [31]; Pieckiel et al., 2024 [30]
	Pelagic	Irz et al., 2006 [28]; Navarro et al., 2009 [25]; Launois et al., 2011 [19]; Blabolil et al., 2017 [18]; Mueller et al., 2017 [22]; Santos et al., 2017 [32]; Pieckiel et al., 2024 [30]
Electric Fishing		Irz et al., 2006 [28]; Navarro et al., 2009 [25]; Paulovis et al., 2012 [29]; Mueller et al., 2017 [22]; Santos et al., 2017 [32]; Bobori et al., 2018 [27]
Baited Traps		Mueller et al., 2017 [22]
Fyke Nets
Seine Nets
Lift Nets with Bait
Line Fishing
Longline Fishing
Snorkel

3.1.3. Q.5. Which Articles Address the Issue of Exotic Species Concerning Water Quality?

Only two of the twelve articles address exotic species as a factor affecting water quality evaluation. Although the final sample size for the review approach is limited to 12 articles, this reflects the scarcity of research specifically addressing the use of exotic fish communities in reservoir water quality assessment under the WFD, thereby reinforcing the relevance of this review in identifying a critical gap and highlighting the need for further development in this underexplored field. Bobori et al. (2018) [27] examined exotic species that are easily adaptable to eutrophic conditions, which are associated with poor water quality. According to the authors, the introduced species are exclusively planktivorous and invertivores, meaning that the trophic cascade effect [33] is insufficient to control eutrophic conditions [27]. Santos et al. (2017) [32] also address exotic species, stating that the dam construction (Sabor River) created two reservoirs with an increase in water temperature and conductivity in upstream watercourses and downstream lakes, especially in the secondary reservoir. Another observed consequence was the accumulation of nitrogen and phosphorus in the two reservoirs formed, originating from sediments through the decomposition of organic matter or from agriculture practiced in the area surrounding the reservoir [32]. The abundance of these nutrients triggered the growth of aquatic plants and microalgae, leading to eutrophic conditions and indicating poor water quality. With the water quality deterioration, a significant change in the composition of native aquatic fauna species occurred [32]. According to the authors, before the dam’s construction, fish communities were mainly represented by native species like the barbel (Luciobarbus bocagei) and the boga (Pseudochondrostoma duriense). However, after the dam’s construction and habitat fragmentation, there was a decline in native fish communities and a proportional increase in exotic species (e.g., bleak—Alburnus alburnus and pumpkinseed—Lepomis gibbosus) [32]. The lentic ecosystem formed upstream created suitable environmental conditions for exotic species dispersion, likely amplified by deteriorating water quality [32].

3.2. Case Study—Aguieira Reservoir

The general physical and chemical support parameters (Table 7) were measured at the Aguieira Reservoir in a single standardized campaign (spring of 2019). Overall, a poor water quality classification was obtained, namely, due to high nutrient concentrations and recurrent cyanobacterial blooms (Table 7), in accordance with previous studies [34,35,36,37,38,39,40].

Table 7. Results of general physical, chemical, biological parameters, and specific pollutants concentrations that allowed us to determine the ecological potential according to WFD metrics and threshold values for Portugal’s northern reservoirs from 3 sampling sites in the Aguieira Reservoir. Reference and boundary values are defined by APA (2016) [41]. Bold values stand for values outside the established threshold.

		Sampling Sites
Physical and Chemical parameters	Reference Values	Ag1	Ag3	Ag4
pH	6–9	9.20	8.31	9.15
O2 (mg/L)	≥5 mg/L	11.9	11.3	12.2
O2 (%)	60–120 %	119.4	112.1	125.2
Nitrate (mg NO3/L)	≤25 mg NO3/L	2.8	4.0	1.2
Phosphorus (mg P/L)	≤0.05 mg P/L	0.01	0.09	0.02
Ecological Potential—Physical and Chemical Parameters		Good	Moderate	Good
Biological Parameters—Phytoplankton
	Boundary Value Good/Moderate
[Chl a] (μg/L)	9.50	26.32	10.16	27.90
Total Biovolume (mm3/L)	1.90	29.33	22.17	39.64
Cyanobacteria Biovolume (mm3/L)	9.20	1.41	5.37	11.37
Algae General Index	10.60	58.95	22.91	38.43
Ecological Potential—Biological Parameters		Moderate	Moderate	Poor
Ecological Potential		Moderate	Moderate	Poor

Regarding the sampled ichthyofauna community, 16 fish specimens were captured in 2018/2019, of which 62.5% were invasive species, and only 1 of the captured species was endemic (37.5%) (Table 8).

