Zingel asper (Linnaeus, 1758)

Béjean, Mickaël, 2019, Reproduction of Zingel asper (Linnaeus, 1758) in controlled conditions: an assessment of the experiences realized since 2005 at the Besançon Natural History Museum, Cybium 43 (1), pp. 17-32 : 22-28

Behaviours and needs of Z. asper in captivity

Acclimatisation

Wild Z. asper acclimatise quickly once in captivity. Care should be taken to ensure that each fish has a hiding place. They are initially fed on small earthworms. Two weeks after their arrival, they start eating dead Chironomidae . Since 2007, five batches of wild Z. asper have been added to the aquarium in the Besançon Natural History Museum. Only a few individuals died during the acclimatisation phase.

Behaviour

The behaviour of Z. asper in captivity can be summarised as passivity during the day and first signs of activity at the end of the day. This very low level of diurnal activity is interspersed with short periods of movement to find hiding places.

We noticed rapid eye movements and sensitiveness to nearby peripheral disturbance. Despite the fish are permanently monitoring their environment, they can be easily caught if hand movement is relatively slow. Z. asper therefore adopt passive behaviour. Escaping is only used as a last resort but Z. asper are capable of sudden acceleration, which can propel them out of the water. If the tank walls are not sufficiently high, they can easily escape. The edges of the tank must rise above the surface of the water by at least 20 cm and a lid on each tank is necessary to prevent Z. asper from jumping out of the tank at night.

It should be noted that each Z. asper occupies the same area for several weeks but, one month before the reproduction period, males return to the riffle and stay there until May. Females only go to the riffle to lay their eggs.

By ensuring that there is at least one hiding place for each fish, up to 40 adult individuals per mº can be assembled in winter and 30 individuals per mº in summer.

Diet and growth

Z. asper have strict dietary requirements both in terms of quantity and quality ( Corse et al., 2017). They will continue feeding at temperatures of 5°C. They favour frozen Chironomidae and live maggots and earthworms. Fish were fed ad libitum and portions readjusted in accordance with water temperature ( Fig. 3B ) and growth of the fish. Food was distributed three times a week for adults and daily for juveniles under the age of 6 months. Removal of detritus and unconsumed food appeared to be essential. Aspiration of the bottom of the tanks before feeding ensures a sufficient degree of cleanliness.

A group of 50 Z. asper born in the Besançon Museum was monitored for nine years. At 1 year old, the average size (from snout to end of caudal fin) was 11.3 ± 1.1 cm, with an average weight of 12.9 ± 3.3 g. Some individuals already measured 13 cm in length and a few males had milt. At 2 years old, some fish could not be sexed but all Z. asper were mature at 3 years and monitoring of both sexes was possible ( Fig. 3A ). Broodstock were therefore measured and weighed in February before reproduction to facilitate sexing of the fish. Z. asper size increases rapidly up to the age of 6 years old and then the growth curve levels out. Since the third year, females are significantly larger than males that rarely exceed 17 cm in length. In contrast, females can reach 19 cm in length since their fourth year. One female reached 21 cm in length ( Fig. 1B ).

Reproduction

No egg laying or reproduction has yet been observed in the natural environment, although broodstock gathering in areas of current have been seen in March and April. In contrast, certain reproduction behaviours were easily observed since the first reproduction tests in aquarium. Moreover, the implementation of nocturnal video monitoring (in 2008) allowed to film most egg laying events. From 2005, using a broodstock individual recognition system ( Fig. 1C ), we found that certain females (in particular the largest ones), could stagger egg laying over several nights at 24- or 48-hour intervals. But most of time, eggs were laid in a single night. On 12 March 2008, a female laid eggs during seven hours, from 9:00 to 16:00. All stages were observed and filmed ( Béjean and Maillot, 2008). Altogether, these observations made in artificial environments since 2005 teach us the following: (1) spawning activity can mobilise one to 20 males and one to two females simultaneously; (2) however, most of the time, only two or three males accompany one female ( Fig. 4 ); (3) depending on their size, females can produce 300 to 3000 ova and expel them in batches of around 40 oocytes; (4) most of the time, eggs are expelled onto gravel, and a small proportion of them may be carried by the current and settle further downstream; (5) most ova are laid on clean gravel in fast-flowing current, although some clutches of eggs are found in areas of weak current (between 0.4 and 0.8 m /s); (6) the females can take up to seven hours to expel all her ova; (7) egg laying can take place from late February to mid-May at temperatures between 8 and 12°C; (8) maximum activity is observed during the months of March and April at temperatures of 10 to 11°C; (9) males do not leave the spawning ground from early February to the end of May, while females only go there to spawn; (10) egg size is 2 mm

