SEM images reveal intraspecific differences in egg surface properties of common nase ( Chondrostoma nasus L.)

Scanning electron microscopy (SEM) has been widely used to describe interspecific differences in egg quality of teleost freshwater fish, but potential intraspecific differences are poorly studied. Eggs of many rheophilic cyprinids are covered with adhesive structures such as attaching villi facilitating egg attachment at substrates of spawning grounds with high currents. Recent findings indicate that the egg quality of the rheophilic cyprinid common nase ( Chondrostoma nasus L.), a target species of conservation, differs in the adhesiveness between spawning populations, poten-tially explaining differences in recruitment success. In this study, a SEM image- based standardized protocol was established to assess egg surface quality of Chondrostoma nasus eggs. Multivariate statistics detected significant differences of egg surface properties among individual females and among three different populations. These differences were mainly attributed to length variability and merging of adhesive villi as well as to coating and filament- like connections of these structures. The findings of this study highlight the need for further investigations to better understand the relationship of egg surface properties, egg stickiness and hatching success to understand the recruitment ecology and performance of early life stages in freshwater fish.

fish eggs, particularly on an intraspecific level, as water chemical effects on egg development is likely to vary between populations spawning in different rivers.
A widely implemented tool for studying egg surface properties is the use of scanning electron microscopy (SEM) Rizzo et al., 2002). Although the ecology of a species cannot be automatically deduced from structure of the egg surface, eggs of most gravel spawning cyprinid species are characterised by a coverage with adhesive structures such as attaching villi (e.g. Patzner et al., 2006;Petz-Glechner et al., 1998;Riehl et al., 2002).
This ensures that a large proportion of the eggs laid adhere to surfaces at spawning sites, which in these species are often characterized by medium to rapid current velocities (Bartoň et al., 2021;Melcher & Schmutz, 2010). This is particularly evident in common nase (Chondrostoma nasus L.) which spawns its eggs at current velocities of up to 1 m/s or even greater (Melcher & Schmutz, 2010;Nagel et al., 2020b). Chondrostoma nasus is a specialist among riverine fish and has formerly constituted a large portion of the fish community in many rivers of Central and Eastern Europe (Kottelat & Freyhof, 2007).
There, this species plays an important role in the food web for lower trophic levels by grazing on benthic algae (Gerke et al., 2018) as well as for higher trophic levels by providing an important food source for apex predators (Šubjak, 2013). Yet, the degradation and fragmentation of habitats has led to severe population declines of this species Peňáz, 1996). As a result, Chondrostoma nasus is listed in several conservation lists (Bohl et al., 2003;Kirchhofer et al., 2007;Wolfram & Mikschi, 2007) and has become a flagship species for river conservation (Schiemer et al., 2002). Spawning of nase occurs in schools in which the sex ratio can reach up to 1 female on 25-30 males (Harsanyi & Aschenbrenner, 1995). During spawning, females scatter large numbers of eggs on the substrate surface of shallow gravel banks where several males immediately fertilize them (Peňáz, 1996). Eggs that are not able to adhere at the surface or the substrate interstices (Duerregger et al., 2018) of the spawning ground drift downstream (Hofer & Kirchhofer, 1996;Nagel et al., 2020b), where development success remains uncertain. Since these eggs are likely to face a higher mortality rate owing from potentially unfavorable development conditions and their availability for drift feeding and weak-swimming predators (Šmejkal et al., 2017), a high proportion of less adhesive and consequently off-drifting eggs may ultimately reduce recruitment potential.
There are anecdotal reports of fish breeders and anglers that suspect great differences in recruitment success of different nase populations. Recent findings also indicate that egg adhesive quality differs between spawning populations of Chondrostoma nasus (Nagel et al., 2020b), stressing the need for studying egg surface properties also on an intraspecific level. Consequently, the aim of this study was to investigate potential differences in egg quality of Chondrostoma nasus by comparing egg surface properties of seven females originating from three different spawning populations. A standardized protocol sheet for the analysis of SEM images was developed to test our hypothesis that surface structure of Chondrostoma nasus eggs shows a significant intraspecific variability.

| Female spawners
Seven females of Chondrostoma nasus were caught in April 2019 during their spawning migration in tributaries of the Inn River (Bavaria, Germany), the largest tributary of the Danube River in Germany ( Figure 1). Two females each were caught in the tributaries Isen (48°26′62.74″ north, 12°66′16.21″ east; April 1st 2019) and Mangfall (12°6'23.52" east; 47°50'46.66" north; April 1st 2019) and three females in the tributary Sims (12°9'1.02" east; 47°51'4.20" north; April 2nd 2019). All fish used for this study were caught in the course of breeding and re-stocking initiatives of local angling clubs using a 1.5 kW electrofishing device (Grassl). Prior to striping of eggs, fish were anesthetized with MS-222 (Tricaine methanesulfonate; concentration according to Adam et al., 2013). Subsequently, total length (TL) of each specimen was measured to the nearest cm and total weight (TW) was determined to the nearest gram. Scales were used to identify the age of each female by counting the annuli.
Immediately after egg release, subsamples of ~10 ml unswollen and unfertilized eggs from each female were preserved in 96% ethanol without any contact to water or other substances. Eggs were fixated for at least 10 days prior to further handling. www.olymp us-lifes cience.com). Eggs that were used for these measurement were not taken for subsequent scanning electron microscope (termed SEM hereinafter) imaging to avoid potential bias on egg surface analysis owing from mechanical damage caused by handling of the eggs.