Table 8. List of caught ichthyofauna community at Aguieira Reservoir, biometric parameters, and inclusion in functional guilds.

Scientific Name	Common Name	Weight (kg)	Size (cm)	Trophic Habitat	Trophic Guilds	Pollution Tolerance	Biogeography
Alosa alosa	Shad	0.980	47.5	Pelagic	Planktivore	Moderately sensitive	Native
Carassius auratus	Goldfish	0.160	20.6	Benthonic	Omnivore	Tolerant	Exotic
		0.150	20.6
		0.185	22.0
		0.175	20.5
		0.250	22.0
Cyprinus carpio	Carp	0.375	28.8
Lepomis gibbosus	Pumpkinseed	0.050	10.4	Pelagic	Insectivore
Luciobarbus bocagei	Barbel	1.495	50.0	Benthonic	Omnivore		Endemic
		0.495	35.7
		0.470	32.5
		0.780	37.7
		0.525	34.2
Sander lucioperca	Pikeperch	0.390	38.0	Pelagic	Piscivore		Exotic
		0.180	28.0
		0.085	13.9

Only one adult shad specimen was captured, which was already dead in the net.

The F-IBIP index was calculated for the Aguieira Reservoir using the information from Table 8 and the web application [23], resulting in a classification of “poor” water quality.

4. Discussion

4.1. Fish Communities as Bioindicators in Lentic Ecosystems

One of the key limitations in developing robust fish-based indices for reservoirs is the predominance of tolerant species with low ecological specialization. These fish are less responsive to environmental pressures, reducing the sensitivity of ecological metrics. In Mediterranean systems, where environmental variability is naturally high, many species are already adapted to harsh conditions, which makes it challenging to detect anthropogenic impacts using traditional biotic indices [17]. This tolerance often results in a reduced number of reliable metrics, as seen in Blabolil et al. (2016) [17]. While it is true that many lentic species are generalists, trait-based approaches can still compose meaningful metrics by focusing on functional redundancy and guild shifts rather than species identity alone [42]. Additionally, the assumed tolerance of lentic fish to stressors may mask subtle sub-lethal effects that are only visible with long-term monitoring or physiological biomarkers (e.g., growth, reproduction, stress proteins, and swimming behavior) [43].

Several studies have demonstrated that fish biomass and certain species’ CPUE are reliable indicators of eutrophication. For example, total biomass is considered a robust proxy for food web productivity [44,45]. Metrics such as the CPUE of planktivores and piscivores are frequently used to signal increased productivity or ecological degradation. However, in Blabolil et al. (2016) [17], the inclusion of tolerant or recently introduced species like common bream (Abramis brama) or pikeperch (Sander lucioperca) may weaken these signals due to differences in behavior and habitat use [17]. A metric employed by Blabolil et al. (2016) [17] in the index was the CPUE in gillnets of invertivores/piscivorous fish, with perch (Perca fluviatilis)/pikeperch (Sander lucioperca) being the most significant relation. Pikeperch is typical of nutrient-rich lowland waters, and its population grows with increasing eutrophication. The rise in the abundance of perch individuals was also recognized as a factor in increasing ecosystem productivity. However, this may be a meaningless figure due to the difference in size and habitat use of the species. Moreover, the system may not be in a steady state given the recent introduction. In particular, Navarro et al. (2009) [25] showed that species richness and diversity/evenness indices were not strongly related to water quality, while the CPUE of invasive species like common carp (Cyprinus carpio) in littoral (electrofishing) and limnetic zones (gillnet) strongly correlated with phosphorus levels and trophic indicators like chlorophyll concentrations. However, the carp is an invasive species widely introduced worldwide, with significant effects on increased water turbidity (nutrient resuspension and macrophyte decline). Argillier et al. (2012) [26] further supported the use of CPUE and omnivorous species biomass as reliable eutrophication metrics. Degraded conditions benefit opportunistic species like omnivorous feeders due to dietary plasticity, contrasting with specialized feeders [46]. However, the authors noted that tolerance-based metrics did not correlate well with anthropogenic pressure at large spatial scales, suggesting challenges in defining tolerance thresholds across Europe [8,26,44,47,48,49,50,51,52]. It can be said that while biomass correlates with productivity, it can be non-specific, since it also increases with invasive species or unbalanced food webs. Moreover, CPUE data is highly sensitive to the type of sampling gear selected, with gillnets often underrepresenting smaller-bodied or pelagic species.

