The distribution of Microcondylaea bonellii (Bivalvia: Unionidae) in Slovenia

The freshwater bivalve Microcondylaea bonellii is classified as Vulnerable according to the last IUCN Red List assessment and also listed in Annex V of the Habitats Directive. According to more than one hundred years old data, the species was present in Slovenia in a restricted section of the Vipava River and its tributary the Lijak Stream, which is also the species’ type location. We surveyed all previously known as well as new potential localities for M. bonellii. Live mussels were found only in the lower part of the Vipava River, but not in the Lijak Stream. Nevertheless, the population in Slovenia is isolated from other populations and restricted only to 16 km of the Vipava River. Unfortunately, key sections of the Vipava River have been subjected to habitat destruction in the past and are threatened by ongoing engineering works and developments. Thus the legislation should be taken into strict consideration immediately if we want to preserve this protected species and its habitat.


Introduction
Freshwater bivalves of the order Unionida, also known as freshwater mussels, are represented in Europe by at least 16 recognized species with several subspecies grouped into two families Margaritiferidae and Unionidae (Cuttelod et al. 2011, Lopes-Lima et al. 2017. Some species are widely distributed, while others have a relatively limited range.  (Oliverio et al. 2016). Isolated populations have also been recorded in the Soča (= Isonzo) River Basin in Italy (Nagel & Hoffmeister 1986, Cencetti & Castagnolo 1997, Nagel et al. 2007) and in the Mirna River on the Istrian Peninsula in Croatia (Fischer 1999, Fischer & Reischütz 1999, Reischütz & Reischütz 2002, Mrkvicka 2018. In the last ten years new populations have been discovered in Albania (Reischütz et al. 2008(Reischütz et al. , 2014. It is considered to be extinct in Switzerland (Rüetschi et al. 2012).
Freshwater mussels, particularly Unionidae and Margaritiferidae, rank among the most endangered organisms in freshwater ecosystems and have experienced a global decline in species richness, distribution and abundance (Lopes-Lima et al. 2017). M. bonellii has undergone a serious decline (Nagel et al. 2007). Until the beginning of the last century it was considered relatively abundant. As in many other freshwater mussel species, the drastic decline of this species is probably due to human impacts, among which habitat destruction and water pollution are the most important factors. Using the last IUCN Red List assessment, M. bonellii is recognized as Vulnerable (Albrecht et al. 2011). It has been assessed as Vulnerable in the European Red List of non-marine Molluscs (Cuttelod et al. 2011). M. bonellii is also listed under the name M. compressa in Annex V of the Habitats Directive (OJ 1992) and protected under the Bern Convention (Appendix III -Protected fauna species). In Slovenia, both the species and their habitats are protected (Ur. l. RS 2004). After adoption of the Habitats Directive in 1992, intensive research activity on species of conservation concern was spurred in many European countries. However, the emphasis was on Annex II and some Annex IV species but unfortunately not for Annex V species. Therefore a knowledge gap has arisen for the ecology and even the basic distribution of many species, also for M. bonellii.
Even the host fishes for M. bonellii glochidia remain unclear (Nagel et al. 2007, Lopes Lima et al. 2017. No research and monitoring of most Annex V species in Slovenia has been carried out and, in turn, no new data have been collected so far. The ongoing unknown status (XX) of M. bonellii was reported under Article 17 of the Habitats Directive in (ZRSVN 2007. Article 11 of the Habitats Directive requires Member States to monitor the habitats and species listed in the Annexes (habitats in the Annex I and species in the Annexes II, IV and V), while Article 17 requires a report to be sent to the European Commission every 6 years. As there are no published and up-to-date data on the species in Slovenia and old literature is commonly overlooked, there is also a gap in the distribution maps in the last review of European Unionidae (Lopes-Lima et al. 2017). At the same time, many recent papers about M. bonellii just mention its presence in Slovenia (Fischer 1999, Fischer & Reischütz 1999, Bössneck 2002, Nagel et al. 2007, Rüetschi et al. 2012, Lapini et al. 2013), but none of them bring new data or proof that this species is still present in Slovenia.
