Distribution Range of the Endangered Species Unio crassus Philipsson, 1788 in Serbia (Western Balkans Region), Historical and Recent Data

: The thick-shelled river mussel, Unio crassus Philipson, 1788, is considered to be one of the species with the highest conservation priority in Serbia. The study represents the first comprehensive research of the distribution of U . crassus in Serbian waters. The research covered a variety of water-body types throughout Serbia, and distribution data were considered over three time periods from 1953 to 2019. The paper summarizes all the available literature data, field research and information obtained during the review of the collection of malacological material of the Natural History Museum in Belgrade. The results show a positive population trend, which is reflected in an extension of the distribution area and an increase in population density. After reviewing the museum collection, 13 synonyms for U . crassus were identified. The study also revealed a better insight into the habitat requirements and the limiting factors of the species. Substrate characteristics, waterbody types, altitude, and nitrate content of the water seem to be of great importance for the occurrence of the species. The results presented here can improve further measures for the conservation of U . crassus , not only in Serbia, but also in the Western Balkans.


Introduction
Freshwater mussels (Unionida: Unionidae) are one of the most important and widespread groups of aquatic organisms, found in a variety of freshwater habitats, throughout the world.These bivalves are an essential component of freshwater [1] and contribute to sediment stabilization, nutrient cycling and water purification with positive effects on freshwater biodiversity [1][2][3].The Unionidae (also known as bivalves, naiads and unionids) are among the most threatened faunistic groups at a global level [4][5][6][7][8][9][10].Of the 16 European species of the order Unionida, nine have the status of near-threatened, endangered, or critically endangered according to the IUCN Red List [10,11].For this reason, comprehensive environmental and population studies are very important from a conservation perspective.Mussels also have economic importance as a food source and in the ornamental industry.Their over-exploitation for industrial purposes has led to the population decline of some species in many regions [12,13] and even to their local disappearance.
The thick-shelled river mussel, Unio crassus Philipsson, 1788 is currently listed by the IUCN classification as endangered-EN at a global level [14].It is listed in Annexes II and IV of the European Commission Habitats Directive [15], and in Resolution 6 of the Bern Convention [16]) but is also covered by Serbian legislation [17].These legislations Water 2023, 15, 4248 2 of 16 promote the conservation of unionids, including habitat restoration and the reintroduction of mussels and host fish [18,19].
The native distribution area of U. crassus extends from France in the west to western Russia in the east, and from Scandinavia in the north to Asia Minor in the southeast.It was also recorded in the basins of the Baltic Sea, the Black Sea, the Azov Sea and the Caspian Sea, up to the Ural River basin in Russia and Kazakhstan [10].The species is widespread in Europe, with the exception of Great Britain, the Apennines and the Iberian Peninsula, where its occurrence has not been recorded [10,14].
Until the first half of the 20th century, U. crassus was the most abundant unionid species in Europe [20].Recently, declining population densities and the endangered status of U. crassus have been observed in most European countries, especially in Western and Central Europe [19].According to the latest evidence, the decline of U. crassus in Europe is estimated to be more than 50% [19].The species is listed in the national Red List as critically endangered in Switzerland, Austria and Germany with only a few intact populations remaining [21,22], endangered in the Czech Republic, Poland and Sweden, and vulnerable in Albania, Belarus, Finland and Latvia [14].
Insufficient knowledge of the unionids, their current status and the quantification of population changes over time are all problems for further research on population trends and for determining effective conservation measures not only in Serbia but also in other European countries.U. crassus is a strictly protected species in Serbia according to national legislation.Taking into account the new population data, we assumed that the population trend in Serbia is continuously changing.To confirm these assumptions, population changes over time need to be documented and quantified, including data on the distribution history of the species.It is also important to understand which factors are potentially responsible for the changes over time.
The aim of this study was to use a large amount of distribution data to gain a better insight into the distribution range of the species U. crassus in different time periods; to identify habitat preferences; to discuss anthropogenic factors affecting the distribution of the species; and to solve the problem of using many unaccepted synonyms and not confidently identifying U. crassus in the past in Serbia.
To achieve these goals, this manuscript compiles all known records of U. crassus in Serbia from the literature and unpublished sampling data up to 2019.The information presented here will significantly improve our knowledge of the current situation of U. crassus in Serbia and support the conservation of unionids and their habitats.

