Ecological status of fish fauna from Razim Lake and the adjacent area, the Danube Delta Biosphere Reserve, Romania

The aim of this study was to determine the ecological status of fish fauna of Razim Lake under the conditions of the water salinity chang - ing from brackish, almost 70 years ago, into freshwater nowadays. The natural processes of siltation and organic deposits, characteristic of Danube Delta lake complexes, intensified in the last decades and included also Razim Lake. The presently reported study of Razim Lake and the adjacent area was undertaken in 2020 with intention to cover fish fauna collected with three different sampling methods (electrofishing, gillnetting, and seining). For each sampling method, Catch per Unit Effort (CPUE), relative abundance


Introduction
The Razim-Sinoie lake complex is situated in the southern part of the Danube Delta Biosphere Re-Staras, unpublished * ). The lake complex has two connections with the Black Sea from Sinoie Lake through the Periboina and Edighiol canals. These two openings to the sea maintain fish diversity and productivity of the entire lake complex (Staras, unpublished). Razim Lake is connected with Sinoie Lake through two canals (named Canal II and Canal V) that provide slightly brackish water for Razim Lake. The hydrotechnical works of the early 1970s transformed Razim Lake into a reservoir with 1 billion m 3 of freshwater (Staras, unpublished). Moreover, the salinity of Razim Lake changed over a short time, as proven by Leonte et al. (1956Leonte et al. ( , 1960, from 2.5‰ in 1951 to 0.5‰in 1956 due to the freshwater influx from the Danube River. The Danube River, via the Sfântu Gheorghe arm as a major path, transports water and solids into Razim Lake via the Dunăvăț, Dranov, Mustaca, and Lipoveni canals. The mean monthly flows on Sf. Gheorghe arm indicate 9.66% of total flow (135 m 3 · s -1 liquid flows) and almost 2 million t · year -1 (solid flows) from the total flows of the arm by continuous lateral discharge to the Razim system (Driga 2004). The general water balance shows that the share of inputs is 90% from supply canals (Dranov, Dunăvăț, and Lipoveni canals), 9% from precipitation, and 1% from small rivers (Slava, Taița, Telița, Agighiol) and the exits from the system are represented by evapotranspiration (15%) and 85% irrigation and evacuation (Bondar cited by Staras, unpublished). The Danube River is the water supplier for all Danube Delta lakes including Razim Lake with which it has also an active exchange of fish fauna, especially at high river water levels because of the high degree of siltation of connecting canals in 2020. The diversity and structure of the fish community varies amongst lakes and can be regarded as a good indicator of the ecological state of the lakes. The aim of this study was to describe the ecological status of fish fauna from Razim Lake and the adjacent area, based on a fish survey conducted in 2020 and to discuss changes, based on earlier scientific publications.