In this paper we present recent data on the distribution of M. bonellii in Slovenia.

Materials and methods
The Vipava River is a 49 km long left tributary of the Soča River, with its mouth in the northern Adriatic Sea. The Vipava River spring system is located at an elevation of 98 m a.s.l. in Vipava town in the western foothills of the Nanos massive. It drains a 598 km 2 catchment (Brenčič 2013, Monegato et al. 2015. In Miren, the Vipava River leaves Slovenia and enters Italy, where it discharges into the Soča River after 5 km, at an elevation of 30 m. The average discharge in Miren is 16.5 m 3 /s, the minimum can be less than 2 m 3 /s and the maximum 341 m 3 /s (Brenčič 2013). At Miren it reaches a width of 30 m. The Vipava River has a Dinaric pluvio-nival regime, in which spring and autumn peaks are fairly similar, while differences between winter highs and summer lows are pronounced (Pavlič & Brenčič 2010). The Vipava River spring is a typical karst, water-rich spring, with a stable average water temperature of 9.5°C which oscillates between 8.5°C and 10°C (Pavlič & Brenčič 2010). In Miren, water temperatures rarely fall below 4°C in the winter and can reach up to 26°C in the summer. Close to Renče, the Lijak Stream flows into the Vipava River. The maximum discharge close to the confluence with the Vipava River is 70 m 3 /s and 2 m 3 /s on average. Water temperature in the Lijak Stream rises up to 20°C in the summer and falls below 4°C in the winter.
Bivalve surveys were conducted between 2007 and 2018 during low water levels. In total, 23 localities were surveyed. Three of these localities were revisited (Tab. 1). Localities were selected, based on the known and predicted distribution of the species. We surveyed shallow water patches to a maximum water depth of 1 m. We combined methods of hand collection, hand netting and the use of surface bathiscope. To spot mussels or their siphons, especially on the gravel sediment, a surface bathiscope was used, while hand netting was performed only in fine river sediments. We also systematically searched for empty shells at gravel bars.
As our focus was only the distribution and not quantitative surveys of M. bonellii, we didn't use methods for searching completely buried mussels, especially small ones. Consequently the results are presented as presence-absence only.  (Mrkvicka 2018). This characteristic allows the identification of mussels on underwater photos, especially on web photo forums.
All bivalves were measured (length, height and width) by using vernier caliper which is accurate to the nearest 0.1 mm. Living bivalves were immediately returned to their habitat, whereas empty shells were taken from the spot and are stored in the authors' private collection. Bivalves were surveyed in accordance with a licence (35603-3/2010-4) issued by the Slovenian Environment Agency of the Ministry of the Environment and Spatial Planning of the Republic of Slovenia to the Centre for Cartography of Fauna and Flora (CKFF).
The range of water quality variables are measured by national authorities (Slovenian Environment Agency -ARSO) within a national monitoring program. We gathered data for Miren, town close to the Italian border, for the period 2007 to 2017 (ARSO 2018; Tab. 2).

Results
Live mussels or empty shells were found at all survey sites in the Vipava River downstream from the Gradišče hydropower plant (Tab. 1: ID 1-10, Fig. 1). At all positive sites for M. bonellii we also found live Unio elongatulus C. Pfeiffer, 1825. In the Lijak Stream, no live mussels or empty shells of M. bonellii were found at any of the survey sites (Tab. 1: ID 14-18, Fig. 1). Altogether, 169 live mussels were found and 287 empty shells collected.
