Background

The Rhine is one of the most “mainstreamed” rivers in Europe, balancing several aspects including transport, drinking water supply, cooling water, energy reservoir, as well as being used for effluent disposal, and providing leisure activities such as fishing (Figure 1). The former structure of the river itself had been dramatically changed as a result of the invasive canalisation project by Tulla at the beginning of the 19th century, affecting the entire river from Rheinfelden (Switzerland) to Rotterdam (The Netherlands) [1]. Based on historical fishing data, it can be concluded that this structural change also influenced the fish fauna of the Rhine [2–9]. However, especially after the Second World War, a drop in both fish species number and abundance was observed due to the contamination of the river as a consequence of increasing industrialisation and further urbanisation of the Rhine valley [10–12]. In the 1970s water quality dropped to its lowest levels resulting in the implementation of various Rhine restoration programs, including the reintroduction of endangered and/or temporarily extinct fish species [13–16]. One of the commitments was the installation of the world’s largest industrial sewage treatment plant (approx. 110 million m3 waste water per year, equivalent to a city of approx. 3 million inhabitants) located at the BASF SE site in Ludwigshafen am Rhein (Rhine-Neckar Metropolis Region, Germany) [12, 17]. Along with this project, electrofishing was carried out by BASF SE at regular intervals in order to investigate potential trends in fish populations close to the industrial site in Ludwigshafen [18]. Starting in 1976, the species caught have been identified, their relative abundance determined and their overall health status observed.

Figure 1
figure 1

Schematic map of Germany (a) and the Rhine-Neckar region (b) including the Rhine River site investigated (km 426 – 435). Black – aquatic areas; dark grey – industrialised area; light grey – urbanised area; white – agricultural, nonurban and forest area.

Full size image

Results

Number of fish species caught per sampling

Starting in December 1976, 6 fish species were found along the Rhine site at BASF SE Ludwigshafen (Figure 2). From 1981 until 1990 the number of fish species found varied from 8 in 1983 to 11 fish per sampling in 1985. In the next decade, a minimum of 10 fish species per sampling (1992) and a maximum of 15 fish species per sampling (1995) were found. From 2001 until 2010 the number of fish species found further increased and ranged from 12 (in 2002) to 17 (in 2006). There is still a high variation in fish species caught during the different samplings. Overall there was a statistically significant (p ≤ 0.05) increase in the species number over the observation period as proven by linear regression by time (SAS procedure Proc Reg).

Figure 2
figure 2

Number of fish species caught on each electrofishing occasion along the BASF SE site at Ludwigshafen am Rhein.

Full size image

Number of fish species found per time period

During the four time periods, the total number of fish species found increased (Figure 3). During the first decade 15 fish species were located at the BASF SE site (Table 1). During the second period up to 22 species were found. In the third period the maximum number of fish species remained the same (22). However, three further fish species were first observed in this river section and three other ones were then absent. In the most recent decade, 24 fish species including three gobidae and one cyclostoma (namely Lampetra planeri (Bloch 1784)) were found. Again, some species, such as S. trutta forma fario, S. trutta trutta, L. lota and C. taenia, were absent in this electrofishing period.

Figure 3
figure 3

Total number of fish species caught during the four sampling periods from 1976 until 2010 along the BASF SE site at Ludwigshafen am Rhein.

Full size image
Table 1 Species list and relative abundance of fish found during each of the four sampling periods at the BASF site at Ludwigshafen am Rhein
Full size table

Relative abundance and trends in the fish community

During the 34-year investigation, 5 out of 31 fish species were found at each sampling period (100 % abundance), i.e. C. carpio, P. fluviatilis, R. rutilus, S. erythrophthalmus and A. anguilla (Table 2). Fish species such as A. aspius, B. barbus, C. nasus, S. glanis and G. cernua were absent within the first observation period, but their relative abundance constantly increased to 75 (C. nasus) and 100 % (all other four species), up to 2010. For S. trutta fario, S. trutta trutta and E. lucius a decrease in species abundance could be observed during the entire catching period. All other species were either commonly found (such as A. anguilla or R. rutilus) or only found occasionally (such as L. lota or C. taenia) and thus, no specific trends in relative abundance could be determined.

Table 2 Statistically significance of increases or decreases in numbers over the entire observation period
Full size table

During the first catching period, nine endangered fish species listed in the red list of Germany and the Rhineland Palatinate were found at this site, whereas in the last decade of catches 14 fish species were found (Table 1). Only S. trutta fario, S. trutta trutta and C. taenia were absent.

Statistical evaluation as shown in Table 2 revealed a statistically significant increase for the following species: B. barbus, S. cephalus, L. leuciscus, G. cernua, C. nasus, A. aspius, L. gibbosus, and S. glanis. A statistically significant decrease was only seen for E. lucius. For all other species no statistically significant changes were observed.

