Materials & Methods
Fish pass
As part of the LIFE project ”LIFE+10 NAT/AT/016 - Netzwerk Donau”, a near-natural bypass river was constructed at the Ottensheim-Wilhering power plant at the Austrian Danube between 2015 and 2016. This fish pass, stretching 14.2 km and designed to resemble a small river, has a discharge capacity of up to 20 m³/s, with a bed width varying between 10 and 25 meters. It was integrated into the existing Aschach and Innbach river systems, which were also restored within the project, thereby reconnecting the Danube with its tributaries (Figure 1). The bypass river enables fish to overcome a height difference of about 12 m, while simultaneously providing key habitats, now largely missing in the Danube, such as shallow shorelines & gravel banks, riffles and pools. The discharge in the bypass system follows that of the Danube, an upstream inlet structure ensures that the required volumes of water are maintained under varying flow conditions in the Danube.
During the construction of the bypass river, the focus extended beyond merely ensuring the passability of the hydropower plant (HPP) dam. A significant emphasis was placed on creating a natural gradient and restoring an original, furcating river landscape with its characteristic flow patterns and substrate conditions. The bypass system thus improves the habitat quality and availability, especially for rheophilic fish species of the Danube. Fish entering the bypass river from either upstream or downstream of the HPP can now utilize the restored sections of the Aschach and Innbach rivers. Additionally, the newly created sections of the bypass river provide migration routes and habitats for spawning, feeding, and juvenile nurseries.
Pit Tags & Antenna arrays
PIT tags (Passive Integrated Transponders) have been used for the individual tagging of fish since 1987 (McKenzie et al., 2012). As they are battery-free passive transmitters, they have an unlimited service life, allowing for the unique identification of individual fish throughout their entire lifespan (Connolly et al., 2008). This makes PIT tags particularly suitable for tracking fish individually over extended periods, ideal for large-scale studies (Greenberg & Giller, 2000; Roussel et al., 2004). For the investigation at the Ottensheim-Wilhering fish pass we used the FDX system from Biomark®, which is known for its high read rate of 30 scans per second, which enhances detection efficiency, especially for shoal-migrating species such as the potamodromous nase (Chondrostoma nasus ) (Nagel et al., 2023). The tags (12.5 mm x 2.1 mm) were carefully implanted into the fish using syringes.
The PIT tags were read automatically and contactless using five permanently installed antenna arrays (Figure 1). These antennas were placed at strategic points in the bypass system: antenna #1 at the upper end and antenna #5 at the lower end of the bypass, covering the entire width of the watercourse. Antenna #2 was positioned in the upstream ramp connecting the bypass system with the Danube, approximately 1 km downstream of antenna #1, Antenna #3 was installed in the Aschach River, 125 m upstream of its confluence with the bypass channel, and antenna #4 in the Innbach River, 430 m upstream of its confluence with the bypass.
We analysed the antenna array data, including the date, time, and location (antenna #) of each fish detection from 18.11.2020 to 13.06.2024 to study the downstream migration and habitat use of the common nase (Chondrostoma nasus ). Due to technical difficulties, Antenna #1 was not operational between 23.06.2023 and 01.02.2024.
To calculate the detection efficiency of the PIT Tag antenna arrays, it is necessary to know both the number of detected fish on an array and the number of fish present above or below this array, requiring data from at least two antenna arrays. The detection efficiency can then be determined using the method outlined by Zydlewski et al. (2006):
E1 = Dcommon to 1+2 / (Dunique to 2 + Dcommon to 1+2)
E1 = Detection efficiency of antenna array #1
D = number of detected tags
The assumptions or necessary conditions for the calculation are as follows:
• The probability of detection on the first array is independent of the probability of detection on the second array
• The fish detected at the first array continues to swim in the direction of the next array
Detection efficiency can be calculated individually for Antennas #1 and #2. However, with additional arrays (#3-5) available to act as a secondary antenna array, there are alternative methods to calculate the overall efficiency.
Two different efficiencies were calculated: (i) the probability of detection during a single downstream passage over the Antennas #1 or #2, and (ii) the probability of being detected on Antennas #1 or #2 throughout the entire study period (2021-2024).
Binomial confidence intervals (CIs, 95%) for detection efficiencies were determined using the normal approximation or Wald method:
CI95 = 1.96 * SQRT((E1*(1-E1)/N), with N = Dunique to 2+Dcommon to 1+2.
We used the chi-square test (χ²) to test whether differences in detection efficiency were related to the antenna arrays used in the calculation.
Fish
In our study, we used the common nase, Chondrostoma nasus , a potamodromous and highly specialized fish with distinct ecological and ontogenetic trophic niches (Reckendorfer et al. 2001, Hudson et al, 2014, Ovidio & Nzau Matondo, 2024). This species served as a surrogate for other common rheophilic fish species of the Austrian Danube such as Barbel (Barbus barbus ) or ide (Leuciscus idus ). Potamodromous fishes exhibit a cyclic sequence of migrations for feeding, refuge and reproduction across four habitat types (feeding, winter refuge, non-winter refuge and spawning) and show habitat-use patterns integrated with their life stages and body sizes (Thurow, 2016, Northcote 1984; Keckeis et al., 1997). In this section of the Austrian Danube, nase is a dominant species (“Leitart”) and was very abundant until the beginning of the 20th century. Populations declined sharply in the 20th century, but in recent years a recovery has also been observed in restored sections of the Danube (Zauner et al., 2015). In spring, nase migrate to their spawning grounds (Penaz 1996), with spawning occurring between mid-March and late April at water temperatures from 9.6 to 10.8° C (Huber & Kirchhofer 1998; Melcher & Schmutz 2010). Suitable spawning habitats are shallow, fast flowing and gravelly streams (Panchan et al., 2022, Hauer et al., 2007, Kamler & Keckeis 2000). The larvae and juveniles of nase require different habitats than adults, favouring shallow, slow-flowing areas along the river shore (Keckeis et al. 1997, Kamler & Keckeis 2000, Penaz, 1996).
To investigate the migration through the fish pass and the movements within it, we captured a total of 190 nase (Chondrostoma nasus ) in the impoundment above the HPP using boat electrofishing. The fish were tagged and subsequently released on November 18, 2020 into the Danube impoundment. Electrofishing, tagging, and handling of the fish were conducted in accordance with the Austrian Animal Welfare Guideline.
Data Analysis
The detection data were aggregated to the number of detections per day and analysed to monitor the movement patterns of individual fish over time. The primary analyses included: i) calculating the number of detections per fish ID per year to identify repeated entries; (ii) conducting a seasonal analysis of detections to determine the timing of migrations and habitat use; (iii) evaluating the duration spent in the bypass to assess its role as a habitat; and (iv) comparing detection locations to understand the utilization of different sections of the bypass.
For data collection, analysis, and presentation, we utilized MS Access®, MS Excel®, and Sigma Plot® software.
We used abacus or calendar plots to analyse the Pit Tag data to estimate the general locations of each nase. The frequency and distribution of detections over time provided insights into how often each individual was detected, how long it remained in the study area, and their movement patterns between antenna arrays. In an abacus plot, the x-axis represents time, while the y-axis lists tagged fish on individual lines, with detections displayed as dots.