Data Collection
The data used for the analysis of the distribution of U. crassus in Serbia cover the period 1953-2019.The distribution was estimated based on all available data: (1) peer-reviewed articles, monographs, dissertations and reports [23][24][25][26][27][28][29][30][31][32]; (2) unpublished data on samples collected during field research of several national projects in Serbia (material deposited in the malacological collection of the Institute for Biological research "Siniša Stanković", University of Belgrade-further referred as the IBISS); (3) material collected during the realization of four international projects (material deposited in the malacological collection of IBISS) [33][34][35][36]; (4) BAES database-biodiversity in aquatic ecosystems in Serbia, ex situ conservation [37]; and (5) collection of unionids of the Natural History Museum in Belgrade (collector Ante Tadić)-referred to as historical data in remaining text.Data (1953Data ( -1973) ) In order to determine the historical distribution of U. crassus in Serbia, the museum collection of unionids was reviewed (collector of Ante Tadić).The analyzed historical material consisted of 244 individuals from 36 sites in Serbia, collected in the Danube and its main tributaries in the Serbian stretch (the Sava, Tisa, Karaš, Tamiš, Nera and MlavaRivers, as well as the Velika Morava and Timok basins)-   collection was inspected and the identification of each specimen was checked and verified.A re-identification was carried out and the presence of potential synonyms for the U. crassus was considered.Each specimen had an inventory number and a label with sampling site, the name of the collector, the date of collection and the identification.A description of the respective sampling location is given in Tadić [38].Key features used in identification were external morphology (shell outline, color, umbo sculpture, hinge characteristics as well as the three measured linear shell distances-shell length, height and width).Review of the status and taxonomic history of the species was carried out according to databases: MolluscaBase [39], MUSSELpdb [40] and WoRMS [41].

Current Data (1990-2019)
Recently, the study of aquatic ecosystems has been intensified and covered the entire territory of Serbia.A total of 540 sites were studied, covering different types of running water-from small and medium-sized streams to large lowland rivers (Figure 1).Various techniques were used to collect mussel samples-kick and sweep sampling and the multihabitat approach (EN 27828:1994) with the FBA benthic hand net (aperture: 25 × 25 cm, mesh size of 500 and 250 µm) according to European Standards [42], benthic dredging and in some cases visual inspection and snorkeling.To obtain comparable data, abundance was expressed as the number of individuals per sample (relative abundance).For the graphical presentation, abundance per watercourse was pooled.
collection of unionids was reviewed (collector of Ante Tadić).The analyzed historical material consisted of 244 individuals from 36 sites in Serbia, collected in the Danube and its main tributaries in the Serbian stretch (the Sava, Tisa, Karaš, Tamiš, Nera and MlavaRivers, as well as the Velika Morava and Timok basins)-Table 1.The museum collection was inspected and the identification of each specimen was checked and verified.A re-identification was carried out and the presence of potential synonyms for the U. crassus was considered.Each specimen had an inventory number and a label with sampling site, the name of the collector, the date of collection and the identification.A description of the respective sampling location is given in Tadić [38].Key features used in identification were external morphology (shell outline, color, umbo sculpture, hinge characteristics as well as the three measured linear shell distances-shell length, height and width).Review of the status and taxonomic history of the species was carried out according to databases: MolluscaBase [39], MUSSELpdb [40] and WoRMS [41].Data (1990Data ( -2019) ) Recently, the study of aquatic ecosystems has been intensified and covered the entire territory of Serbia.A total of 540 sites were studied, covering different types of running water-from small and medium-sized streams to large lowland rivers (Figure 1).Various techniques were used to collect mussel samples-kick and sweep sampling and the multihabitat approach (EN 27828:1994) with the FBA benthic hand net (aperture: 25 × 25 cm, mesh size of 500 and 250 μm) according to European Standards [42], benthic dredging and in some cases visual inspection and snorkeling.To obtain comparable data, abundance was expressed as the number of individuals per sample (relative abundance).For the graphical presentation, abundance per watercourse was pooled.

Current
The distribution data for U. crassus are considered over two time periods (1990-2008 and 2009-2019).The number of detections per study period and per river kilometer was carried out to examine the distribution of the species in Serbian waters.The distribution data for U. crassus are considered over two time periods (1990-2008 and 2009-2019).The number of detections per study period and per river kilometer was carried out to examine the distribution of the species in Serbian waters.

