Restricted – Commercial (When completed)

FIN22 1^st May 1999 31^st July 2003

Tel: 0131 4402116 Fax:      Email: Jim.Buchanan@onyxnet.co.uk

Studies of structural, behavioural and microbiological attributes will be carried using larvae reared at the SFIA hatchery, to provide baseline information on relevant aspects of larval biology and to determine the timing and nature of problematic events in commercial systems.

Theme 2 of the project will investigate whether tolerance to mechanical handling, or temperature change, vary according to developmental stage. The aim is to determine the most appropriate developmental stage for tank transfer and temperature acclimation, and to devise less damaging handling and acclimation protocols.

Theme 3 of the project will investigate whether the behavioural attributes of halibut larvae are responsive to differences in physical, chemical and biological variables. The aim is to identify appropriate environmental conditions that enable yolk sac larvae to be retained within the water column in commercial rearing tanks and that promote foraging behaviour during the feed initiation phase.

Some of the results of our research were used as the basis for successful commercial rearing trials conducted at Seafish, Ardtoe in 2002.

This part of the project is a collaboration with a UHI Millennium Institute Postgraduate Student, David Nairn. The primary focus of this work is on the development of mechanoreception. Larval fish have superficial neuromasts which, later in development, are enclosed in the lateral line canals. These organs detect water displacement and are responsible for detecting near field vibrations and water velocity but are vulnerable to damage. The gelatinous cupula which is deformed by water displacement is prone to breakage and may not reform properly or may result in malformation of the neuromast itself.. Neuromast development has been described and damage to both the cupulae and absence or malformation neuromasts has been observed. The large neuromasts that first appear on hatching are broken in 50% of larvae. The neuromast arrays surrounding the olfactory epithelia are frequently damaged or some absent in larvae that fail to feed.

The distribution of yolk-sac larvae in rearing silos was observed using infra red illumination and video recording. These experiments have confirmed the preliminary results which showed that the distribution of larvae within the silo depends upon salinity and not temperature variation. The response of larvae to salinity variation changes during development. Larvae can be found at all depths in the tank throughout the yolk sac phase but mean depth decreases during this period. Early in the yolk-sac phase larvae tend to move lower in the tank during bouts of low salinity rising again when salinity increases. Later larvae are less likely to sink when salinity is low but tend to move towards the surface temporarily when salinity increases again. Neuromasts are likely to be damaged when larvae interact with surfaces (walls or surface tension layer) or when there is considerable shear stress due to turbulence. Experiments with a ¼ scale glass model silo have shown that when water is flowing into the bottom of the silo at reduced salinity convection currents cause mixing of the water redistributing the larvae. Even when water salinity is stable flow at the base of the conical part of the silo is turbulent rather than laminar. Water flow in the silos is, therefore, likely to be the cause of neuromast damage to the larvae.

In an additional experiment (not in the original plan of work) we have used a density gradient column to test the density of yolk-sac larvae during development and have revealed that density increases during this period. The observation that larvae move nearer the surface later in the yolk sac stage can therefore only be explained by an increase activity or other changes in swimming behaviour. Video recordings will be reanalysed to address this issue.

It is clear that for hatcheries with seawater supply prone to salinity variation the salinity of water supplied to yolk sac silos needs to be controlled. To this end a salinity controlling system has been designed by the project team and applied in our most recent experiments.

A manual frame by frame analysis of in situ video recordings was use during the early part of the project to describe foraging behaviour. The development of suitable automated analysis software has limited the application of this type of analysis to video recordings from “green water” experiments. New software has recently been purchased by colleagues in Italy who we are working with in an EU funded project and will be evaluated for use in this project.

The major objectives were completed in the first year of the project. In larval halibut there was a defined progression in the development of the bacterial gut flora through the egg yolk sac phase and first feeding. In the yolk sac phase the gut was colonised by low numbers of bacteria of random types but on first feeding large numbers of bacteria colonised the gut, originating from the live feed. Colonisation was similar in different hatcheries and was highly selective, maturing such that vibrios and pseudoalteromonads not found initially came to dominate the flora of mature feeding larvae. The work has resulted in 2 publications – see below.

Objective 2a: Determine effects of mechanical handling on larvae at different developmental stages

Following the monitoring meeting in 2001, this experiment was successfully repeated in 2002, using a more robust handling method in which the contents of one bowl were simply poured into another, providing a greater stress. Survival results were similar to the original experiment (no significant difference; F_5,25 = 0.70, p = 0.632). However, unlike the above findings, there was also no significant effect on larval mass at first-feeding (F_5,24 = 0.08, p = 0.995).