Blabolil et al. (2017) [18] state that the CZ-FBI clearly defines fish metrics, allowing rapid classification of ecological potential and, when necessary, providing guidance for restoration and fisheries management actions. The authors note that the methodological steps used to develop the new index are transferable to other regions with similar reservoirs and environments that share the same species. The CZ-FBI is based on well-defined indicators, correlating with eutrophication, measured as total phosphorus concentration, as well as in the assessment of fish abundance, community composition, age structure, biomass measures, and species or family-level fish composition. According to the authors, the selected species are widely distributed and constitute a significant part of fish communities across Europe. Additionally, the CZ-FBI reflects the inherent spatial heterogeneity of reservoirs in terms of fish distribution, covering longitudinal and depth gradients in benthic and pelagic habitats. The authors also compared the CWE-FBI with the CZ-FBI, noting that the latter primarily uses single-species indicators, whereas the CWE-FBI relies on functional guilds. It is worth mentioning that, in terms of WFD criteria, the CWE-FBI does not cover age structure. Fish biomass metrics are used in both the CWE-FBI and the CZ-FBI, with the CWE-FBI metric being an integral value for the entire benthic habitat, while pelagic biomass metrics were calculated separately in the reservoir and tributary sections. This approach’s advantage is accounting for the longitudinal productivity gradient present in lentic ecosystems [18]. Additionally, CZ-FBI development was based on paired correlations with P_total concentration, while CWE-FBI uses hindcast procedures. This advanced hindcast model can reveal causal relationships in large datasets but is generally not applicable to smaller datasets like those used for CZ-FBI. Instead, simpler methods can identify clear patterns in the dataset. Expert assessment was also involved in CZ-FBI development. This subjective approach is appropriate when reference conditions do not exist, and only a limited dataset is available. Despite the different underlying approaches, both indices yield very similar results, confirming either index’s ability to classify reservoirs’ ecological potential across an anthropogenic pressure gradient. Similarly, Launois et al. (2011) [19] demonstrated that the hindcast approach to fish community-based metrics made it possible to assess lakes’ and reservoirs’ current conditions. A set of metrics, showing significant responses to anthropogenic pressure, was selected and combined into a biotic integrity index for French lakes and reservoirs. In reservoirs, all trophic guilds appeared to relate to anthropogenic pressures, strongly suggesting that a fish community-based index is a relevant eutrophication indicator [48,53,54]. According to the authors [19], while the FBI (fish-based biotic integrity index) for reservoirs fully meets the set of criteria recommended by Karr et al. (1986) [55], the FBI developed for natural lakes in France lacks a metric responding to agricultural pressures, possibly due to low agricultural land use in the sampled lake watersheds. Using the hindcast model, the authors assumed that reference conditions were adequately simulated, setting all anthropogenic factors to zero. However, under the WFD, the reference for a given lake corresponds to the “absence of anthropogenic pressures or near-pristine conditions” [56]. Therefore, the method was even more restrictive in reference assessment than anticipated by the WFD, leading to overestimated FBI values. Although this hindcast method is increasingly popular [57], users should remember that it includes some subjectivity. However, reproducibility and index precision can be enhanced by adding new fish data. Overall, hindcast approaches are insightful but heavily rely on historical data quality and assumptions about optimal reference conditions, which may not exist for many reservoirs, and indices based on functional guilds (e.g., the CWE-FBI) are often more transferable, while species-based indices (e.g., the CZ-FBI) may be more ecologically precise but regionally restricted [58].

Regarding biomass-based metrics, Paulovis et al. (2012) [29] questioned whether they might better reflect ecosystem function than abundance-based ones. Their analysis suggested that while the two types of metrics sometimes diverged (e.g., for omnivores and piscivores), no clear pattern or correlation emerged. Biomass is often skewed by a few large individuals, while abundance may better capture population dynamics, especially of rare or sensitive species. Still, biomass may offer more sensitive insights into community structure and energy flow, especially when assessing the role of dominant species or functional groups. However, combining both (e.g., BPUE + CPUE) provides a fuller picture of ecosystem productivity and structure, which is recommended in recent ecological assessment guidelines [59].

