Different Feeds and Feeding Regimens have an Impact on Zebrafish Larval Rearing and Breeding Performance

This boom in research has led its aquaculture to grow exponentially worldwide without sufficient accompanying studies on novel methods for husbandry and larval rearing which would optimize an intensive production of fish for research with adequate standardization and fish welfare. With new transgenic tools being rapidly developed, the zebrafish mutation project as well as its current role as a powerful tool in drug screens, it became essential to find better protocols and methods to breed and raise zebrafish faster, more reliably and with better welfare. However, little is known on the nutritional requirements of zebrafish [29, 30], being mostly reared with information available for Cyprinaformes [30, 31]. This is becoming a major concern within the research community as it imposes a difficulty to standardize a husbandry protocol in different facilities [32, 33]. Proper nutrition is not only important for individual growth and survival [29] but also to the reproductive success [34] which directly affects offspring fitness [35].

feeding protocol [33,34,[41][42][43][44].A major drawback of these studies is that they focused mostly on the juvenile and adult stages leaving the larval stage largely unstudied.The few studies contemplating first-feeding have used live feeding [45,46], rendering few data on dry food as first-feed.
In this study we present a comparison analysis for three commercial feeds (ZM, Zeigler Larval AP100 and Skretting Gemma Micro) in three different zebrafish strains (wild-type, mutant, transgenic), under three different feeding regimens after day 9 post-fertilization (dry feed only, a combination of dry feed with live feed and live feed only) regarding growth, survival, sexual maturity, fecundity and embryo viability.The main goal was to understand which dry feed would better complement live feeding and up to a certain extent would reduce the need for live feeds, taking into account the growing demand of zebrafish facilities to decrease the fish life-cycle and increase breeding performance and welfare.

Fish
Three different zebrafish strains were used, representing the main types of fish used in labs -wild-type, mutant and transgenic fish -in order to check if there was different feeding requirements for the different strains.Representing the wild-type, AB was chosen as the most common wild-type strain used in research.Representing the mutants, Nacre (mitfa -/-), a pigmentation mutant, was chosen due to its broad distribution among labs for being often used as a background strain for imaging in neuroscience and cancer research.HuC:GCaMP5G, a transgenic strain expressing the calcium indicator GCaMP5G in neurons [47] in a heterozygous background (AB x Nacre) was chosen as an example of a transgenic strain.

Embryo generation
Sexually mature zebrafish from each of the three study lines were set up to cross in pools at dawn.Eggs were collected, transferred to Petri dishes with embryo medium and kept in an incubator at 28°C until 6 dpf.Each dish contained groups of 50 embryos that were then vested to a 3.5 L tank filled with 500 mL of embryo medium and introduced in the main holding system with no water flow (Tecniplast ® recirculating system with the following water parameters: pH 7.2; 28.0°C; 1000 µs).During the following 2 days, 200 mL of system water was added to each tank per day.From 9 to 15 dpf, larvae were kept with system water drop-flow and afterwards started on constant flow.

Feeding assays
Tanks from each line were assigned to nine experimental groups according to the diet and regimen Table 1.For each group there were two replicate tanks.
All tanks were fed four times per day in working days and once a day in weekends and holidays, always at the same hour of the day.In working days, tanks from the dry live groups were fed 2 times Artemia + 2 times dry feed.In the weekends and holidays, tanks from dry only and dry live groups were fed only dry food whereas tanks from the control group were fed only Artemia.The first two feed sizes (up to 30 dpf -Skretting ® Gemma Micro 75 and 150; Zeigler ® Larval AP100<100 and 100-150; ZM ® 00 and 100) were dispensed with a 1000 µL pipette tip, one dip tip per feeding.Subsequent feed sizes (After 30 dpf -Skretting ® Gemma Micro 150; Zeigler ® Larval AP100 150-250; ZM ® 200) were given with a disposable Pasteur pipette tip, one dip tip per feeding.Freshly hatched decapsulated Artemia (ZM ® Prime) was also administered with a disposable Pasteur pipette at a mean concentration of 750,000 Artemia per liter of system water.Fish were fed in excess in all feeding events, i.e. were fed enough to feed continuously for 5 minutes with food still present in the water after this period of time.