Incubation and hatching

Full embryonic development ( Fig. 5 A-C) requires between 250 and 380 degree-days. This corresponds to an incubation time of 20 to 30 days at a temperature varying from 10 to 13°C. The “eyed” stage appears after around 10 days. The length of the development period varies widely, even between eggs subjected to identical incubation conditions.

Many trials were conducted and the method selected consisted of applying a temperature of 11°C for the first seven days. After a first sorting operation, the eggs were subjected to a temperature of 13°C until the 19 th day. They were then placed in the hatching module at 15°C. With this procedure, full embryonic development requires 280 degree-days. However, incubation success rates fluctuate greatly depending on the thermal cycle the broodstock were subjected to, and special attention was paid to the influence of winter temperatures on hatching rates (see below).

Hatching takes place in just a few minutes. However, some larvae extricate themselves in two stages: initially, the envelope is broken but the head and abdomen remain inside it. Several hours later, or even the following day, the larva manages to free itself. In some cases, however, it dies without completing its exit. Before this occurs, it can be removed from its envelope using very fine tweezers. Several hatching techniques were tested. The most effective consisted of leaving the eggs to float freely in a gentle current. In 2016, this method achieved a 95% success rate in 5,398 eggs. The results and zootechnical parameters are summarised in table I.

Larvae

The first few weeks of larvae development can be broken down into 4 phases: post-hatching phase until the first intake of food ( Fig. 6A ); pelagic phase ( Fig. 6B ); benthic phase ( Fig. 6C ) and juvenile phase ( Fig. 6D ).

Particular behaviours are seen at each stage and they determine the rearing parameters used. Just after hatching, the larvae stay immobile on the bottom or close to the substrate. They can already move around, even in the current. The larvae only attempt to reach the surface to fill their swim bladder after several hours. They gather together in the lightest corner of the tank. At this stage, they work their way into the slightest crack or slit, where they can stay hidden in. The larvae were placed in small rearing tanks at a concentration of 25 per litre with a water temperature of 15°C. After two to five days, the larvae colonise the entire water mass and start to feed. It is easy to catch them using a pipette and the transparency of their body allows to check if they have eaten or not. They are fed three times a day with Artemia nauplii .

The benthic phase starts after 15 to 20 days. The larvae then abandon the open water, and gradually colonise the bottom as well as the vertical walls of the tank. However, they systematically avoid placing themselves on the glass panels. They start to eat pieces of bloodworm and pigmentation commences. One-month survival rates are around 80% in these conditions at temperatures varying from 15 to 18°C. The appearance of nostrils is the signal to move the larvae to a larger tank at a concentration of 400 to 500 fish per mº. The water temperature may then exceed 20°C.

The juvenile phase starts 40 to 50 days after hatching, as small Z. asper acquire their definitive morphology and eat whole bloodworms. At this stage, they are nocturnal and adopt the same behaviour as adults. From this point onwards, losses are very rare and rearing them no longer presents any difficulties, unless they are raised in aquariums made entirely of glass. At 2 months, they measure between 35 and 40 mm and we chose this size to transfer them to their natural environment. To transport them, the fry were packaged in batches of 300 in 50-litre bags containing one-third water and two-thirds pure oxygen. They were able to survive in these conditions at least nine hours.