The presence and dominance of non-native or opportunistic species complicate the use of fish metrics in reservoirs, and their presence can skew index values and undermine attempts to assess native biodiversity health. For those situations, Pieckiel et al. (2024) [30] propose and test a fish-based index, tailored for assessing the ecological potential of reservoirs in Poland, representing a significant step forward in using fish community metrics to assess water quality under the WFD. The core innovation in this study is the development of a pilot ichthyological indicator called the Cyprinidae/Percidae Index (C/P Index), which reflects the impact of eutrophication on fish communities. This index centers on the relative abundance of two fish families, Cyprinidae and Percidae, as indicators of nutrient load and water quality. The C/P Index draws on the established sensitivity [26,60,61] of Percidae species (like Lepomis gibbosus—pumpkinseed and Sander lucioperca—pikeperch) to low nutrient levels, in contrast to the tolerance of Cyprinidae species (such as Rutilus rutilus—roach and Cyprinus carpio—common carp) to eutrophic conditions. Essentially, the ratio serves as a proxy for water quality. A higher C/P Index suggests a shift toward nutrient-rich, eutrophic conditions. This pilot index is promising, not only for Poland but for broader European applications, offering a scalable, transferable tool for evaluating ecological potential in various reservoir settings. The use of species-based indicators addresses the need for consistency across regions, since the selected fish families are widely distributed across Europe. The correlation of the C/P Index with TSI values further underscores the practical advantage of combining biological and chemical metrics to assess human impacts in modified water bodies.

Hydrological and morphological features such as waterbody size, depth, and retention time significantly influence fish community structure and water quality, and many fish indices fail to correct for these natural gradients, leading to misclassification of reservoirs with inherently low diversity. Česonienė et al. (2020) [31] explore the connection between the hydrometric parameters and the ecological status of Lithuanian lakes and ponds. The study emphasizes that larger and deeper water bodies often have better water quality. The key findings reveal that lakes with greater depths generally have lower levels of total phosphorus and nitrogen, which are indicators of higher ecological status, and that fish community indicators, such as the Lithuanian Fish Index (LFI), also correlate with better water quality, suggesting that fish communities can effectively assess water quality. Additionally, the study demonstrates that faster water exchange rates in lakes can positively impact the reduction in nitrogen and phosphorus levels, leading to healthier ecological conditions. Consequently, these findings indicate that fish community characteristics, lake depth, and water exchange rates are reliable hydromorphologic metrics for assessing and enhancing the ecological status of reservoirs.

A recurring theme across studies is the difficulty of standardizing fish metrics across diverse reservoir types, namely, due to differences in sampling protocols, species pools, and pressure gradients. Species-specific, regionally restricted indices (e.g., the CZ-FBI) may not transfer well to other countries or ecoregions without extensive recalibration [21]. On the other hand, family or guild-level indices (e.g., the C/P Index) may offer broader applicability but with reduced ecological specificity [30]. Additionally, challenges such as sampling bias, gear selectivity, and lack of reference conditions persist across methodologies and remain barriers to broader adoption. This suggests the use of hybrid indices (guild + species + functional traits) and pressure-specific sub-indices (e.g., eutrophication and hydromorphology) to enhance standardization, aiming for a WFD-defined index for reservoirs.

In fact, Birk et al. (2012) [62] claim an underrepresentation of fish in WFD assessments, with only 15% of the 297 methods reviewed in their study being related to fish as a biological quality element, making them one of the least-used groups for bioassessment, especially in lakes and transitional waters. This is due in part to practical difficulties with sampling fish at large scales and because it is not required in all water categories under the WFD. There are also concerns that many fish-based methods were developed with insufficient empirical support, making their sensitivity and diagnostic power questionable. This does not mean fish indicators do not possess potential, but, in fact, that they require more robust calibration, long-term datasets, and better integration with physical habitat and hydromorphological data. As agreed in Poikane et al. (2014) [63], efforts of intercalibration among countries should be encouraged, as they have been marked as valid approaches and can be the key for long-term and effective protection of member state water bodies, in this case, reservoirs.