The mean length of live bivalves was 79.3 mm (min-max: 57.7-95.1 mm), the mean length of empty shells 75.6 mm (min-max: 31.3-95.1 mm) (Fig. 2).  In the Prvačina and Renče area we were able to survey the Vipava River from bank to bank (20 m width), as well as its central part, where water was up to 1 m deep at low discharge rates. M. bonellii and U. elongatulus were present only in a 1.5 m strip from the bank with the low but present water current. Bivalves were buried in fine sand and gravel up to 1 cm diameter. In the central part, where water current was higher and consequently larger fractions of gravel were present, we didn't find any mussel of either species. Compared to U. elongatulus, M. bonellii was totally buried into the sediment and thus only siphonal openings were visible. In the Miren area, we were only able to survey a 5 m strip from the bank due to deep water in the main river channel. In a patch, where a living M. bonellii was found, evident water current was observed. Bivalves were buried between stones of the same size as was their own. Right next to the riverbank in fine sand with low flow, M. bonellii was very rarely recorded, while U. elongatulus prevailed. This was the opposite of the survey results for the Prvačina and Renče areas.

Discussion
We confirmed the presence of M. bonellii in the Vipava River main channel downstream of the Gradišče hydropower plant. 120 years ago, Gallenstein (1894) wondered whether the species really does occur further upstream of the confluence with the Lijak Stream. He concluded that more detailed research was required. We were able to find it close to Prvačina town, 3 km upstream from the confluence of the Lijak Stream, 2 km downstream from the Gradišče hydropower plant. Live mussels were also observed in the Italian part of the Vipava River (Reischütz & Reischütz 2002, Lapini et al. 2013). In the Lijak Stream, dozens of living U. elongatulus were found, but no living M. bonellii. Neither did we find any empty shells, although they had been found by Reischütz & Reischütz (2002). The downstream section of the Lijak Stream was very difficult to survey due to its steep river banks and deep water. In the Vipava River basin, the species' range is limited to 20 km of the Vipava River from the Gradišče hydropower plant downstream to its confluence with the Soča River, 16 km of those in Slovenia. The hydropower plant was built in 1922 (Brenčič 2013) and is poorly or even totally impassable for upstream fish migration. Thus it has probably been limiting the upstream distribution of M. bonellii for already a century. Additionally, small weirs were built on the Vipava River for mills and sawmills. Today at least 4 weirs over 1.5 m in height exist in the Vipava River downstream from Gradišče. All weirs are poorly or even totally impassable for fishes. Consequently the habitat of M. bonellii and probably populations are fragmented.
Oxbow lakes of the Vipava River, where Erjavec (1877) found the species, weren't surveyed, since most of them don't exist anymore. Although bivalves can be totally buried into the sediment (Nagel et al. 2007), we believe our results of the species range are representative, since we combined the method of surveying both live mussels and empty shells.
Correspondingly we also found U. elongatulus at some locations where M. bonellii was not present. We can conclude that the range of M. bonellii in the Vipava River is similar to the previously described situation. The statement by Reischütz & Reischütz (2002) that M. bonellii is extinct in the Lijak Stream needs further focused research. In streams where mussels might be present in very low densities, other methods (e.g. sieving sediment) on more patches should be performed.  (Nagel & Hoffmeister 1986). In the upper regulated section of the Birša Stream (Slovenian name for the Versa Stream) in Slovenia, no mussels were found. The Birša Stream partly dries up before the national border and shows signs of excessive nutrient intake. At the bottom of the stream fine sediments are present as a result of runoff from adjacent agriculture areas, with the signs of anoxia. South of Slovenia, the first known population of M. bonellii lives in the Mirna River on the Istrian Peninsula in Croatia (Fischer 1999, Fischer & Reischütz 1999, Mrkvicka 2018. The IUCN range map (Albrecht et al. 2011) shows a totally inaccurate range in the region. The Vipava River in Slovenia, the Versa Stream in Italy and the Mirna River in Croatia are excluded, while the Dragonja and Reka River basin are included as an area of distribution (Albrecht et al. 2011). Consequently, references which use IUCN range maps are incorrect (e.g. Darwall et al. 2014). There is no evidence that this species is present in the catchment area of the Dragonja and the Reka Rivers in Slovenia. A record for the Reka River from Slapnik (2005) is doubtful and should be omitted, as Slapnik (2013) mentioned the species' distribution only for the Vipava River basin. In addition, our findings of individual bivalves from the genus Anodonta in the Reka River (Govedič, unpublished), which is not mentioned in Slapnik (2005), is another reason to omit the record from the Reka River.