In 2016, several rearing trials using larvae aged between 1 and 2 months old were conducted with concentrations varying from 315 to 559 individuals per mº. The average survival rate in these experiments was 93 ± 5.8%. The range of concentrations tested does not allow to relate the survival rate and the original larvae concentrations (Rº = 0.0049). However, the average survival rate and that obtained with the higher concentration (94.9%) indicates that rearing 1-monthold larvae at a concentration of about 500 larvae per mº does not cause any supplementary loss.

Up to the age of 1 year, small Z. asper were reared at a concentration of 100 per mº. At 1 year old, they were treated as adults. They measured between 7 and 13 cm.

Sensitivity

Although this species sometime copes with unfavourable physicochemical conditions (nitrite concentrations of 0.2 mg /l measured via twice-weekly monitoring), it appears to be sensitive to excess organic matter, even in small amounts. Surplus quantities over long periods can promote the development of mycosis in the gills ( Fig. 7A, B ). The first symptoms are fast breathing, the fish then keeps its mouth open and eventually dies through asphyxiation. This sensitivity is increased during the reproduction period when the majority of deaths is due to mycosis. Treatment is possible using chloramine-T or malachite green (Tab. III). In 2017, Z. asper from the natural environment were infested by Dactylogyrus parasites. After seven months in captivity, they caused fatalities by weakening the fish. A course of treatment with praziquantel successfully eradicated this widespread infestation in the gills and at the base of the fins.

Longevity, fertility and fecundity

In captivity, some Z. asper live over than nine years. Observation of a group of Z. asper born in 2008 at the Besançon Museum and monitored until 2017 allowed us to study changes in the number of ova per female as a function of age ( Fig. 8 ). The number of eggs per female starts reducing after the fifth year. The number of eggs per female then diminishes rapidly, reaching very low production in the 8 th year. This led us to stop using the data from this group for interpretations of the experiment on winter temperatures presented below. However, it is worth noting that for two remaining females, egg production started again in the ninth year. Regarding the average hatching rate, at the age of 7 years old, it was still close to 74%. At 8 years, the rate dropped to 45% and at 9 years it rose to 64%, without any obvious explanation, as the environmental conditions were the same. By the 7 th and 8 th year, there were only three females, and two in the 9 th year. By the end of 2017, seven broodstock were still alive in this group, including one female.

Physical characteristics

Each fish can be individually identified thanks to the black bands on its body since their patterns are sufficiently distinctive. Dorsal and lateral photographs (Fig. 1C) allow to produce an accurate register for each cohort. This identification proved extremely useful during observation of reproductive behaviour. It was gradually abandoned as the number of Z. asper in the breeding programme increased. However, this technique was successfully used to count the wild Z. asper population in the Swiss part of the Doubs River ( Bonnaire, 2012).

A characteristic of the species is the ability of its eyes to reflect torchlight. This property was used from 1998 onwards to count the wild populations in the first LIFE programme. Observations in aquariums have shown that pupil dilation in Z. asper is affected by light levels.

Influence of winter temperature on reproduction

The broodstock used to study the influence of vernalization period length came from 2 sources. The first group was reproduced in 2008 from wild broodstock from the Beaume River, and 50 fries were taken as samples from the egg clutches obtained in 2008 to take part in the experiment. These Z. asper fish of the “Beaume” stock were used from 2012 to 2015. The second group consisted of two batches of 30 wild Z. asper from the Durance River ( Tab. I ), caught in the frame of the reintroduction programme in the Drôme River in 2015 and 2016. The Z. asper fishes of the “Durance” stock (used in 2016 and 2017) were replaced between the two years. Tanks DR1 and DR2 (Appendix 1) were selected for this experiment as they were equivalent and allowed easy monitoring of reproduction. Feeding and lighting conditions were similar.