Environmental Variables and Data Processing
The GPS position and elevation of each site in recent research period were recorded using a GarmineTrex 20× handheld GPS receiver (Garmin ltd).The chemical parameter(NO 3 -N) considered in the study was provided by the Serbian Environmental Protection Agency (SEPA), as an official accredited institution (SRPS ISO/IEC 17025:/2017) [43] for the national water-monitoring programs.All water samples were analyzed in the accredited SEPA laboratory according to the following method: nitrates (NO 3 -N): UP 1.98/PC 12.
Water parameters were collected once a month.This parameter was selected based on the literature data as potentially one of the main elements affecting the U. crassus community.
According to the modified national typology [45], all surface waters in Serbia are classified into six categories: Type 1-large lowland rivers; Type 2-large rivers; Type 3-small to medium rivers with elevation below 500 m; Type 4-small to medium rivers and streams with elevation above 500 m; Type 5-watercourses of the Pannonian Plain; Type 6-small waterbodies including springs and upper stretches of streams.For this study, the material was collected in three categories of waterbodies (Types 1, 2 and 3).

Statistical and Graphical Analysis
The study of the ecological preferences of species in terms of the elevation gradient, waterbody types and the gradient of substrate types with the response curve was performed using the STATISTICA 8 software (StatSoft, Inc., Tulsa, OK, USA) [46].The nitrate-nitrogen content in different watercourses was analyzed using General Discriminant Analysis (GDA).The values of abundance and nitrogen content are graphically represented by mean, maximum and minimum values.The maps were created using Adobe Illustrator CC15 (Adobe Inc., 2015) [47].

Historical Data
By analyzing historical data, 94 individuals of the 244 examined specimens were identified as U. crassus.A re-identification of each specimen from the museum collection was carried out and then the scientific names were validated in the database.After reidentification, 13 synonyms for U. crassus were identified (Table 1).The majority of species names are not considered valid based on current knowledge of the freshwater mussel diversity.Many of these synonyms were introduced into Europe by the French Nouvelle École in the late 19th century [48].According to the historical data, considering the period from 1953 to 1973, it can be assumed that U. crassus was a common species in Serbia with a continuous range.The distribution range of the species based on historical data is shown in Figure 2. The majority of species names are not considered valid based on current knowledge of the freshwater mussel diversity.Many of these synonyms were introduced into Europe by the French Nouvelle École in the late 19th century [48].According to the historical data, considering the period from 1953 to 1973, it can be assumed that U. crassus was a common species in Serbia with a continuous range.The distribution range of the species based on historical data is shown in Figure 2. In the archive material, the species was detected at 20 of the 36 examined sites (55%).The species was found to be widespread in the Serbian stretch of the Danube.It was recorded at five out of nine examined sites (55%), from Apatin (1402 rkm) to Golubac (1040 rkm).A total of 16 specimens of the species were collected in the studied section of the Danube.The species was also detected in the Sava River with moderate occurrence, In the archive material, the species was detected at 20 of the 36 examined sites (55%).The species was found to be widespread in the Serbian stretch of the Danube.It was recorded at five out of nine examined sites (55%), from Apatin (1402 rkm) to Golubac (1040 rkm).A total of 16 specimens of the species were collected in the studied section of the Danube.The species was also detected in the Sava River with moderate occurrence, being recorded at three out of the eight examined sites (37%).Five individuals of U. crassus were collected in the Serbian stretch of the Sava (from 76 rkm to the mouth of the Danube).Thirtynine specimens were collected at four sites.A dense population and the highest frequency of occurrence of U. crassus was observed in the Zapadna Morava.Even 39 specimens were collected at four sites in Z. Morava.U. crassus was detected in the Tamiš (in 33% of the examined samples), Mlava (in 50% of the examined samples) and Nera (in all samples) Rivers, but also detected in the Tisa and Crni Timok Rivers (at only one site).The occurrence of the species is observed in the Veliki Bački and Karaš channels.
Between the 1970s and 1990s, a period of intensive industrialization, U. crassus became locally and even regionally extinct.After historical data, there were no records of the species in Serbia until the early 1990s (Figure 3).Danube).Thirty-nine specimens were collected at four sites.A dense population and the highest frequency of occurrence of U. crassus was observed in the Zapadna Morava.Even 39 specimens were collected at four sites in Z. Morava.U. crassus was detected in the Tamiš (in 33% of the examined samples), Mlava (in 50% of the examined samples) and Nera (in all samples) Rivers, but also detected in the Tisa and Crni Timok Rivers (at only one site).The occurrence of the species is observed in the Veliki Bački and Karaš channels.
Between the 1970s and 1990s, a period of intensive industrialization, U. crassus became locally and even regionally extinct.After historical data, there were no records of the species in Serbia until the early 1990s (Figure 3).