4.2. Case Study—Aguieira Reservoir

Building on the broader issues discussed above, case in point, the use of fish communities as indicators of water quality, it becomes important to provide a relevant real case to examine how fish-based indices perform in practice. The Aguieira Reservoir was chosen as a case study for this reason, being the only case in Portugal included in the intercalibration exercise of the WFD for defining elements to be assessed in reservoirs. The ecological classification of this ecosystem has consistently been poor [40,64], a pattern also observed in this study, based on the quantified physical, chemical, and biological results (Table 7). The application of the F-IBIP in a Portuguese reservoir characterized by the dominance of tolerant and invasive species offers a way to test and illustrate some of the methodological constraints identified in the literature, including sampling bias, index sensitivity, and the adequacy of standardized metrics in lentic settings. This case study was not intended to serve as a comprehensive ecological characterization of the reservoir but rather as a practical demonstration of how the limitations highlighted in the previous sections may manifest in real-scenario applications of fish-based indices, providing a representative example of the difficulties faced when assessing ecological potential in such systems under the Water Framework Directive.

Regarding the fish specimens found in Aguieira, it is important to highlight that the use of multi-mesh gillnets alone may introduce sampling bias and limit the representation of small species. Different sampling methodologies ought to be employed together, as explored in Mueller et al. (2017) [22]. Despite this limitation, gillnets remain useful for detecting dominant or larger-bodied species, which, in the case of highly disturbed or simplified fish communities, can still provide relevant information about ecological condition and potential. As sampling was limited to a standardized multi-mesh gillnet protocol [18] at three representative sites, this small sample size constrains representativeness. However, it allows the index’s sensitivity to be tested in a highly disturbed system and illustrates the methodological constraints of fish-based bioassessment in reservoirs, as reported by Blabolil et al. (2016) [17], Blabolil et al. (2017) [18], and Mueller et al. (2017) [22]. The shad is native to Portuguese rivers, and its abundance has drastically decreased, especially in the rivers Tagus, Sado, Guadiana, and Douro, with the latter facing a serious reduction [65,66]. However, significant shad populations still exist in the Mondego, Lima, and Minho Rivers, where fishing for this species is economically important for fishing communities [66]. In addition to migratory populations, there are land-locked populations, particularly in reservoirs, such as Castelo do Bode (Tagus River) and Aguieira (Mondego River), where they remained after the construction of dams [66]. This species primarily feeds on crustaceans and filamentous algae by filtering water, while in brackish water, it feeds almost exclusively on crustaceans [66].

The barbel (Luciobarbus bocagei) is an endemic cyprinid of the Iberian Peninsula with a wide distribution across most river basins in mainland Portugal, except for the Guadiana, Mira, and Algarve Rivers, and the Minho River [66]. It is highly adaptable to various freshwater systems, occupying a broad spectrum of lotic and lentic ecosystems, preferring areas with weak-to-moderate currents while avoiding cold waters. These traits often correspond to the middle and lower sections of water courses. This species is well-adapted and is present in a significant percentage of reservoirs [67]. The common barbel (Barbus barbus) is a bottom-feeding, omnivorous, and opportunistic species, feeding on plant material (aquatic macrophytes and filamentous algae), insects, and crustaceans [67]. The introduction of some exotic species has also contributed to the barbel population decline on the Iberian Peninsula, mainly due to competition for resources (e.g., food and habitats) and predation, particularly of juveniles, by introduced piscivorous species, such as the largemouth bass (Micropterus salmoides), northern pike (Esox lucius), and pikeperch (Sander lucioperca) [67].

Regarding the invasive species captured, we found pikeperch (Sander lucioperca), which inhabits deep, calm areas with rocky bottoms and turbid waters, though it can also be found in the water column. Its distribution includes the Azores, the Ermal and Aguieira Reservoirs, the Mondego River and its tributaries, and Lamas de Olo (Vila Real) [66]. Adults feed exclusively on fish (bleak, barbel, nase, and crayfish, which can be their primary food source during certain times of the year), while juveniles feed on crustaceans (e.g., Daphnia) [68]. Carps (Cyprinus carpio) prefer lentic habitats with fine substrate, abundant aquatic vegetation, and warmer waters. They are found in reservoirs and are highly resistant to low dissolved oxygen levels and poor water quality [11], a fact recorded in this ecosystem regarding physical, chemical, and biological parameters (Table 7). During winter, they remain near the bottom, sometimes burying themselves in sandy/muddy substrate, limiting their activity until spring arrives. Spawning occurs in May/June in shallow areas with abundant vegetation to which the eggs adhere and follow an omnivorous feeding regime (vegetable detritus, insects, crustaceans, and fry of other fish) [11]. Like carp, the goldfish (Carassius auratus) prefers lentic habitats with fine substrate and abundant aquatic vegetation, often found near the shore in calm zones [13]. It is also present in reservoirs and is highly resistant to low dissolved oxygen levels and poor water quality (Table 7). Spawning occurs between May and June, in shallow areas with abundant vegetation to which the eggs adhere, with an omnivorous diet [13]. Pumpkinseed (Lepomis gibbosus) is also a lentic species found in reservoirs and shallow watercourses with low current speed, aquatic vegetation, and warmer temperatures. Like most non-native species, it is resistant to deteriorating environmental conditions [69]. Reproduction occurs between May and July, during which they build nests (small depressions in sandy or gravelly substrate) defended by males, who display strong territorial behavior. Most individuals primarily feed on insects, though they may also prey on eggs, larvae, and small fish [69]. These species occupy similar habitats and share identical ecological characteristics, making them unsuitable for distinguishing the quality of aquatic ecosystems.