Objective 2b: Determine the response of different larval developmental stages to a range of temperature acclimation regimes

Phase I: this objective is similar in design to 2a, with temperature increase (from 6 to 10^oC at a fixed rate) instead of standardised handling, at 150, 190, 230 and 270 (control) dd. As with the above objective, the suitability of the design has already been tested, and water baths (to raise the temperature at predetermined developmental stages) were installed in a dedicated facility. Similar data to Objective 2a were collected. There was no overall effect on end survival, although survival was lowest at 150 dd and highest at 190 dd. Although neither result was significantly different from the control, they were significantly different from each other (F_(1,8) = 5.25, p = 0.05). Temperature increase at different stages had no effect on larval size at the end point (290 dd).

Phase II of this objective involved raising the temperature at the least vulnerable developmental stage but at varying rates. On the basis of the above data, the temperature was raised at 190 dd at three different rates (1, 2 and 4 ^oC.d^-1), and compared with a control of a temperature increase of 1^oC.d^-1 at 270 dd This experiment was carried out using yolk-sac larvae from Ardtoe’s June/July 2001 spawning stock, and performance parameters assessed. The ‘high’ stress group (4^oC.d^-1) showed significantly higher mortality. The ‘low’ and ‘medium’ stress groups did not significantly differ from the control. However, the control group lost more mass per degree-day and subsequently weighed less than the experimental groups, with the ‘medium’ stress group (2 ^oC.d^-1) performing best in terms of mass at first-feeding.

Results from the above two objectives suggest that to maximise larval survival, mass and condition, larvae should be handled and subjected to temperature increase between 180 – 200 dd.

Objective 2c and 3b (Phase III): Application of results on a commercial scale

By day 60 post first-feeding, growth in one experimental group was significantly greater than it’s control counterpart (68 % greater; p<0.05), with a slight increase in percentage survival (21 vs. 19). In the other experimental group, there was no improvement in growth, but a large increase in survival (24 vs. 15 %). Therefore, in terms of biomass (dry weight; g), the experimental groups weighed 28.4 and 45.5 g at day 60, compared with 11.9 and 26.7 g for the controls. Although each group started with 20 000 stocked eggs, the early removal and high algal levels contributed to improvements in production of 138 and 70 % per experimental group.

Our first experiment showed that adding Nannochloris to rearing tanks resulted in better feeding, survival and growth of the larvae when compared with Isochrysis or Pavlova. More recent experiments, concentrating on the use of this Nannochloris, have shown that increasing microalgal density increases survival and growth. Densities below 2x10⁵ cells.ml^-1 had no effect and survival, growth and feeding incidence continued increase up to the highest density used (1.2x10⁶ cells.ml^-1). This is a much higher cell density than has been used routinely in the past. These effects could not be attributed to alteration in the bacterial flora, nutrient loading or ammonia levels in the rearing tanks.

Further experiments designed to ascertain the sensory stimuli provided by “green water” that stimulate feeding have focused on chemical and physical stimuli. When algae were added to dialysis bags floating in the rearing tanks larvae did not perform any better than no algae controls suggesting that a chemical stimulus alone may not be sufficient to stimulate feeding. In the final experiment the effect of inert particles, algae and algal filtrate on survival and first-feeding response were compared. Two types of inert particles, Kaolin and Vinacryl beads (Vinamul, Ltd.), were used with sufficient particles added to match turbidity with the Nannochloris treated tanks, 5 NTU. At this turbidity level absorbance (optical density) was similar in both Kaolin and Nannochloris treatments but much lower when Vinacryl was used. An additional treatment used Vinacryl at double the density in order to match absorbance with the other treatments. Survival after 11 days was similar in the Vinacryl and algal treatments at over 50% but Kaolin resulted in only 10% survival, similar to controls (11%). Although 25% of larvae in the algal filtrate treated tanks survived feeding and growth were very poor and similar to the Controls. We can conclude that the feeding and survival enhancing effect of “green water” is not due to chemical feeding due primarily to a physical effect and that it may be possible to replace algae with an inert substitute. Very low bacterial counts were found in the algal group compared with others with barely detectable levels of bacteria. This is not wholly unexpected as some algae have pronounced anti-bacterial effects. It is possible that this accounts in part for the high survival of larvae in this group. The vibrios isolated were almost all sucrose -ve, except for those from the kaolin group. In addition, the Vinamul group contained a higher proportion of non-vibrios than other groups (17/52 isolates) compared with (8/132) from all other groups combined. The poor survival of the Kaolin treated larvae is intriguing especially since survivors had fed and grown well but may be explained by these differences in bacterial flora. Further analysis will be carried out by identifying dominant organisms of the bacterial populations for each group. This will take some time to achieve as limited resources are available for this task.