More than 64,000 eggs were collected during these trials and all egg clutches were accurately counted at four key moments of the incubation process: on laying, at 10 days, at transfer for hatching and at hatching. Average survival rates up to hatching differed as the experiment conditions changed ( Fig. 9A ) (Tab. II).

Firstly, the broodstock subjected to clement winter temperatures produce egg clutches of mediocre quality, or even no egg at all in the case of the group subjected to no more than a 30-day vernalization period at 6°C. From the first development stages, egg mortality declines as the length of the winter cold period increases. After 10 days of incubation, a significant improvement in the survival rate is observed when the vernalization period exceeds 90 days. However, survival rates continue to moderately decrease over the next stages. Within the same group, survival rate sometimes varies widely. The graph in figure 9B illustrates, clutch by clutch, survival rates for the tests conducted on the “Beaume” broodstock born in captivity. The egg clutches produced during the same experiment are identified by an identical colour. The hatching rates for broodstock subjected to 90 days of vernalization are the most variable. They range from 18 to 52.4% with an average of 36.2% ±14.5%. This great variability suggests that the 90-day period of cold temperatures induce a boundary, which determines whether reproduction is successful or not. For this period, four out of five females produced egg clutches with 10-day survival rates of 70 to 88%; the rate dropped dramatically between 29 and 52% at hatching. The survival rates of egg clutches subjected to a vernalization period shorter than 90 days show a very sharp drop before the tenth day and then later stabilised. In contrast, the experiment with a cold period of more than 90 days produced very good results, with survival rates of between 70 and 80% from the tenth day up to hatching.

These tests were completed with two experiments using wild broodstock of various ages from the Durance River in 2015 and 2016. Variation in survival rates of egg clutches from wild broodstock is similar to those of broodstock born in captivity with the same vernalization period of 120 days. The influence of vernalization period length on the hatching rate is illustrated in the graph in figure 9C. The hatching rate varies in accordance with the duration of the cold period experienced by the broodstock during the winter. For the “Beaume” broodstock, Kruskal-Wallis tests show that comparisons between the results obtained with the 120-day period and with the other periods are significant (p <0.05) and that only the test between the data of the 75-days and the 90-day periods is not significant (p = 0.06). The length of the vernalization period has a marked influence on the hatching rate. This test was also applied to data from the experiments involving both sets of stock and 120-day periods and show that both sets reacted similarly to this vernalization period (p = 0.07). The Wilcoxon-Mann-Whitney test was applied to these data and confirmed the results of the previous statistical test.

The duration of this cold period probably also impacts on broodstock mortality. In 2013, a large number of fishes of the “Beaume” broodstock died (Tab. II). This increase in mortality could have been caused by a higher concentration of fish. However, it was primarily females that died, either just after laying or without having laid eggs. During the 2013 reproduction period, 70% of the females died. We noticed that the clutches contained eggs 1.5 times larger than normal. This anomaly was also presented by the ova of the females that were not able to lay. In contrast, no large eggs were seen in the clutches of broodstock subjected to 120 days of vernalization. It seems that overly clement winter temperatures caused early maturation of the ova and ultimately adversely affected their expulsion during spawning, leading to significant mortality among females.

The best hatching rates were therefore obtained with a vernalization period of 120 days. However, it is important to apply this period at the correct time. The graph in figure 10 summarises the different life phases of Z. asper in captivity, as a function of temperature. Four key moments in the life of Z. asper can be identified where water temperature plays an important role. When the temperature reaches 14°C to 15°C in May, Z. asper consume triple the quantity of food eaten at 5°C to 10°C. From 20°C in July, consumption levels are more than 5 times higher than in the winter. The fish take advantage of the summer season to grow and build reserves for the production of gametes. This time period with temperature of 14°C to 20°C (and higher) is the section shown as “growth” on the graph ( Fig. 10 ). The “gametogenesis” phase starts when temperatures fall in September and continues throughout winter but slows down in November when the temperature decreases and reaches 5°C. This time period with temperature around 5°C (“vernalization” in Fig. 10 ) lasts from November to February. Rapid and regular increase of temperature to 10°C triggers reproduction from the beginning of March until the end of April.