Current Data
Of the 540 sites surveyed, mussels were detected at 46 sites.The current distribution of the species is shown in Figure 4 A, B. The number of detections in the watercourses per study period and the river kilometers are shown in Table 2 and Figure 3. U. crassus was detected in the Kolubara [24], Pusta reka [25], Tisa (site Novi Bečej-63rkm) Rivers in 2001 [23], Crni Timok upstream in 2004 [37] and at two sites on the Danube (Stari Banovci and Smederevo) in the period 1990-2008 [26] (Table 2, Figure 4A).

Current Data
Of the 540 sites surveyed, mussels were detected at 46 sites.The current distribution of the species is shown in Figure 4A,B.The number of detections in the watercourses per study period and the river kilometers are shown in Table 2 and Figure 3. U. crassus was detected in the Kolubara [24], Pusta reka [25], Tisa (site Novi Bečej-63rkm) Rivers in 2001 [23], Crni Timok upstream in 2004 [37] and at two sites on the Danube (Stari Banovci and Smederevo) in the period 1990-2008 [26] (Table 2, Figure 4A).
The species was detected in the Danube, Tisa, Sava, Velika and Zapadna Morava Rivers, as well as in the Kolubara River basin (three sites on the main course of the Kolubara and in the Peštan and Ljig Rivers), and according to the literature data, it was also detected in the Južna Morava [39] and Nišava Rivers [40] (Table 2).
During this period, the species was sporadically detected along the Danube, with a low frequency of occurrence and abundance (up to 0.48% of the total mussel community).The species was detected in the Danube only in 2013 at two sites ( Čerević-1273 rkm and Tekija-956.2 rkm).It more frequently occurred in the Tisa River.The species was detected at the sites Titel-11 km upstream of the Danube confluence with the Danube (2010), Ada-130 rkm (2013), Martonoš-155 rkm and Tisa, confluence-2 rkm (2019) with low relative abundance (up to 5.78% of the total mussel community).U. crassus was also detected in the Sava and Velika Morava Rivers along almost the entire stretch with a higher relative abundance (with a percentage participation of 25.42% and 11.59%, respectively) and in repeated sampling occasions.The occurrence of U. crassus in the Kolubara River basin was also confirmed in repeated sampling in the period 2009-2019, but with a low abundance.
The mean value of the population abundance of U. crassus is shown in Figure 5, with the minimum and maximum deviation of abundance in the different watercourses, with the highest population abundance recorded in the Sava River.In the period 2009-2019, U. crassus was detected in 40 of the 120 examined w course sites (33.3%) in Serbia (Figure 4B).
The species was detected in the Danube, Tisa, Sava, Velika and Zapadna M Rivers, as well as in the Kolubara River basin (three sites on the main course Kolubara and in the Peštan and Ljig Rivers), and according to the literature data, also detected in the Južna Morava [39] and Nišava Rivers [40] (Table 2).
During this period, the species was sporadically detected along the Danube, w low frequency of occurrence and abundance (up to 0.48% of the total mussel commu The species was detected in the Danube only in 2013 at two sites (Čerević-1273 rkm Tekija-956.2 rkm).It more frequently occurred in the Tisa River.The species was de at the sites Titel-11 km upstream of the Danube confluence with the Danube ( Ada-130 rkm (2013), Martonoš-155 rkm and Tisa, confluence-2 rkm (2019) with low tive abundance (up to 5.78% of the total mussel community).U. crassus was also de in the Sava and Velika Morava Rivers along almost the entire stretch with a highe tive abundance (with a percentage participation of 25.42% and 11.59%, respectively in repeated sampling occasions.The occurrence of U. crassus in the Kolubara River was also confirmed in repeated sampling in the period 2009-2019, but with a low dance.
The mean value of the population abundance of U. crassus is shown in Figure 5 the minimum and maximum deviation of abundance in the different watercourses the highest population abundance recorded in the Sava River.

