Characterization of the reproductive strategy of invasive round goby (Neogobius melanostomus) in the Upper Danube River

Originating from the Black and Caspian Seas, the round goby (Neogobius melanostomus) has become one of the most successful invaders of freshwater ecosystems. This dataset was created to provide a characterization of the reproductive strategy of an established round goby population in the Upper Danube River including sex ratio, fluctuations of gonadosomatic index (GSI), analysis of timing of spawning as well as of clutch and egg size. Data was collected for 1.5 years and time-series analysis was used to examine the influence of temperature, photoperiod and discharge on the reproductive strategy of round gobies. Further, clutch and egg sizes of round gobies as well as sex ratio of the population was analyzed.

Methods
Study area: The Danube is the second longest river with the second largest catchment area in Europe.The studied population of round gobies was identified based on previous populations screenings across the Upper Danube River and was located between Bad Abbach (48°56'22.9"N 12°02'18.3"E) and Saal an der Donau (48°54'35.9"N 11°55'27.6"E), Bavaria, Germany. The sampling site was located in the navigable part of the River, where river banks consisted mainly of artificial rip-rap structures without canopy cover.

Field sampling: Round gobies were collected from April 2022 until November 2023 twice a month by electrofishing or baited fish traps at three locations between Bad Abbach and Saal an der Donau, Bavaria, Germany. In total, eight electrofishing campaigns between June 2022 and June 2023 were conducted along the rip-rap structures at the shoreline of the River. Three 30m-replicate stretches were sampled at each location during electrofishing.  On all other sampling dates, baited fish traps were placed along the same stretches in ~0.4-1.5m depth for 24h per sampling date (35 sampling dates in total). After collection, all gobies were euthanized using an overdose of the anaesthetic MS-222 (Tricaine), while all native fish were carefully returned to the River. To avoid degradation of tissue, gobies were transported to the laboratory (Aquatic Systems Biology, Technical University of Munich) on ice in a cooler and were frozen afterwards at -20°C for further analysis. Temperature data was obtained from a nearby measuring station in Kelheim (48° 55' 6.24 N 11° 53' 11.76 E )(Gewässerkundlicher Dienst Bayern 2024). Discharge data was obtained from the same measuring station (Hochwassernachrichten Dienst Bayern 2024).  

Gonado-somatic index (GSI): A total of 1041 females (size 84mm +/- 21) and 1026 males (size 84mm +/- 21) were dissected. Before dissection, gobies were patted dry with a paper towel, total length was measured to the nearest millimeter and fish were weighed to the nearest 0.001g using precision scales (Kern ADJ 100-4). After dissection, the wet weights of the gut content and gonads were recorded. The somatic Mass (Ms) and GSI were calculated according to Brandner et al. 2018:

Ms = Mt – (Mindexed organ + Mg) with Mt =  total body mass and Mg = gut content mass

GSI = 100 x Mgonads Ms -1

Clutch size: Round goby eggs were retrieved using spawning traps consisting of a roof tile with a paving stone attached on top with cables ties. This method allowed easy detachment of stones from the roof tiles and access to the clutches without damaging them. The spawning traps were placed at the same locations where round gobies were collected and additionally in the Marina in “Saal an der Donau” (48°54'36"N 11°55'31"E) close to the uppermost stretch of goby collection. The traps were installed in a depth of ~0.4-1.5m in the Danube and ~3m in the Marina, respectively and were checked for clutches every week. If clutches were present, a translucent foil with an imprinted grid was placed over the clutch and multiple pictures were taken. The photos were used to determine the clutch area in cm² and the average number of eggs per cm². Subsequently, the number of eggs per clutch was calculated (clutch area in cm² x number of eggs per cm²). Eggs were taken from each clutch using tweezers and preserved in 96% (v/v) ethanol. Height and width of eggs were measured using a stereo microscope (Binocular Olympus SZX10, Olympus Germany GmbH) with a 20-fold-magnitude and the software cellSens (Olympus corporation).

Statistical Analysis

Sex ratio: For the analysis of sex ratio, data was pooled and analyzed per individual month. Additionally, an overall sex ratio was calculated. To test for differences between the observed sex ratio and an expected 1:1 equilibrium, a chi²-test was computed.

Clutch Size and Gonadosomatic Index (GSI): Shapiro-Wilk and Levene Test were performed to test for normal distribution and homogeneity of variances of data. In a first step, a regression of GSI with fish size was computed for both female and male round gobies to explore the relationship between the two parameters. Further, differences in GSI between fish caught by electrofishing and traps were examined to check whether both can be included in the following analysis. For the analysis of GSI, data were then pooled and differences between individual months were analyzed. Since assumptions for parametric tests were not met for GSI, non-parametric Kruskal Wallis test with Bonferroni correction for multiple testing was performed. Assumptions of normal distribution and homogeneity of variances were met for clutch and egg size. Therefore, ANOVA with post-hoc Tukey test was performed to analyze differences of clutch and egg sizes between sampling dates.

Time-series analysi: To analyze the influence of water temperature, photoperiod and discharge on GSI, the correlation between these time-series was examined using cross-correlation. As data are required to be stationary and prewhitened for cross-correlation, Kwiatkowski-Philipps-Schmidt-Shin (KPSS) test was used to check for stationarity. All data showed to be stationary, so no further differencing (elimination of a linear trend by computing differences between consecutive observations to make a time series stationary; (Shumway and Stoffer 2016)) of data was conducted. To account for the influence of long-term trends in the input (temperature, photoperiod or discharge) or the output time series (GSI), that may hide or suggest a significant correlation between the two time-series, prewhitening of data was performed. For prewhitening of data an ARIMA Model (Box and Jenkins 2013) was fitted to the input time-series (temperature, photoperiod or discharge), which was then applied to the output time-series (GSI) according to Probst et al. 2012 and the cross-correlation between the two prewhitened time-series was performed. As round gobies were sampled on 35 occasions within 625 days one lag in the time series analysis (time gap between two time series) was calculated as follows:

625 days/35 = 17.8 (≈ 18) days

All analysis were computed using the statistical software R (Ver. 4.1.1, R Development Core Team, 2009) and significance was accepted for p<0.05.