Figures

Figure 1. – Zingel asper specimens. A: Lateral view Zingel asper, male, 2 year old specimen born in 2008 in Besançon Natural History Museum; B: Lateral view, 4 year old female, size of 21 cm TL (captive born 2008); C: Dorsal photo of a Z. asper group, each fish can be individually identified thanks to the black stripes on their bodies (1-6).

Figure 3. – A: Growth curve of the Z. asper group born in captivity (2008) descended from wild Z. asper from the Beaume. At 2 years old, some fish could not be sexed but all Z. asper were mature at 3 years and monitoring of both sexes was possible. At 5 years old, females are significantly longer than males. B: Quantity of food per week for 20 asper (Beaume 2008) in 2015. When the temperature reaches 14 to 15°C in May, Z. asper consume 3 times the quantity of food eaten at 5 to 10°C. From 20°C in July, consumption levels are more than 5 times higher than in the winter.

Figure 4. – Female (in front) with 3 males in the current just before the expulsion of ova.

Figure 5. – Development of Z. asper before hatching. A: Embryo at 4 days; B: Embryo at 12 days; C: Hatching at 20 days.

Figure 6. – Development of Z. asper after hatching. A: Larvae post-hatching; B: Pelagic larvae at 5 days; C: Benthic phase at 15 days; D: Juvenile phase at 40 days.

Figure 7. – Health of gills. A: Early signs of mycosis on gills; B: Healthy gills.

Figure 8. – Egg production as a function of age. These results were obtained from the number of eggs laid throughout the breeding season divided by the number of females that laid. The same blood- stocks (born in captivity in 2008) were used for 9 years. The maxi- mum production was obtained at 6 years. The number of females was 17 at the beginning and 2 at the end (Tab. II).

Figure 9. – A: Comparison of incubation results as a function of cold period length from 30 to 120 days. More than 64,000 eggs were collected during these trials and all egg clutches were accurately counted at four key moments of the incubation process: on laying, at 10 days, at transfer for hatching and at hatching. From the first development stages, egg mortality declines as the length of the winter cold period increases. B: Clutch by clutch detail of the incubation survival rate of eggs for batches of “Beaume” broodstock subjected to 3 different temperature periods. The egg clutches produced during the same experiment are identified by an identical colour. Within the same group, survival rate results sometimes varied widely.The hatching rates for broodstock subjected to 90 days of vernalization are the most variable. This wide variability suggests that the 90-day period of cold temperatures creates a boundary, which determines whether reproduction is successful or not. C: Variation in hatching rate (mean and standard deviation) as a function of vernalization period length. Unlike those from the “Beaume” stock, Z. asper of “Durance” stock we used in 2016 and 2017 (captured in 2015 and 2016), were replaced between the two years. The hatching rate varies in accordance with the duration of the cold period experienced by the broodstock during the winter. For the “Beaume” broodstock, Kruskal-Wallis tests show that comparisons between the results obtained with the 120-day period and the other periods are significant (p <0.05). The length of the vernalization period clearly has an influence on the hatching rate. This test was also applied to data from the experiments involving both sets of stock and 120-day periods.As this test was not significant (p = 0.07), it demon- strates that both sets of stock reacted similarly to this vernalization period.

Figure 10. – Annual thermal cycle for different life phases of Z. asper. The “gametogenesis” phase starts when temperatures fall in September and continues throughout the winter but slows down in November when the temperature reaches 5°C. This period of time at 5°C, indicated as “vernalization” on the graph, lasts from November to February. The duration of this period is very important because it directly conditions the success of the reproduction. Rapid and regular rising of temperatures to 10°C triggers reproduction from the beginning of March and continues until the end of April.