Ecological Preferences
The distribution of U. crassus is observed predominantly in the littoral reaches of large lowland rivers (waterbody Types 1 and 2), where fine substrate predominates (psammal/psammopelal (6 µm-2 mm) and in small to medium watercourses (Type 3), where coarse substrate (mesolithal 6-20 cm and microlithal 2-6 cm) predominates, at elevations of up to 500 m (Figure 6A-C).It can be characterized as a rheo-to limnophilous species, preferring habitats with slow to moderate water flow.

Ecological Preferences
The distribution of U. crassus is observed predominantly in the littoral reaches of large lowland rivers (waterbody Types 1 and 2), where fine substrate predominates (psammal/psammopelal (6 µm-2 mm) and in small to medium watercourses (Type 3), where coarse substrate (mesolithal 6-20 cm and microlithal 2-6 cm) predominates, at elevations of up to 500 m (Figure 6A-C).It can be characterized as a rheo-to limnophilous species, preferring habitats with slow to moderate water flow.
The distribution of U. crassus is observed predominantly in the littoral reaches of large lowland rivers (waterbody Types 1 and 2), where fine substrate predominates (psammal/psammopelal (6 µm-2 mm) and in small to medium watercourses (Type 3), where coarse substrate (mesolithal 6-20 cm and microlithal 2-6 cm) predominates, at elevations of up to 500 m (Figure 6A-C).It can be characterized as a rheo-to limnophilous species, preferring habitats with slow to moderate water flow.Comparing nitrate-nitrogen levels for the same water bodies between the different monitoring years shows that nitrate-nitrogen levels were higher in the period 1999-2007 than in the most recent monitoring period (2011-2019) (Figures 7 and 8).Comparing nitrate-nitrogen levels for the same water bodies between the different monitoring years shows that nitrate-nitrogen levels were higher in the period 1999-2007 than in the most recent monitoring period (2011-2019) (Figures 7 and 8).

Discussion
This study represents the first comprehensive research on the distribution of U. crassus in Serbian waters, based on historical, literature and field data.
Considering all the collected data, it can be observed that the distribution and abundance of U. crassus varies in the different study periods (Figures 2 and 4A,B) and in different watercourses (Figure 5).Re-identification of the archive samples from the Serbian Natural History Museum (period from 1953 to 1973) revealed that U. crassus was a common species with continuous distribution throughout Serbia until the mid-1970s.After re-identification (Table 1) of the museum collection, it was observed that there are

Discussion
This study represents the first comprehensive research on the distribution of U. crassus in Serbian waters, based on historical, literature and field data.
Considering all the collected data, it can be observed that the distribution and abundance of U. crassus varies in the different study periods (Figures 2 and 4A,B) and in different watercourses (Figure 5).Re-identification of the archive samples from the Serbian Natural History Museum (period from 1953 to 1973) revealed that U. crassus was a common species with continuous distribution throughout Serbia until the mid-1970s.After re-identification (Table 1) of the museum collection, it was observed that there are