Overall health status

Visual assessment of the overall health status, any external abnormalities, and infection status with ectoparasites showed no remarkable results [19]. Only sporadic nematodes infections of the airbladder of the eel in 1988 and 1989 are worth noting.

Discussion

The investigation of the fish population along the Rhine from kilometre 426 to 435 (including the BASF SE industrial site) over the past 34 years shows an overall increase in both species number and abundance. In fact, the fish biodiversity has increased during the whole observation period. Especially rheopilic species, such as A. aspius, B. barbus, S. cephalus and C. nasus, which were either absent or only rarely observed in the first sampling period, are now found regularly in the Rhine along this site. Although water flow has not changed essentially during this time, water quality has improved dramatically in recent years, thus improving the habitat of endangered fish species. In fact, both the nutrition load (such as nitrates and phosphates) and the concentration of organic contaminants decreased due to installation of sewage treatment plants (both municipal and industrial) along the Rhine [20, 21]. Furthermore, the implementation of the large industrial sewage treatment plant at BASF SE Ludwigshafen may have had a major impact on water quality.

The reduction in organic carbon load also had a positive effect on the oxygen content of the water, thus increasing the survival rate of the more oxygen-dependent fish species historically typical of this section of the Rhine (socalled barbel region). In the early 1970s, oxygen concentrations were close to the lower limit for fish survival, resulting in a reduction of the macroinvertebrate species to about one third of the original 80 species known to be native in this area [22]. So more robust fish species like R. rutilus/S. erythrophthalmus ( these species were not exactly differentiated at that time), A. brama and A. anguilla were the predominant species in this section of the Rhine. This is in line with previous observations from the late 1960s when R. rutilus and A. brama were found to be most abundant and fish species such as B. barbus, S. cephalus and S. lucioperca were rare in this area [23]. However, by the early 1990s, water quality measurements showed oxygen concentrations close to 100% saturation with corresponding NH4-N, NH3-N and P-total concentrations of < 0.5, < 4 and ≤ 20 mg/L, respectively [21]. The improvement in water quality around the BASF SE site is in line with the overall improved water quality observed at other sites along the Rhine [14, 24].

Rheophilic fish have increased in the past two decades but were low in numbers for the first 15 years [18]. In fact, taking in to account the whole time period of investigation this increase in the number of species found was highest between the first two decades (1976 – 1980 versus 1981 – 1990). Based on the historic data 48 fish species were found to be indigenous for the river Rhine, however recatchment analysis revealed that only 42 fish species were found in 1986/1987 [25]. Additionally, this number of fish species refers to the whole river section including also estuaries and bayous and is therefore not restricted to the main river section investigated [26]. This would explain why fish species such as the Coregonus spp. L. (2 species), Osmerus eperlanus L., Misgurnus fossilis L., Rhodeus amarus (Bloch), Gasterosteus aculeatus L. and Pungitius pungitius L. were listed as indigenous for the river but should not be considered to inhabit this specific area [27].

Those fish species found are representative of the Rhine, and can be found further upstream at the socalled Hochrhein (from Lake Constance to Basel, Switzerland), thus confirming the good water quality and high oxygen content [28]. Nevertheless it cannot be excluded that rheophilic fish species has recovered due to other reasons (e.g. fish passes). The wide variety of fish species caught at each sampling, however, can be explained by the method used as it is limited to the surface and river bank areas and thus strongly dependent on the given water flow conditions, which beyond water temperatures (high or low water; see Table 3) [23, 29, 30]. Despite these limitations, a clear trend in fish population dynamic can be seen during the past decades of investigation.

Table 3 Corresponding water levels and water temperatures on the month of electrofishing
Full size table

In addition to the increase in species numbers, several other observations on fish fauna have been made which are not necessarily linked to the water quality improvements in this area in particular. Restocking measures which were carried out during the past tenth of years for S. salar, S. trutta and A. anguilla might also have and direct (species affected) or an indirect impact (predator) on the river rhine fish populations [15, 31, 32]. Nevertheless, the number of both S. trutta fario and S. trutta trutta has decreased over the past 34 years and are absent nowadays. These two species, however, have always been rare as they are more likely to be found in the lotic tributaries of the river (socalled trout region) than in the Rhine itself.

Furthermore, the S. glanis, G. cernua and C. carpio are not typical of the barbel region as they are more native to lentic habitats such as lakes and the lowlands of the Rhine. The increase in their relative abundance might be explained either by the direct connection to lentic harbours and bayous or especially by the rocky structure of the river bank which provides an excellent hiding habitat for at least the S. glanis and G. cernua. In addition, since the number of catfish was low in the Rhine in the past, this species was restocked several years ago and is now to be found in great numbers. Considering other species found more or less frequently over the 34-odd years of investigation, it should be mentioned that fish usually migrate along the river but show a preference for a specific river section (as indicated by trout or barbel or bream region). Thus, it is probable that a fish species can be found in a river section where they are not usually abundant. The E. lucius, for example, was usually caught in the lentic harbour at the BASF SE site which in fact is the more appropriate habitat than the river itself und thus their relative low abundance compared to P. fluviatilis and S. lucioperca is not unexpected [27].