Materials and methods
Study area, sampling period, fish, and water measurements. The study area was represented by five sectors of Razim Lake, a large-surface lake: Fundea Gulf (1), Holbina Gulf (2), southern lake (3), Mustaca sector north and south and Oaia Lake (4), west Lake Enisala (5), and canals (Dunăvăț, Mustaca, Dranov) ( Fig. 1), with each sector being sampled at multiple sites. The ichthyofauna was sampled in Razim Lake and the adjacent area in July, August, and September of 2020. For biometric measurements, an ichthyometer with an accuracy of 1 mm per 50 cm for fish length and for weight, an electronic scale with an accuracy of 1 g per 5 kg were used. Geographical coordinates and physical-chemical parameters observed in the area were recorded with a Garmin device and Hach multiparameter, as well as a Secchi disc for water depth and transparency.
Fish sampling. The fish sampling and Catch per Unit Effort calculation (CPUE) was done in accordance with EU recommendations by use of common methods: • Electriofishing with SAMUS 1000 W electrofisher device, transect with multiple electric points during 10 min per site, the catch being standardized at individuals or g · h -1 of fishing effort (for shoreline or small canals from compact reed developed nearby lake). • Passive gillnet fishing (stationary 12 h by night, the catch being standardized at 100 m 2 gillnets per night): commercial gillnets or Nordic gillnets multi-meshes fishing tools (30 m length × 1.8 m high each). The Nordic gillnets have 12 randomly joined panels, each panel being 2.5 m in length, with multiples meshes: 6, 6, 8, 10, 12, 16, 20, 24, 30, 35, 45, and 55 mm (Nyberg and Degerman 1988;Năvodaru 2008) (main tools used in Razim Lake and adjacent area). • Seine fishing with 2 wings of 100 m length each and a codend of 7 mm knot-to-knot mesh size. Standardization to one haul of active fishing (1 h). • Directly observed species from angling and some traditional fishing tools (fyke net, hand cast net, fish landing) just for fish species identification, without other standardization.
Taxonomy and ecology. The fish species scientific names used are consistent with the Eschmeyer's Catalog of Fishes (Fricke et al. 2021). The specimens collected were identified after Antipa (1909), Cărăusu (1952), Bănărescu (1964), and taxonomic name and support knowledge after revision by some authors (Otel et al. 1992(Otel et al. , 1993Kottelat 1997;Otel 2001Otel , 2007Sindrilariu et al. 2002;Nelson 2006;Kottelat and Freyhof 2007;Năvodaru and Năstase 2011;Năstase et al. 2017Năstase et al. , 2019aFroese and Pauly 2021;Năstase, unpublished * ). Relative abundance and biomass for each species and sampling methods were calculated as standard CPUE (Catch Per Unit Effort). The relative abundance or dominance (D) for each species and sampling methods was calculated as the proportion of species to total catch (Mühlenberg 1993; Sindrilariu et al. 2002). The relative abundance or dominance (D) for each species and sampling methods was calculated as the proportion of species to total catch (D i = n i · 100N -1 (%), where, D i = dominance of species i, ni = individuals of the species i, and N = total number of individuals) (Mühlenberg 1993; Sindrilariu et al. 2002). The frequency of occurrence (F) or constancy (C) for each species and sampling method was calculated as the proportion of samples containing a species from the total number of samples (C i = b i · 100a -1 (%), where, C i = frequency of occurrence of species i, b i = the number of samples in which species i was observed and a = total number of samples) (Schwerdtfeger 1975;Sindrilariu et al. 2002). Ecological significance (W) is a relation between frequency (C) and dominance (D) (W = D · 100C -1 ). For frequency, five classes were used; six classes were used for abundance/dominance data analysis, and seven classes were used for ecological significance (Table 1).
To determine ecological status of the lake, some quantitative ecological parameters were chosen as most expressive for fish communities: Relative Abundance in Number per Unit Effort (NPUE), Relative Biomass in Biomass Per Unit Effort (BPUE), the biodiversity index according Shannon-Wiener Index H s , and Equitability Index = Evenness index (E) as in Năstase et al. (2019aNăstase et al. ( , 2021 (Table 2). * Năstase A (2009)  An ecological status classification matrix in accordance with the Water Framework Directive (WFD) is presented in Table 2 regarding the fish community. The Biodiversity Index (H s ), according to the Shannon-Wiener formulae, as well as maximal fish Diversity (H max ) and Equitability (Evenness) Index (E) were calculated. The Equitability Index describes the quantum of unequal distribution of different effective species proportion as an ideal community, ranges between 0 and 1. The Shannon-Wiener Index varies from values of 0 for communities with one species, to various other values for more mixed species (Odum 1975;Botnariuc and Vădineanu 1982;Gomoiu and Skolka 2001;Sârbu and Benedek 2004). Formulas used: according Shannon-Wiener formulae  where p i is the dominance; N r is the number of individuals belonging to a certain species; and N = total number of individuals in a sample.
According to the Water Framework Directive, it is desirable to test and apply known ecological parameters that could improve the methods of assessing the ecological status, using, when no other methods are available, even expert judgement analysis (this analysis from papers was thought of and used in a European project in 2014: Black Sea e-Eye -Innovative Instruments for Environmental Analysis in NW Black Sea Basin, to improve methodology after Moss et. al. (2003) and Ibram et al. (2015). The ecological lake classification matrix is in accordance with the Water Framework Directive. EU Water Framework) has five (I-V) limits classes marked with different colors. Actually, there are yet no developed statistical threshold limits classes (I-V) for those chosen ecological parameters (NPUE, BPUE, H s , E) according to the WFD water quality regarding fish, but expert judgement was used as a future proposal. Class limits was proposed by the present authors, based on field experience and expert judgement in the Danube Delta (Năstase et al. 2019a(Năstase et al. , 2021. In the summer of 2020, sampling was conducted using 77 Nordic gillnets, totaling 2310 m of passive nets per night, 190 minutes of electric fishing, five seine active hauls and 48 commercial gillnets 1440 m in total of passive netsnights-1 in total (Table 3).