The drastic decline of M. bonellii in Europe is probably caused by human impacts, among which habitat destruction and water pollution are the most important factors (Nagel et al. 2007). Nitrogen compounds, particularly ammonia, and temperature are suspected major stressors for these aquatic organisms (Beggel et al. 2017). Studies from Central Europe have shown a relationship between the impaired population status of threatened freshwater mussel species and elevated nitrate (NO3) concentrations in running waters (Douda 2010). The quality of water in the Vipava River entirely depends on its catchment area. Thus the protection of M. bonellii in the Vipava River entirely depends on Slovenia since the complete catchment area is geographically within the country. Most of the flood plain of the Vipava River is used for intensive agricultural purposes. After World War II, the Vipava Valley was subjected to extensive agromeliorations and regulation of tributaries (Brenčič 2013). Before, mostly meadows and pastures existed along the Vipava River, which were regularly flooded. The increasing intensification of agriculture in the Vipava Valley can lead to elevated concentrations of nitrogen in the water and, due to the use of phytopharmaceuticals, also some other chemicals. The Vipava River receives all wastewater from the entire valley. There is a population of at least 65,000 inhabitants in the wider drainage area (Kladnik 2013). The probability of accidental pollution incidents increased after the change of discharge from the waste water treatment plant of the town of Nova Gorica, which had been redirected a few years ago to the Vrtojbica Stream (a tributary of the Vipava River upstream from Miren).
There is no information on water quality requirements for M. bonellii habitat, so we can't compare the results of water quality of the Vipava River at Miren with others studies. Average concentrations of nitrate (6.6 mg/L NO3) in the Vipava River that corresponds to 1.5 mg/L of nitrate nitrogen (N-NO3), is in range of the reported value for high abundance and reproduction of U. crassus (Douda 2010, Köhler 2006, Zając et al. 2018, Denic et al. 2014, Zettler & Jueg 2007. There can be some misunderstanding while some research present results as nitrate (NO3) values (Zając et al. 2018), others (Douda 2010, Denic et al. 2014, Zettler & Jueg 2007) as nitrate nitrogen (N-NO3). Only Köhler (2006) pays attention to both values. Furthermore, many other chemicals of which impact has not been studied and which are not regularly monitored can have negative effects on bivalves. Knowledge on chemicals that are produced and released to the Lijak Stream from sediments of accumulation in Lake Vogršček is also lacking. Also human activity in Goriška Brda in Slovenia in the catchment area of Versa Stream can potentially affect the quality of water at M. bonellii populations downstream in Italy. Goriška Brda is an agricultural region with intensive vineyards and orchards. Population density in Goriška Brda is high (69 inhabitants/km 2 ) as well (Glavan 2011). Due to the intensification of agriculture in Goriška Brda, special attention on water quality should be paid, as well as on quantity. Especially in summer, the stream should be regularly monitored with the purpose to detect early threats to the M. bonellii populations downstream in Italy.
In terms of size structure smaller live mussels are missing. The reason for this is most probably the hand-searching surveying method, which is insufficient for identifying the presence of small mussels. Digging in the substratum to detect small specimens was not performed. The biggest live specimen and empty shell measured 95 mm in length. This is comparable with related studies (max. 91 mm in Bössneck (2002), max. 81 mm in Fischer & Reischütz (1999)). The biggest one ever found measured 102 mm in length (Erjavec 1877).