Fish measurements
Individual fish counting was done at 3 dpf, 30 dpf, 60 dpf and 90 dpf.The 3-dpf counting was done under a light stereoscope whereas subsequent counts were done using a fish net.
Fish length was measured at 45 dpf, 60 dpf, 75 dpf and 90 dpf.Each tank was photographed three times with a SONY® SteadyShot (DSC-W510) and the best photo was used to measure the fork length (from snout to tail split) in ten randomly chosen individuals (except in tanks with a density lower than 10 individuals, in which case all available individuals were measured) using ImageJ software (v.1.45s, Wayne Rasband National Institutes of Health, USA).

Density control
Embryos were divided in 50 individuals per dish and at 6 dpf vested to 3.5 L tanks (density of ~14 fish/L).At 30 dpf all fish were counted and tanks holding more than 25 individuals were divided in two tanks for a maximum density of ~7 fish/L in order to decrease the influence of density in the overall results.

Accessing reproductive parameters
Starting at 75 dpf and once every other week until 180 dpf or until six successful matings (i.e.egg laying), all groups were set up to cross, one male and one female per breeding tank.In the next morning, eggs were collected and counted at 0 dpf to determine the number of live fertilized eggs and at 3 dpf to determine embryo survival.

Statistical analysis
All data was analyzed with the software IBM SPSS Statistics (v.21, SPSS Inc., Chicago, IL).After assumption verification, t-student and one-way ANOVA test were used.Effect dimension was evaluated by η2p and when significant differences were observed, they were analyzed with the Tukey's HSD test in order to perceive which groups differ.Significant statistical effects were considered for results with p-value <0.05.

Results
Results were initially treated individually for the three fish groups (wild-type, mutant and transgenic).No differences were detected showing that fish strains do not significantly influence results (data not shown).Thus, all results for these fish groups were pooled and subsequently treated as a single group.

Breeding performance and Sexual maturity
In Figure 3 we show the age at which fish laid their first eggs upon being set in pair crosses after sexual dimorphism has been visually observed, which age reflects sexual maturity.In the dry only feeding group, Gemma Micro produced significantly different results (F(2,7)=17,703; p=0,002; η2p=0,835; π=0,99), with the age of fish sexual maturity being greatly reduced (less than 90 dpf) as compared to fish fed with the other two brands (more than 125 days).In the tanks fed with dry live, there were no significant differences betweenall feed brands (F(2,15)=0,599; p=0,562; η2p=0,074; π=0,131).This reveals that both Larval AP100 and ZM benefit from the complementation with artemia.In the control group differences between Larval AP100 and ZM-fed fish (F(2,12)=4,14; p=0,043; η2p=0,408; π=0,615) were observed, again reflecting a negative impact of ZM as a first feed (as shown also for the survival) on the sexual maturity.

Different Feeds and Feeding Regimens have an Impact on Zebrafish Larval Rearing and Breeding Performance
Copyright: © 2016 Certal et al.