| Egg size and SEM imaging
Nine eggs from each female were randomly selected for SEM imaging. First, egg moisture was removed using a vacuum (0.05 mbar) freeze dryer (Alpha 1-4, Christ,) at −47℃ for 120 s. Second, eggs were fixed to a SEM sample holder with conductive carbon adhesive pads and gold-coated using a Polaron SC502 Sputter Coater (Fisons Instruments).
Subsequently, eggs were examined with a SEM (S-2300, Hitachi) at a voltage of 25 kV, a geometric working distance of 10 and a magnification of 1,500. Nine photographs of the egg surface from each egg were taken, following the pattern displayed in Figure 2.
Technical settings of the SEM remained constant during imaging of all photographs.
Since the image quality of some photographs was not sufficient for a reliable assessment, which was especially true in the S3 sample, these were excluded. In order to obtain an equal number of images for each egg, seven images from each egg were randomly selected from the remaining photographs. This resulted in a total number of F I G U R E 1 Map of the study area and photographs of the rivers with studied spawning populations of nase F I G U R E 2 Left side: Egg of Chondrostoma nasus (×20) with visible microphyle (red arrow) and an overlaid schematic indicating the areas where photographs were taken. Right side: Magnification (×1,500) used to assess egg surface properties; note the adhesive villi covering the zona radiate externa of Chondrostoma nasus eggs TA B L E 1 Origin, ID, female attributes and egg size (mean ± SD) for each specific Chondrostoma nasus female used in this study as well as number eggs used for SEM imaging and number SEM images used for egg surface assessment. Note: All measurement of egg sizes was done with preserved eggs, which causes a volume reduction of ~ 25% (Patzner et al., 2006). TL = total length; TW = total weight 428 images for the assessment (Table 1). In a final step, all images were encoded and put into a randomized order by an external person. Subsequently, these images were reviewed by the same person and then recoded to their original ID.

| Assessment of egg surface properties
First, density of adhesive villi (AV) per image was determined by counting the number of AV on the egg surface. For each egg, only the image from the centered photograph was evaluated (Figure 2), as only this shooting angle allowed an accurate counting of all AV. Only fully visible AV were counted.
To systematically assess further egg surface properties, six criteria were defined and rated at a level of 0 (low), 1 (medium) or 3 (high). This rating scheme was adapted from a protocol that has been established to assess external injuries in fish and is capable of distinguishing possible differences between groups as well as to identify the underlying causes when combined with multivariate statistics (Mueller et al., 2017). Assessment criteria were defined according to a combination of results from a literature search (Patzner et al., 2006;Rizzo et al., 2002) and own observations on egg surface characteristics.

| Statistical analysis
Univariate statistics were used to test for differences in the den-  (Bray & Curtis, 1957) was computed using each image as a sample and each assessment criterion as a variable. Non-metric multidimensional scaling (nMDS) was performed to visualize differences in egg surface properties. The one-way analysis of variances (ANOSIM) was used to check for significances in egg surface differences of individual females and spawning populations. Subsequently, a similarity percentages analysis (SIMPER) was performed to reveal the criteria causing similarities and differences in and between the groups. All multivariate analysis were conducted in Primer v7 (Plymouth Marine Laboratory). For all analysis, significant differences were accepted at p < .05. and populations (Kruskal-Wallis-Test: χ 2 = 7.650; df = 2; p < .05).

| RE SULTS
When comparing individual females, AV density was lowest in M1 (171 ± 10) and significantly higher in all other Chondrostoma nasus ( Figure 4).
Based on a total of 428 images assessed according to the criteria of the protocol, ANOSIM detected significant differences between eggs from different populations (Global R: .32; p < .001) and overall females likewise (Global R: .34; p < .001; Table 3, Figure 5).
Differences were most pronounced among eggs from the Mangfall population with the Sims and Isen populations, but only small differences occurred in the comparison of females originating from the Isen compared to the River Sims, as reflected by the widely overlapping ordination of the symbols in Figure 5 and the low R value of this group comparision of only .064 (Table 3).
Egg surface images from the Isen population revealed an aver- Consequently, these criteria also caused the differences in the comparisons between the populations and individual females (Table 3).