Discussion
This study represents the first comprehensive research on the distribution of U. crassus in Serbian waters, based on historical, literature and field data.
Considering all the collected data, it can be observed that the distribution and abundance of U. crassus varies in the different study periods (Figures 2 and 4A,B) and in different watercourses (Figure 5).Re-identification of the archive samples from the Serbian Natural History Museum (period from 1953 to 1973) revealed that U. crassus was a common species with continuous distribution throughout Serbia until the mid-1970s.After re-identification (Table 1) of the museum collection, it was observed that there are many synonyms for U. crassus.Many of these taxa were first described by Henri Drouët (French 'Ecole Nouvelle) in his conchological study of the unionids of Serbia, which also contains an overview of the systematics of only the unionids of Serbia [48].The great intra-species morphological variability led to an expansion of species' descriptions in the XIX century.Based on all available data on unionids in Serbia and considering the species names according to the valid taxonomy, it can be concluded that the largest number of synonyms exists for the species U. crassus.European mussel diversity was also overestimated in the early 1900s due to unreliable taxonomic identification and numerous synonyms, mainly due to the influence of the French 'Ecole Nouvelle [19].The number of described species in Europe was up to 1500 in the XIX century, but currently 16 species of Unionida are recognized after many synonymies were resolved [19].
Subsequent studies (after the 1970s) showed a decline in population density and a restriction of the distribution range, as well as sporadic findings of the species.In fact, until the early 1990s, there was no data on U. crassus in Serbia.Later, the occurrence of U. crassus was reported for the Kolubara River in the period 1991-1994 [24], for the Pustareka in the period 1998-1999 [25] and for the Crni Timok River [37].U. crassus was detected in Serbian waters in the Crni Timok in 2004 [37].Martinović-Vitanović et al. [26] reported findings at two sites in the Serbian stretch of the Danube (Stari Banovci and Smederevo) in the period 2003-2008.All of the above-mentioned findings could be characterized as rare and/or individual findings, indicating that the species was present, but with low population density.
More recent investigations (2009-2019) confirmed the presence of the species in the Danube, Tisa, Sava, Velika and Zapadna Morava Rivers, and in the Kolubara River basin (three sites on the main course of the Kolubara and in the Peštan and Ljig Rivers) (IBISS database), as well as in the Južna Morava [39] and Nišava Rivers [40] (Table 2, Figure 4B).According to the results of the survey of the Sava River in 2012 [28,29], and especially in 2019, a stable population of U. crassus was found in the upper and middle stretches of the river.The species was detected at all investigated sites from the site Jamena (204 rkm) to the mouth of the Danube (3 rkm).During the 2019 survey, a high abundance of the species and an almost uniform population of U. crassus was detected at the Jamena site.A similar distribution pattern of the species was observed in the KolubaraRiver basin and in the Velika Morava (from 179 rkm to 2 rkm) and Zapadna Morava (from 172 rkm to 2 rkm) Rivers.According to recently published data, an extension of the known range of the species in Serbia and its occurrence in the Južna Morava [28] and Nišava [32] Rivers was also detected.During the investigation of the Tisa River in 2001 [23], the presence of U. crassus was detected at one (Novi Bečej-63 rkm) of four investigated sites in the Serbian stretch (lower Tisa), while subsequent surveys from 2010 to 2019 showed an increasing population trend with detection of the species along almost the entire Serbian river section, from the Martonoš (155 rkm) to the Titel (11 rkm) River.Considering that the presence of the species has been confirmed in repeated sampling with significant abundance in the Velika Morava and especially in the Sava River (Figure 5), it could be assumed that the population is recovering, but stable populations are still localized.Furthermore, the permanent finding in the Kolubara River basin could indicate either a recovery of the population or that the population has reached its optimal density for that particular river type.The decreasing trend of the population and the fragmented distribution of U. crassus were also confirmed for Europe during studies in the second half of the XX century, with the exception of the northern part (the Baltic basin area), where the species is still considered to be relatively widespread [14].In contrast to the current data, U. crassus was also formerly widespread and the most common unionid in Europe [14].
Knowledge of the habitat requirements of endangered species is of great importance for the implementation of effective conservation strategies, which usually include habitat restoration [49].In this study, the species was registered in different waterbody types (Types 1, 2 and 3) in areas up to 350 m a.s.l.(Figure 6), in the littoral part of rivers, mostly in fine substrate but also on larger sediment fractions.Most European unionids are lowland species, whereas U. crassus can inhabit higher elevations than other unionids [19,30] and can even reach very high densities in mountainous rivers [50], which supports the hypothesis of a wider niche for habitat variables than expected [51].