Discussion
A previous study had already shown feeding trials presenting a better performance of Gemma Micro against another commercial dry feed in another fish species [48].Our results have now shown the same better performance in zebrafish.On a feeding frequency study, this same diet had been already used in zebrafish with good results in adults without however being neither compared with any other feed nor tested in larval rearing [49].In our facility conditions, Gemma Micro stood out as best performer in almost all assessed parameters, with fish presenting higher survival rates, shorter sexual maturity age and highest fecundity.
When using only live feed after 9 dpf (control group), Larval AP100-fed fish showed a higher mortality at 30 dpf (74%) as compared to the other diets.This probably suggests that the type of food intake from 5 to 9dpfis crucial to provide adequate nutrients for the larvae to grow and attain the necessary size to be able to eat Artemia.Larval AP100-fed fish showed also the lowest growth, oldest sexual maturity (102, 5 days) and lowest number of eggs laid per breeding pair event within the dry live group which further supports the idea that this feed does not offer adequate nutrients (as compared with the other two) to support zebrafish development even when complemented with Artemia.
Tanks fed under the dry only regimen with ZM showed the worst survival rates (16%) and growth, which makes this feed a poor replacement for live feeds.In the control group it showed a high mortality rate in the survival assessment at 30 dpf (45%) probably due to the same reason as Larval AP100: it did not provide enough nutrients for the larvae to growth and start eating Artemia.Regarding the breeding parameters in the control group, although ZM-fed fish had a slight poorer performance as compared with the other two brands, all feeds showed a very poor breeding performance, revealing that Artemia by itself is a poor diet for optimal breeding success regardless of the first feed.However, when dry and live feeds are combined, ZM showed results as good as Gemma Micro, suggesting that this feed may be suitable to complement, but not to replace, live feeding practices at later stages.
It should be noted that in this study no live feed was used from 5 to 9 dpf in order to assess the impact of the dry feeds as first exogenous feed.However, if a live feed such as Paramecium or rotifers would be used as first feed, we believe that survival rates would be improved for all feeds, possibly allowing for a dry only feeding regimen after day 9 if using Gemma Micro in view of the results presented here.
Importantly, Larval AP100 and ZM-fed fish had both higher mortality rates than Gemma Micro, which resulted in a lower density, even with the density control measures taken.Density is known to affect reproductive performance [50] and lower densities can also make individuals grow faster than in higher densities which could account for the difference attenuation in both growth and reproductive performance within the control group.Interestingly, even with higher mortality rates in the dry only regimen, and consequently lower densities, fish fed with Larval AP100 and ZM did not surpass the ones fed with Gemma Micro.
Several studies in zebrafish fed with different commercial/ non commercial feeds had shown effects of different feeds and feeding regimens in reproductive performance [41,49,51,52].However, fish were fed with test feeds only after young juvenile or adult stages, thus not revealing the impact of feeding during development nor assessing the quality of dry feeds during larval rearing.Our study has shown that fish may attain sexual maturity as soon as 75dpf depending on the feed and feeding regimen.As expected, our results reflect the effect of the different feeding protocols in the reproductive performance through the number of eggs laid per breeding event, where fish fed with Gemma Micro produced the best results.However, no impact of the tested feeds was seen in the embryo viability, with 30-40% embryo mortality being observed for all feed types in all feeding regimens.Similar to what was discussed above for the survival, we believe that if a live feed would be used as first feed, this embryo mortality may be greatly reduced due to the impact on embryo quality [46].We also analyzed the female/male ratio per tank but it resulted in no significant differences (data not shown).
We believe that the difference in composition of the 4 types of feed tested (3 dry feeds + Artemia, Table 2) is not likely to support Performance.SOJ Aquatic Res 1(1): 1-8.

Different Feeds and Feeding Regimens have an Impact on Zebrafish Larval Rearing and Breeding Performance
Copyright: © 2016 Certal et al.
the significantly different data obtained in this study.Although we did not have access to the detailed composition of each feed, in particular the fatty acid n-3/n-6 ratio, the general content in protein and fatty acids is not dramatically different.However, Gemma Micro is produced by a new process substantially different from the other two dry feeds (and to the best of our knowledge all other dry feeds currently used in zebrafish) -the method called cold extrusion spherizer agglomeration production, a process of cold extrusion of raw materials.As stated by Skretting and others [53,54], with this method the protein, carbohydrates, essential oils and vitamins save their own specific values without any loss, the nutrients being more digestible and stable in water.As a result, they either prevent water pollution in the tanks or the fish consume it with more desire and digest very easily, which may explain our results.Accordingly, and as a qualitative observation, tanks fed with Gemma Micro were recorded had having cleaner environments when compared with the other two brands.
Zebrafish had been previously reared successfully without any live feeds [41], however the diet was custom-formulated which imposes a difficulty to use it in others facilities.Our data have now shown relatively good results with an available commercial feed.Gemma Micro's performance when used alone shows moderate results during development (especially in survival rates), which in a large scale practice may have a negative impact suggesting that it may not be suitable to fully replace live feeds, at least during developmental stages.However, when compared with the other tested feeds, it appears as the most suitable feed to complement live feed protocols.Notwithstanding, Artemia did have a negative effect on Gemma Micro breeding performance Figure 4, which altogether points to the idea that Gemma Micro may be suitable to completely replace Artemia in adult zebrafish.

Figure 2 :
Figure 2: Fish growth assessment by mean fork length (mm): a) 45 days post-fertilization; b) 90 days post-fertilization.Error bars: ± standard error of the mean.

Figure 3 :
Figure 3: Mean fish age (day's post-fertilization) of first egg lay.Error bars: ± standard error of the mean.

Figure 4 :
Figure 4: Mean number of fish eggs laid per breeding event.Error bars: ± standard error of the mean.

Figure 5 :
Figure 5: Embryo viability as the percentage of live embryos at 3 days post-fertilization compared to the number of eggs laid.Error bars: ± standard error of the mean.