The results obtained for fish communities showed low representativeness and diversity, regarding the capture method used (multi-mesh gillnets), and future efforts ought to include complementary methods (e.g., seine nets) and increase temporal coverage to account for diel activity of species. According to Mueller et al. (2017) [22], while electrofishing is inefficient in lakes and reservoirs, other capture methods, such as seine nets and gillnets, are more effective and less selective, resulting in broader representativeness and diversity of fish species. The authors also noted that repeating capture methods with seine nets, electrofishing, and lift nets with bait should be considered to increase capture efficiency and representativeness, although this study was conducted in the backwaters (lentic systems) of a river. Mueller et al. (2017) [22] also advocate using capture methods at different times of the day (day, night, dusk, and dawn) for a more realistic assessment of fish communities. The higher capture rates of certain species at night, dusk, and dawn may be explained by species-specific behavioral differences, such as preferred hunting times during dusk and dawn (e.g., Sander lucioperca). In addition to the low representativeness of fish species recorded, the high percentage of invasive species was noted as a parameter with known negative effects on water quality evaluation [27]. Although the physical and chemical results indicated poor water quality (Table 7), invasive species are known to easily adapt to eutrophic environments and contribute to water quality deterioration [27]. Conversely, some exotic species (e.g., pikeperch—Sander lucioperca and Micropterus salmoides—largemouth bass) prey on native species (e.g., barbel—Luciobarbus bocagei) [68], decreasing water quality classification, as most fish indices use native species as a parameter for the ecological assessment of the ecosystem.

The ichthyofauna of the Aguieira Reservoir has been studied since 1999, and it was evaluated and characterized through sampling using gillnets, electrofishing, surveys, and data from previous studies [70]. Thirteen species were recorded: shad—Alosa sapidissima, eel—Anguilla anguilla, common barbel—Barbus barbus, goldfish—Carassius auratus, Iberian nase—Pseudochondrostoma polylepis, goby—Gobio lozanoi, carp—Cyprinus carpio, northern chub—Leuciscus carolitertii, ruivaco—Achondrostoma oligolepis, spined loach—Cobitis taenia, pumpkinseed—Lepomis gibbosus, largemouth bass—Micropterus salmoides, and trout—Oncorhynchus mykiss. The community showed low diversity, with about 42% comprising introduced species, with some very common, such as pumpkinseed (Lepomis gibbosus) and largemouth bass (Micropterus salmoides). The disappearance or decline of several migratory species (lamprey—Petromyzon marinus, shad—Alosa sapidissima, twaite shad—Alosa fallax, eel—Anguilla anguilla) is a clear indication of the artificialization of the ecosystem due to dam construction [70]. There is no reference to the presence of lamprey (Petromyzon marinus) in the reservoir, and the presence of twaite shad (Alosa fallax) was indicated only by local sport fishers. However, fishers sometimes call smaller shad (Alosa sapidissima) specimens “twaite shad” due to morphological similarities between the two species [70].