Although the number of fish species has risen over the past 34 years, the species list differs from the one drawn up 100 years ago, since some anadromous fish such as the allis shad Alosa alosa (L.) and common sturgeon Acipenser sturio L. have been extinct for more than 60 years due to former river contamination (at the beginning of industrialisation), barricades along the length of the river and intensive fishing [33]. For the Atlantic salmon Salmo salar L. several reimplementation programmes were established in the late 1990s to restock populations [34]. This species has still to be found during the electrofishing procedure as the method used is limited to catches along the river banks rather than the middle of the river, and the number of returning S. salar is still low [30].

However, some new invaders have been observed: the North American sunfish L. gibbosus has been found in the area since the early 1990s, originating from either sport fisheries (1st invasive period) or pet shop trading (2nd invasive period). In 2010, high population densities of another invasive group of fish, the pontocaspian gobies (Fam. Gobidae), were noted. Although the exact species has not been determined, it is very likely that they found their way into the Rhine via the Rhine-Main-Danube Canal since at least the Proterorhinus marmoratus (Pallas 1814) was found in the River Main in the late 1990s [35]. Like the invasion of Danubian/Caspian fish species also the Rhine macroinvertebrate fauna is highly affected by Danubian species [36].

Both sunfish and gobies are strong predators especially of young fish, but at least for the latter species there is little information on their possible impact on the Rhine fish community to date.

For some fish species found in the past one or two catching periods, such as the A. bipunctatus, L. leuciscus and the cyclostoma L. planeri, no definitive population trend analysis is currently available since the time period of their abundance is too short. However, as they are “native” to the barbel region it is very likely that their numbers will rise in future.

Overall, even in the highly industrialised river section at Ludwigshafen am Rhein, the Rhine fish population along with the water quality do not markedly differ from other comparable river sections, showing a good recovery over the past 34 years of investigations.

Conclusions

Overall, the water quality of the Rhine has improved since the 1970s as indicated by the number and species of fish collected by electrofishing in the period of investigation. The improvement of water quality is also enhancing the biodiversity of fishes in this river. Since the water quality has almost reached the optimum which can be expected for this river section (water quality index of about 2.0 – 2.3) the number of fish species found may likewise have reached maximum levels for this water quality. However, current trends indicate a shift towards invasive Neozoa so there might be a further biological impact on the Rhine fish population in future. Therefore, regular electrofishing is considered to be a valuable tool not only to monitor water quality but also to constantly investigate the fish populations along the BASF SE site with respect to new invaders and potentially new species.

Methods

Samples were taken by electrofishing at regular intervals between 1976 and 2010 along the left bank of the Rhine from Gate 6 of BASF SE (Rhine km 426) to Rheindürkheim (Rhine km 435, Figure 1).

Fish were caught using a small boat with equipment on board [18]. Briefly, a 500 volt (12 ampere) direct current generator was connected to an anode dip net (diameter of about 40 cm; mesh width about 1 cm) as used to catch the fish. Fishing takes places about 3 to 5 m from the western shore of the river Rhine at about 1 m water depth.

Species were determined and their local abundance was assessed semiquantitatively. Since the species found vary for each sampling, the total number found was clustered into four time periods (until 1980 [period 1], 1981 – 1990 [period 2], 1991 – 2000 [period 3], and 2001 – 2010 [period 4]) in order to investigate trends in abundance over the 34-year sampling period. All fish were also checked visually for overall health status, external abnormalities, and infection with ectoparasites.

From 1976 to 2010, 29 samplings were taken. In some years between 1976 and 1984 the fish were caught in the spring and autumn in order to determine any differences in the number of species between the two seasons. From 1988 on, electrofishing was mostly executed in autumn. Monitoring was introduced by BASF SE as a commitment to responsible care and also to measure the effects of enhanced water cleaning measures and improved treatment procedures at the site itself [37].

Statistical methods

For the binary response (species present or not present) increase or decrease in the occurrence by time was tested using SAS procedure Proc Logistic.

This analysis was done separately for each species using logistic regression [38]. The Wald Chi Square test was used to test for an increase or decrease by time. For the number of species a linear regression by time was done using the SAS procedure Proc Reg [39]. It was tested if there is an increase or decrease by time.

Authors’ information

All authors, but Katja Hempel are working at the Department of Products Safety at BASF SE Ludwigshafen as experimental and regulatory ecotoxicologists and toxicologists, respectively. Katja Hempel is now working as a Toxicologist for Abbott GmbH & CO. KG in the Department of Global Preclinical Safety at Ludwigshafen.