Results
In the summer of 2020, we captured 8042 fish individuals with more than 573 kg of fish and 36 individuals weighting in a total of almost 1.5 kg of crayfish (Table 3).
Species richness. All captured individuals belong to 43 fish species and one crayfish species-Pontastacus leptodactylus. Overall, Razim's ichthyofauna is dominated by limnophilous or stagnophilic-rheophilic species, such as white bream, Blicca bjoerkna (Linnaeus, 1758) and roach, Rutilus rutilus (Linnaeus, 1758), followed by characteristic-complementary-associated species, such as ziege, Pelecus cultratus (Linnaeus, 1758); European perch, Perca fluviatilis Linnaeus, 1758; pike-perch, Sander lucioperca (Linnaeus, 1758); common bream, Abramis brama (Linnaeus, 1758); bleak, Alburnus alburnus (Linnaeus, 1758); and gibel carp, Carassius gibelio (Bloch, 1782), but the majority of species occur sporadically in the Lake, with a significant number of species being accidentally found here (Table 4). The numbers for the goby species-monkey goby, Neogobius fluviatilis (Pallas, 1814); round goby, Neogobius melanostomus (Pallas, 1814); racer goby, Babka gymnotrachelus (Kessler, 1857); bighead goby, Ponticola kessleri (Günther, 1861); syrman goby, Ponticola syrman (von Nordmann, 1840); mushroom goby, Ponticola eurycephalus (Kessler, 1874)-are worrying, as they are in a continuous decrease, being limited only to certain favorite places of the Lake, especially in the areas with submerged stones (used to avoid clogging of the mouths of the canals) and gravel areas, compared to the previous years when they dominated even sandy areas. It can be said that this phenomenon of numerical reduction of the gobies populations in Razim Lake is due to the obvious habitat changes which include increase of siltation, the mud of the Razim Lake transforming the lake into a pond, typical for lake complexes from the Danube Delta. Another question mark is the existence of percarina, Percarina demidoffi von Nordmann, 1840 (Percidae), a non-native not invasive, but sensitive species, first recorded 1986 (Otel and Bănărescu 1986). In recent years, it has not been found in Razim Lake, in the place where this species had formed vigorous populations in the past, even stronger populations than in its native range (Don River), the cause probably also being habitat change.
Out of the 43 fish species captured or observed in Razim Lake, nearly 1/3 are without commercial value (small fish) and 2/3 (30 fish species) have commercial value. From these 30 commercial fish species, more than 1/4 have high commercial value-pontic shad, Alosa immaculata Bennett, 1835; pike-perch, Sander lucioperca; Wels catfish, Silurus glanis Linnaeus, 1758; common carp, Cyprinus carpio Linnaeus, 1758; European eel, Anguilla anguilla (Linnaeus, 1758); and northern pike, Esox lucius Linnaeus, 1758). Almost half of the species have medium market value (like gibel carp, rudd, roach, tench, perch, bream, etc.) and almost 1/4 have low economic value (goby species). Of the 43 fish species, the majority are native and four are non-native species: Chinese sleeper, Perccottus glenii Dybowski, 1877; silver carp, Hypophthalmichthys molitrix (Valenciennes, 1844); grass carp, Ctenopharyngodon idella (Valenciennes, 1844); pumpkinseed sunfish, Lepomis gibbosus (Linnaeus, 1758). While some of the species are migratory, reophilous or reofilous-stagnofilous, such as Alosa immaculata, Anguilla anguilla, and white-eye bream, Ballerus sapa (Pallas, 1814), occur rarely in the Lake, others are stagnofilous-reophilous or limnophilous species which are the majority. The stagnoph-  (Table 4). The parameters used in the ecological characterization of Razim Lake from the point of view of the ichthyofauna show that they fall into the moderate class, the majority of the indicators having moderate and good values, but according to the "one out, all out" principle there are some indicators in the moderate state class, which makes us assert that Razim Lake has a Moderate ecological status in 2020 (Table 5 and 6). Some large fish individuals like Sander lucioperca, Silurus glanis, and Abramis brama were rarely found during our sampling campaign in Razim Lake, probably due to legal and illegal overfishing. Extensive poaching with nylon and small mesh-size gillnets fishing is one of the most dangerous practices in reducing the quality and size of fish populations in the area. There is no precise estimate of the extent of poaching in Razim Lake since 1990, but it is believed that poaching is threatening all animals, especially fishes. Razim Lake, the largest lake of Romania has always been fascinating for studies of fish fauna, especially due to the contact of freshwater with the brackish water, which make it a "natural biological laboratory" of living fish population species, with a lot of hybrid individuals or subspecies. The diversity indices of Razim Lake and adjacent water bodies indicate a stable ecosystem, so a stable fish coenosis, with values of equitability (E) more than medium 0.5 for each sampling method. Shannon-Wiener Index values are increased, the Table 4. Ecological significance of fish species from Razim Lake and the adjacent area (also included classes "Present = P" for species which could not be standardized, just observed). boundaries are more than 1.955 with the maximum on the shorelines or canals from the reed band on the shoreline of the Lake (Fig. 2).

Species Nordic gillnets Commercial gillnets
Relative abundance and biomass. Relative abundance (CPUE) is dominated by bream species (especially white bream), roach, giebel carp, perch, bleak, rudd, and ziege, but for the majority of fish species, it has low values, with some differences between sampling methods (Fig. 3). Relative biomass (CPUE) was dominated by common carp, gibel carp, roach, white bream, ziege, perch, pikeperch, and rudd with some differences between sampling methods (Fig. 4).