| D ISCUSS I ON
The findings of this study point at distinct differences in the surface structure of Chondrostoma nasus eggs among populations and individuals, which likely affect adhesiveness and thus recruitment success in this species. The reasons for these differences may be explained by genetic effects such as local adaptation, by maternal effects or ambient environmental conditions which needs to be clarified in future studies. The protocol developed in this study has demonstrated its applicability to assess egg surface properties and, when used in combination with multivariate evaluation methods, its ability to identify potential intraspecific differences in the egg surface structure of Chondrostoma nasus. Egg quality in general is affected by several components, ranging from endocrine status and diet composition of the female during growth of the oocyte, nutrient composition of the oocyte to female attributes such as size and age as well as physico-chemical water conditions affecting egg incubation after egg release (Brooks et al., 1997;Keckeis et al., 2000). Yet, an effect of the latter can be excluded in our study, as eggs were directly striped and fertilized without any contact to water. However, a variety of reasons remain that could explain the differences observed. Keckeis et al. (2000) found that egg size and to a lesser extend also the chemical composition of the egg is highly influenced by the age of the female spawner. As female Chondrostoma nasus of the Sims population were older (10-11 years) than females from the Mangfall and TA B L E 3 Group comparisons of the different populations and individual females. R-and p-value are based on the one way analysis of similarities (ANOSIM). Average dissimilarity (AVDIS) and ranked criteria contribution (given in %) is based on the results of the similarity percentages (SIMPER) analysis Isen population (all 9 years), this could also explain the differences observed in our study, which were mainly caused by higher occurrence of filament-like connections, coating of the egg surface as well as a lesser length variability of AV in the Sims population.
Yet, the rather small differences in age of 1-2 years suggests that this is unlikely to be the case and stresses the need for further investigations. Future research should also include endpoints such as stickiness and hatching success, as it remains unclear if these are related to the differences in the observed eggs surface properties. However, previously observed differences in adhesive abilities (Nagel, et al., 2020b) and hatching success between the Mangfall and the Sims population in the wild (Duerregger et al., 2018) suggest that this is likely the case. This stresses the need of linking observations on egg surface properties to general egg quality expressed by egg stickiness and hatching success and other important incubation conditions such as physico-chemistry of the water (Kincheloe et al., 1979;von Westernhagen, 1988) and substrate composition (e.g. Nagel et al., 2020a;Sternecker & Geist, 2010). This is of particular importance as severe recruitment problems may arise from a combination of stressors such as a poor egg quality, a reduced adhesive ability resulting in higher off-drift of eggs and deteriorated habitat conditions on spawning grounds. Additionally, recent findings demonstrate that egg adhesiveness at spawning grounds can be extremely reduced in rivers with hydropeaking effects (Bartoň et al., 2021). In turn, improvement of spawning ground quality might partially compensate for reduced egg quality as a loose and porous interstitial as well as low fine sediment infiltration rates positively contribute to hatching success (Nagel et al., 2020a;Nagel et al., 2020a). In addition, a porous spawning substrate can incorporate a higher share of laid eggs, even if they have less adhesive abilities, and eggs infiltrating to the hyporheic zone are incubated in more sheltered conditions compared to those, that could not adhere at spawning sites (Duerregger et al., 2018;Persat & Olivier, 1995).
In light of still declining Chondrostoma nasus populations and intensive efforts to conserve and restore this species, future research is needed to better understand the relationship between egg surface properties and constraints for recruitment success in the early life history of this species. Assessing differences in egg surface properties in relation to adhesiveness and recruitment success in species with similar eggs other than nase may also be an important future direction in understanding fitness differences and resilience among individuals, populations and species in relation to changes of their habitats.

CO N FLI C T O F I NTE R E S T S TATE M E NT
The authors declare no conflict of interest.

ACK N OWLED G EM ENTS
We would like to thank E. Schulz, A. Huber (Kreisfischereiverein Rosenheim e.V.), R. Zillmer (Anglerbund Rosenheim) and M. Holzner (Bezirkzfischereiverein Mühldorf Altötting e.V.) for their permission to conduct this study and their assistance with fieldwork. We are also grateful to B. Gum and L. Egg (Fischereifachberatung Oberbayern) for their support, to M. Mueller for her involvement in the planning phase of this project and to U. Raeder and W. Kuefner for technical assistance during REM analysis.

DATA AVA I L A B I L I T Y S TAT E M E N T
Data are available from the corresponding author upon reasonable request.

Juergen Geist
https://orcid.org/0000-0001-7698-3443 F I G U R E 5 Non-metricmultidimensional-scaling (nMDS) of egg surface properties for different females (indicated by different symbols) and spawning populations (indicated by different colors). The distance between the symbols corresponds to the dissimilarity in egg surface properties (small distance = large similarity). Abbreviation of the IDs follow Table 1