As in other basins, the shad (Alosa sapidissima) maintains a resident population in Aguieira, whose migration was hindered by the physical barrier (dam). These individuals have adopted a completely different lifestyle from their anadromous migrating counterparts, a phenomenon known as “land-locked populations” [70]. Studies by Pedroso (1997) [71] and Sales-Luís (1998) [72] using electrofishing in this reservoir reported pumpkinseed—Lepomis gibbosus as the dominant species, with a significant presence of largemouth bass—Micropterus salmoides, goldfish—Carassius auratus, common barbel—Barbus barbus, Iberian nase—Pseudochondrostoma polylepis, eel—Anguilla anguilla, and shad—Alosa sapidissima [70]. Surveys from fishing clubs corroborate these results, indicating pumpkinseed—Lepomis gibbosus as the predominant species, while largemouth bass—Micropterus salmoides, common barbel—Barbus barbus, Iberian nase—Pseudochondrostoma polylepis, and carp—Cyprinus carpio are also common in this reservoir. The latter species can reach large sizes, according to local sport fishers (e.g., total length: 70 cm; total weight: 6 kg). In general, the information collected indicates a decline in Pseudochondrostoma polylepis – Iberian nase, Carassius auratus—goldfish, and Squalius carolitertii—northern chub populations, coinciding with the proliferation of introduced species [70]. In the littoral zone, mainly sedentary species adapted to lentic waters, such as pumpkinseed—Lepomis gibbosus and largemouth bass—Micropterus salmoides, were found. The pelagic zone (surface and deep) is primarily inhabited by exotic limnophilic cyprinids like the common carp (Cyprinus carpio) and goldfish (Carassius auratus), and native cyprinids, particularly potamodromous species like Pseudochondrostoma polylepis—Iberian nase, and Barbus barbus—common barbel [70].

The classification of “poor” water quality from the F-IBIP highlights the challenges and opportunities in using fish-based metrics to evaluate reservoir water quality, which, despite representing only a preliminary exploratory study, is in agreement with WFD metrics (Table 7). In this assessment, a “poor” ecological potential indicates that the biological (ichthyofauna) quality of the aquatic ecosystem is severely degraded. This classification is characterized by low native fish diversity, a high prevalence of invasive species, and disruptions to natural ecological balance. These issues reflect broader challenges in bioassessment, such as the difficulty in distinguishing natural variability from anthropogenic impacts and the need for region-specific indices. Moreover, the fish community structure in Aguieira Reservoir points to environmental stressors, including pollution, habitat alteration, and the loss of natural stream features, factors that particularly impact sensitive species and reduce overall biodiversity.

The case study of Aguieira Reservoir reflects the same key patterns identified in the literature review, particularly the low representation of native fish species and the dominance of invasive taxa [27,32]. These findings are consistent with the challenges commonly reported across European reservoirs, including limitations in sampling methodologies [22,26] and the sensitivity of fish-based indices to anthropogenic pressures and artificialized ecosystems [17,26]. The prevalence of tolerant, non-native species and the reduced ecological quality, as indicated by the F-IBIP classification, emphasize broader concerns about biodiversity loss and the need for improved monitoring strategies within the Water Framework Directive context. The F-IBIP output needs to be carefully evaluated, given its original application in river systems, although Česonienė et al. (2020) [31] applied the LFI in reservoirs, which is also used in rivers [73], and this approach provides a comparative basis and a starting point for the development of a reservoir-specific index.

Nevertheless, the application of the F-IBIP to Aguieira highlights several limitations of the index itself, many of which are reflected in the literature. This index still does not account for population parameters such as age, gender, or biomass/catch unit effort, which have been previously indicated by other authors [17,18,25,29] as important and sensitive markers in demonstrating water quality and ecosystem disruption, nor reproductive metrics, which are critical in reservoirs affected by long-term stressors. Moreover, its reliance on taxonomic metrics alone may overlook functional changes in the community, such as the rise of opportunistic feeders or shifts in trophic dynamics. Compared to hybrid or trait-based indices, such as the CZ-FBI or the C/P Index, the F-IBIP appears less sensitive to the nuanced ecological degradation seen in Aguieira. A trait-based approach might have captured functional homogenization within the fish community or better reflected the dominance of stress-tolerant, invasive species, as discussed in the previous section. Another key limitation is the calibration of the index itself. Since it was originally developed for lotic systems, its application to reservoirs may require adjustment or recalibration, particularly in accounting for lentic-specific dynamics such as depth gradients, water retention time, and pelagic vs. benthic species distributions. These limitations are consistent with the broader concerns raised in Section 4.1, as shown in, e.g., Launois et al. (2011) [19], that used different metrics (% of lithophilic species biomass, omnivorous species biomass, herbivorous species count, % piscivorous species, strictly lithophilic species %, and BPUE) to define a specific index for French lentic systems due to the lack of adequate FBIs for these scenarios. This reinforces the need for regionally calibrated, functionally inclusive indices to ensure meaningful ecological assessments in Mediterranean reservoirs.