Discussion
Since the 19 th century, when Grigore Antipa drew attention to the decline in fish production in Razim Lake, reaching less than 1/3 of what it was 15 years before his studies   (Antipa 1894), the trend in 2020 remains the same, mainly due to legal and illegal overexploitation, even with the appearance (1895 first fishing permit) and periodic updating of fishing laws. Even at the beginning of the 21 st century, contravention of the fishing laws is usually not considered a serious offence in courts of law. In the past, the marine species entering Razim Lake in significant quantities were: blunt-snouted mullet,  Leonte et al. 1960;Otel et al. 1992Otel et al. , 1993Staras, unpublished), but succession of species happens due to changes in water salinity. Namely, in 2020, only rare, accidental entry of Alosa immaculata, Anguilla anguilla and some mullets species was observed, with higher presence of Alosa tanaica, Atherina boyeri and freshwater species. Historic data (Leonte 1969 cited by Staras, unpublished) cite around 55 fish species, a considerable number being marine and euryhaline. In 2020, 43 fish species were described in Razim Lake, with 39 native and four non-native (Perccottus glenii;    Otel et al. (1993) and Staras (unpublished) in the Razim-Sinoie Lake complex in the 1990s. Carassius gibelio and Cyprinus carpio are given as native species from Central Europe to Siberia (Kottelat and Freyhof 2007;Otel 2019). The current living conditions favor the development of freshwater eutrophic species with less than 0.5‰ salinity, large variations in dissolved oxygen and increased quantities of nutrients in water. A new non-native fish species was recently recorded in the natural environment of the Lower Danube River Basin, Perccottus glenii, first recorded in the Romanian River Suceava (Nalbant et al. 2004). It was first recorded in DDBR by Năstase (2007). Its range has expanded to Razim Lake, being first recorded in 2016 in Holbina Gulf of Razim Lake (Năstase et al. 2019a). Its population has increased in the Danube Delta (Năstase et al. 2019b) also in the Razim-Sinoie Lake complex, having a strong invasive behavior (Vilizzi et al. 2021), well adapted to new biotope conditions in Razim Lake. Qualitative and quantitative decreases in species numbers and abundance is undesirable throughout the DDBR, not only for Razim Lake. For that reason,  (Lukasch, 1933) were present in the Razim-Sinoie Lake complex in the 1990s (Otel et al. 1992(Otel et al. , 1993Staras, unpublished), some of them in considerable quantities. However, in 2020, only a few species in Razim Lake are migratory, reophilous or reofilous-stagnofilous, such as Alosa immaculata, Anguilla anguilla, and Ballerus sapa, which occur rarely in the Lake and the majority are stagnofilous-reophilous or limnophilous species.

Conclusions
The main species (eudominant, very frequent) in Razim Lake and adjacent waters were white bream, Blicca bjoerkna; roach, Rutilus rutilus; and bleak, Alburnus alburnus, but mostly are accessory, also a significant percentage of species being accidental, with some differences between sampling methods. Relative abundance (CPUE) was dominated by bream species (especially Table 7. Geographical coordinates and physical-chemical parameters observed in some fishing points from Razim Lake and the adjacent area in summer 2020. T = water temperature, WD = water depth, Tr = transparency, Sal = salinity, C = conductivity, Ox = oxygen content, OxS = oxygen saturation (%).
white bream), roach, gibel carp, perch, bleak, rudd, and ziege with low values for the majority of fish species, but relative biomass (BPUE) is dominated by common carp, gibel carp, roach, white bream, ziege, perch, pike-perch, and rudd with some differences between sampling methods. The diversity indices of Razim Lake and the adjacent area point to a more than medium stable fish coenosis, with the most stable being the shoreline area. The parameters used (according to Moss et al. 2003) and four selected ecological parameters used according to the WFD) in the ecological status characterization of Razim Lake from the point of view of the fish fauna, categorise Razim Lake into the moderate class, using the "one out, all out" principle of the WFD. The ecological indicators have not completely captured a decreasing trend in commercial fishing. This aspect is studied for fisheries resources using stock estimations from fishery landings. However, the absence of large fish (pike-perch, wells catfish, common bream) is a sign of overfishing, especially when adult individuals are missing or an insignificant number is spawning, that could have negative repercussions on future generations, such as for pike-perch). The investigation of Razim Lake has always been a challenge for researchers and this paper aims to be a benchmark for future fish ecological studies. From another perspective, monitoring of fish fauna from Razim Lake is vital because it represents the main reservoir of some commercial fish species like pike-perch, common bream, common carp, but also for some important ecological species, such as Percarina demidoffi, Ponticola syrman, and Umbra krameri, as well as to adjust ecological parameters as support for the determination of conservation status.