The population decline and local extinction may be related to the general environmental degradation due to pollution and habitat degradation in the second half of the 20thcentury.It was observed that this species is generally vulnerable to environmental degradation, especially to changes in water chemistry [14].The high level of eutrophication caused by agricultural drainage is considered to be the main reason for the decline of U. crassus [10,22,30,[52][53][54].Our data show that the mean nitrate-nitrogen concentration varies between the study periods (Figures 7 and 8).The maximum variation in nitrogen concentration indicates highly polluted rivers and poor water quality conditions in almost all the studied rivers in the period between 1999 and 2007 (Figure 7).In the recent period (2011-2019), an improvement in water quality in terms of nitrogen concentration was observed (Figure 8).The most favorable conditions are in the Sava River where the maximum values do not exceed 1.5 NO 3 -N mg/L (Figure 8).A significant improvement in water quality was observed in all the studied rivers, which is consistent with our most recent investigations.According to the latest studies, the distribution, the number of detections in selected watercourses and the localities of the first findings, clearly indicate an increasing population trend and an expansion of the distribution range in recent years in Serbia, with a focus on the Sava River basin (Figures 4B and 5).According to research by Zettler and Jueg [22], the increased nitrate-nitrogen caused by eutrophication is one of the main factors in the decline of U. crassus.In particular, it is a limiting factor for the growth and maturation of juveniles.A prevailing concentration below 2 mg/L throughout the year and between years indicates successful growth [22].According to the same authors, limited recruitment of juveniles was observed in moderately polluted streams with nitrogen concentrations between 2 and 10 mg.When the nitrogen concentration exceeds 20 mg/L, the mortality of the mature U. crassus population strongly increases [22].Increased mortality was observed in juveniles above concentrations of 2.3 mg NO 3 -N/l [51].
U. crassus was common in the Danube River during the period 1953 to 1973 [37,38], but according to our recent data, its presence in the river was detected only in 2013, with low abundance at only two sites ( Čerević-1273 rkm and Tekija-956 rkm) (Table 2).In addition to pollution, the disappearance of this species from the Danube in recent decades could also be related to the hydromorphological changes caused by the construction of dams (Iron Gate) and their impact on the river.Dam construction is probably one of the major threats to the mussel community with direct (damage or removal) or indirect effects on mussels (loss of suitable mussel substrate and decline of host fish) [19,22].The construction of dams creates barriers to the migration of fish that are potential obligate hosts for the unionid larvae.A lack of suitable host fish can lead to a lack of juvenile recruitment, reducing population density and can potentially lead to species disappearance from habitats or even to extinction [22,55].The construction of the dam and the forming of a large accumulation lake on the Danube River in Serbia has led to changes in the natural river regime, i.e., the slowing down of the river flow and permanent sediment deposition [56].Although the dam was built in the lower section, the changes in the river character are noticeable over a long distance downstream and also upstream of the dam.The change in the flow velocity of the river has led to an increased sedimentation rate in the Danube [57].The increase in sedimentation rate and the change in substrate as a result of the dam [22,58] indirectly affects the mussel community by affecting the potential microhabitats of the species.Changes in river flow due to dam construction and their impact on mussel fauna have already been confirmed for streams and rivers in Europe [14,19,22,59].
On the territory of Serbia, the beginning of mussel exploitation dates back to the 1930s.In the 1950s, organized mussel collection for industrial purposes was performed [38].This long-term overexploitation has certainly significantly contributed to the decline of mussel populations in our rivers, which can still be observed today.Since the 1850s, freshwater mussels have been exploited for the extraction of pearls and nacre for button making [60].At the peak of this exploitation, up to 50,000 tons of shells were harvested from North American rivers [61].Strict laws now prohibit these activities, but poaching continues in some countries [19].According to Ferreira-Rodríguez et al. [62], overexploitation is only locally significant and is often of secondary importance compared to other pressures that currently exist.
Among other factors, the introduction of exotic species is a possible contributing factor to the decline of freshwater mussels [10].Over the past 20 years, research on allochthonous species has intensified in Serbia [61].According to Zorić et al. [63], the Danube is the main corridor for the introduction and spread of alien species in Serbia and their spread to the other major rivers, i.e., the Tisa, Sava and Velika Morava Invasive mussels are widely recognized as an important threat to native biodiversity [64].The ecological impact of invasive species on native communities is not well documented in Serbia but there is evidence of widespread distribution in Serbian waters, dense populations and coexistence with native fauna [63].They can cause direct biotic interactions with the native community (e.g., predation and competition) and also indirect changes in habitat conditions (e.g., habitat structure and turbidity) [65,66].Evidence of the negative impact of invasive species on native unionids has already been observed in many European countries as well as in North America [19].
At global, regional and local levels, species important for nature conservation are selected, protected areas are designed and an ecological network is established to link protected areas important for biodiversity conservation and the remaining priority habitat types [64].
The NATURA 2000 network is the main tool for biodiversity protection in the European Union.It is now considered to be the world's largest network of protected areas, covering 30,000 sites that occupy 20% of the EU territory [67].
Nature conservation efforts in the Republic of Serbia are aimed at fulfilling obligations in the framework of preparations for accession to the European Union (EU), which mainly refers to the establishment of the NATURA 2000 ecological network.When the conditions for EU accession are met, biodiversity and habitat diversity in Serbia will become part of the European ecological network NATURA 2000, with the obligation to implement the Directive.Serbia will propose areas important for the conservation of endangered plant and animal species for the ecological network NATURA 2000 and habitat types, as well as other EU member states.
The Balkan Peninsula served as a glacial refuge for several species of freshwater macroinvertebrates [68,69].The establishment of an ecological network of protected areas will make it possible to ensure the survival of the most valuable species and habitats, promote the protection of numerous ecosystems and ensure that the natural system of Europe, and the Balkan Peninsula in particular, remains healthy and resilient.