M. bonellii can live up to ten years, but negative environmental impacts, specifically water pollution, shorten their life expectancy (Fischer & Reischütz 1999). According to the length structure of M. bonellii in the Vipava River they still reach the expected length and age. From the size structure of living specimens and from empty shells in the years 2015 and 2018, we can conclude that they still successfully breed in the Vipava River. But the alarming fact is that we found large amounts of empty shells at Miren, where they were smaller than 65 mm. This size corresponds to the age of 4 years (Nagel & Hoffmeister 1986) and indicates the mortality of mussels before they reach their final size. We don't know what is causing this, but the quality of water should be monitored more thoroughly as is provided under the national monitoring program, especially during summer flow minimums.
The number of specimens we found at each locality is related to the surface of shallow habitat, which was accessible for survey and with time spent for searching. Our goal was not to assess the density of the species, but to confirm its presence. It seems that, as in the Versa Stream, M. bonellii in Vipava River also prefers sections with a distinctive flow. In the main flow section of the Versa Stream, 25 mussels/m 2 were observed, but higher densities were also observed in pools and among the roots of plants on shallow banks (Nagel et al. 2007). In the Mirna River, mussels were observed in high densities up to 15 mussels/m 2 in shallow waters of eroded banks with coves and sandy substrate or accumulated fine substrate but not in the straight parts of the riverbed with a stronger current and gravel substrate (Mrkvicka 2018). According to Rüetschi et al. (2012), M. bonellii can also live in lakes and in slow flowing streams with sandy banks. Erjavec (1877) found it in oxbow lakes of the Vipava River. Thus it is questionable whether the mussel lives also in deep sections of the Vipava River and its oxbow lakes, according to the fact that it wasn't found in fine sediment next to the riverbank with low flows. Most of the lower part of the Vipava River is between weirs, where water is deep and its riverbanks are steep and consequently inaccessible for conventional survey methods. It would only be possible to check deeper sections of the Vipava River with the help of divers or underwater cameras. Thus the overall population size of M. bonellii in Slovenia remains unknown. If mussels prefer only shallow parts with a distinctive flow, then the population will be very patchy scattered as a result of the arrangement of a suitable substrate.
We still consider M. bonellii in Slovenia to be a poorly known and endangered species. New records give the impression that the overall population has increased in the Vipava River, but they are probably only rediscoveries and a confirmation of the former range of the species. Its presence in deeper parts of the Vipava River between weirs should be surveyed as soon as possible as well as in the Lijak Stream. Distribution and population size of the species in the Vipava River remain unknown. Ongoing unknown status (XX) of M. bonellii needs to be reported under Article 17 of the Habitats Directive in the next reporting period in 2019. The Vipava River and Lijak Stream underwent large scale habitat destruction in the previous century, but local construction works are still in progress. In order to maintain high flows within the riverbed and to prevent flooding, the habitat is changing because of higher riverbed shear stress during high water discharges. After adoption of the Habitats Directive in 2004, there were few construction interventions in the Vipava River, without confirmation of the species at the site of intervention. Locally eroded banks are replaced by blocks of stone and small locally shallow parts are destroyed. Many suitable habitats and eventually existing mussels' populations were probably destroyed. At least the exact distribution of protected species should have been studied before. In the future all construction works, which can cause the death of mussels or impair their habitat, shouldn't be carried out in the Vipava River or the Lijak Stream. Further comprehensive studies focusing particularly on the ecology and habitat requirements are warranted to better understand the conservation status of the species. Also weirs should be passable for all fish species and not only for spring spawning migrations, also all year around, especially during the development of glochidia and at the end of the parasitic phase of bivalves. Reischütz & Reischütz (2002) evaluated that the conservation status of M. bonellii in Slovenia should be altered to CR (Critically Endangered).
We fully agree with this assessment, since the existence of the species is limited to 16 km of the highly fragmented Vipava River in Slovenia and is isolated from other populations.