This case study also provides a concrete illustration of the key challenges highlighted in the literature review. The limited sample size and reliance on a single sampling gear also highlight a key methodological challenge in reservoir bioassessment of balancing the need for standardized, WFD-compliant approaches with the representativeness required for robust community assessment. As reported by Blabolil et al. (2016) [17], Blabolil et al. (2017) [18], and Launois et al. (2011) [19], fish-based indices originally developed for rivers often show reduced sensitivity when applied to reservoirs, particularly in systems dominated by tolerant, generalist, or invasive taxa, a pattern that emerged clearly in Aguieira, where 62.5% of the captured individuals were non-native species, and native assemblages were strongly underrepresented. The poor F-IBIP classification reflects this imbalance and underscores the limitation of relying solely on taxonomic metrics without incorporating functional traits, biomass-based measures (e.g., BPUE and CPUE), or pressure-specific indicators, such as the Cyprinidae/Percidae ratio proposed by Pieckiel et al. (2024) [30]. Thus, the Aguieira results not only confirm but exemplify the gaps identified in the literature, supporting the call for regionally calibrated, hybrid indices that integrate species, guild, and functional metrics to improve ecological potential assessments in Mediterranean reservoirs.

This underscores a key challenge. Fish metrics must account for multiple stressors and their interactions to provide accurate water quality assessments. However, these challenges also present opportunities. Advancements in fish-based bioindicators, coupled with improved monitoring techniques and adaptive management strategies, could enhance the reliability of these assessments. Due to data limitations, no statistical correlations between fish metrics and environmental variables could be conducted. Future studies should, therefore, expand sampling effort and integrate complementary gears (e.g., seine nets and baited traps) to overcome these limitations and support the development of hybrid, functionally inclusive indices for reservoirs. Future applications should incorporate multivariate approaches (e.g., RDA and PCA) to better understand the environmental drivers shaping fish community composition and improve the ecological interpretation of index responses. By improving fish indices and integrating them with complementary ecological data, such as hydromorphological and physical and chemical parameters, water managers can develop more effective conservation and restoration strategies for reservoir ecosystems.

5. Conclusions

Fish communities have become an increasingly important component in water quality classification since the implementation of the WFD. Although there is extensive literature on fish communities as bioindicators for rivers, the scientific literature on reservoirs remains underdeveloped, as shown in the present study, with a small final sample size of twelve articles reviewed. This limited sample size does not merely reflect a constraint of methodologies but highlights a significant gap in the current scientific literature, particularly regarding the role of exotic species in water quality assessment. The scarcity of studies specifically addressing fish communities in reservoirs, especially in the context of invasive species and their ecological impacts, underscores the urgent need for more targeted research in this area. Only in the past decades have studies been conducted under the WFD that focus on these communities as bioindicators. Indices using ichthyofauna in their metrics have been created, but several challenges hinder their development. They must be paired with other organisms (benthic invertebrates and phytoplankton) and physical and chemical water parameters; insufficient data prevent their inclusion in biotic indices due to the lack of literature on the best sampling methods; anthropogenic pressures are a limiting factor; and the presence of exotic species negatively impacts the assessment of the ecological potential of water bodies.

Some authors argue that the indices created could soon be improved, eventually aiding in the assessment and classification of the ecological potential of ecosystems and guiding future restoration efforts to mitigate the effects of anthropogenic pressures. Indices that primarily use biomass and catch numbers in their classification could benefit from improvements that account for various anthropogenic pressures, increased sampling effort, and the systematic integration of invasive species as a biological indicator in their parameters, especially considering their demonstrated association with degraded water quality conditions in reservoirs, as observed in this review.

In the case study presented, the lack of representativeness and diversity of fish species, along with the high percentage of invasive species, was a limiting factor in classifying the water quality of the Aguieira Reservoir. Future studies will need to consider these factors and identify other relevant parameters (e.g., biomass per unit effort—BPUE and catch per unit effort—CPUE) to classify water quality. Additionally, using more sampling methods simultaneously (e.g., seine nets, gillnets, bait fishing, and baited traps) and increased sampling efforts (repeating sampling methods at different times of the day) are recommended to obtain more accurate data on species diversity and representativeness. However, it is essential to assess and include the effects of anthropogenic pressures in calculating biotic indices, as well as to use all available information on the occurrence and diversity of invasive species within the metrics of an ichthyofauna biotic index. Therefore, advancing the integration of fish communities into ecological assessment frameworks while addressing current limitations represents not only a scientific necessity but also a critical step toward more accurate, holistic, and actionable water management strategies that reflect the true health and resilience of aquatic ecosystems.