Conclusions
The results presented show considerable progress in the restoration of the former distribution range of U. crassus in Serbian waters.Based on a dataset that includes historical and current data, population trends of this mussel over time were identified and a better understanding of the basic ecological requirements of the species was gained.The interaction of eutrophication, hydrological changes, overexploitation as well as the introduction of invasive species may be possible factors that influenced the local disappearance of U. crassus in some sections or the decrease in population density in Serbia.The results of this study can be used for the further development of effective and sustainable conservation strategies for endangered U. crassus populations, which usually include habitat restoration.Despite the high conservation status of this species, knowledge about its biology and ecology is insufficient.To improve conservation strategies for U. crassus, a systematic understanding of the limiting factors in the species' life cycle is crucial.Further studies on U. crassus should include more comprehensive ecological, biological and genetic investigations, as well as detection of new populations on a larger geographical scale.

Figure 1 .
Figure 1.Map of all observed locations of U. crassus in watercourses in Serbia in the period (1973-2019).Figure 1. Map of all observed locations of U. crassus in watercourses in Serbia in the period (1973-2019).

Figure 1 .
Figure 1.Map of all observed locations of U. crassus in watercourses in Serbia in the period (1973-2019).Figure 1. Map of all observed locations of U. crassus in watercourses in Serbia in the period (1973-2019).

Figure 2 .
Figure 2. Map of the distribution range of U. crassus in studied grid squares 10 × 10 km in the period 1953-1973.

Figure 2 .
Figure 2. Map of the distribution range of U. crassus in studied grid squares 10 × 10 km in the period 1953-1973.

Figure 3 .
Figure 3.The number of mapping fields in different time periods.

Figure 3 .
Figure 3.The number of mapping fields in different time periods.

Figure 7 .
Figure 7. Nitrate-nitrogen concentration in rivers currently or formerly populated by U. crassus in period 1999-2007.Mean values are shown as columns with minimum and maximum deviation indicated by lines.D-Danube; S-Sava; T-Tisa; VM-Velika Morava; ZM-Zapadna Morava; K-Kolubara Rivers.

Figure 8 .
Figure 8. Nitrate-nitrogen concentration in rivers currently or formerly populated by U. crassus in period 2011-2019.Mean values are shown as columns with minimum and maximum deviation indicated by lines.D-Danube; S-Sava; T-Tisa; VM-Velika Morava; ZM-Zapadna Morava; K-KolubaraRivers.

Figure 7 .
Figure 7. Nitrate-nitrogen concentration in rivers currently or formerly populated by U. crassus in period 1999-2007.Mean values are shown as columns with minimum and maximum deviation indicated by lines.D-Danube; S-Sava; T-Tisa; VM-Velika Morava; ZM-Zapadna Morava; K-Kolubara Rivers.

Figure 7 .
Figure 7. Nitrate-nitrogen concentration in rivers currently or formerly populated by U. crassus in period 1999-2007.Mean values are shown as columns with minimum and maximum deviation indicated by lines.D-Danube; S-Sava; T-Tisa; VM-Velika Morava; ZM-Zapadna Morava; K-Kolubara Rivers.

Figure 8 .
Figure 8. Nitrate-nitrogen concentration in rivers currently or formerly populated by U. crassus in period 2011-2019.Mean values are shown as columns with minimum and maximum deviation indicated by lines.D-Danube; S-Sava; T-Tisa; VM-Velika Morava; ZM-Zapadna Morava; K-KolubaraRivers.

Figure 8 .
Figure 8. Nitrate-nitrogen concentration in rivers currently or formerly populated by U. crassus in period 2011-2019.Mean values are shown as columns with minimum and maximum deviation indicated by lines.D-Danube; S-Sava; T-Tisa; VM-Velika Morava; ZM-Zapadna Morava; K-KolubaraRivers.

Table
. The museum

Table 1 .
Re-idetification of the collection from the Natural History Museum in Belgrade (collector Ante Tadić).

Table 2 .
Findings of U. crassus in the period 1990-2019, according to the literature data and field investigation.

Table 2 .
Findings of U. crassus in the period 1990-2019, according to the literature data and field investigation.