Executive Summary of Research and Results/continued:
Objective 4a Investigate the interactions between bacteria isolated from halibut larvae.

Screening of a large number of bacteria isolated from halibut and turbot has revealed three which are highly antagonistic towards other bacteria, including the major pathogen Vibrio anguillarum.

In several infection trials 20 isolates from halibut larval rearing trials, both those which performed well and those performing poorly, were tested for their activity when added to larvae reared in the absence of bacteria. None of the 20 isolates caused larval mortalities, in contrast to Vibrio anguillarum, which was highly virulent. A manuscript has been submitted – see below.

These experiments (5 replicates) showed that the bacterial flora of larvae could be manipulated by (a) rearing in a recirculation system with water passed through a biofilter, or (b) addition of antibiotics, in comparison with the standard flow-to-waste system. Higher survival occurred in recirculated water and in antibiotic-treated larvae than with flow-to-waste. The concentration of bacteria in the gut was greatest in the recirculated water showing that the types of bacteria not numbers was important in determining survival. A manuscript is in preparation.

The initial work concentrated on determining whether potential probiotic bacteria could prevent mortalities on yolk sac larvae induced by V. anguillarum. None of the 3 isolates from objective 4a were capable of this and it is proposed that such studies should be limited to the first feeding phase when loading of the gut is possible by colonisation of Artemia with probiotic bacteria.

As an additional part of this section, linked to Objective 3b (Effect of microalgae on feeding behaviour) we have begun to investigate the changes in microbial flora in the larval gut under rearing conditions with different types and concentrations of microalgae. Molecular biology techniques are being evaluated to monitor changes in the gut bacterial flora. This involves surface sterilisation of the larvae, extraction of DNA, PCR amplification of the 16S rRNA genes of bacteria present and cloning the amplicons into a plasmid vector to form a clone library. From each sample we have selected 192 clones for interrogation with DNA probes specific for eubacteria and vibrios. Other probes are under development. Initial results have revealed the presence of some unusual bacteria, not likely to be detected by culture, as well as some routinely found by culture. This work will be continued over the next few months to provide a complete picture of the changes in bacterial flora induced by culture in the presence of different microalgae.









Future R&D resulting from this project (Include any other non-tangible benefits and state any Teaching Companies Scheme action if appropriate)

Our experiments on the distribution of yolk sac larvae in culture silos have highlighted inadequacies in the design of these tanks. Flow pattern and control of salinity could be improved. Trial and error would not be appropriate but with some input from physicists and modellers computational fluid dynamics could be applied to design an improved tank.

Birkbeck, T.H. and Verner-Jeffreys, D.W. (2002) Development of the intestinal microflora in early life stages of flatfish. In: Microbial Approaches to Aquatic Nutrition within Environmentally Sound Aquaculture Production Systems. (C.-S. Lee and P. O'Bryen, eds.) pp. 149-160. The World Aquaculture Society, Baton Rouge.

Verner-Jeffreys, D.W., Shields, R.J., Bricknell, I.R. and Birkbeck, T.H. (2003) Changes in the gut-associated microfloras during the development of Atlantic halibut (Hippoglossus hippoglossus L.) larvae in British hatcheries. Aquaculture, 219, 21-42.

Verner-Jeffreys, D.W., Shields, R. and Birkbeck, T.H. (2003) Bacterial influences on Atlantic halibut, Hippoglossus hippoglossus L., yolk-sac larval survival and start-feed response. Diseases of Aquatic Organisms, in press.

Verner-Jeffreys, D.W., Shields, R.J., Bricknell, I.R. and Birkbeck, T.H. Effect of different water treatment methods and antibiotic addition on larval survival and gut microflora development in Atlantic halibut (Hippoglossus hippoglossus L.) yolk sac larvae. Submitted to Aquaculture. March 2003.

Oral papers were presented at the British Marine Finfish Association Workshop in November 2000, November 2001 and November 2002.

To be completed by the Project Leader

Declaration

I declare that the information given has been approved by all the project participants and is correct to the best of the best of my knowledge and belief I understand that the information contained in this form may be held on a computer system.

Signed:Jim Buchanan Position:      Date:15/12/03

Programme Co-ordinator

LINK Aquaculture, FRM Ltd.,

Freshwater Fisheries Laboratory

Faskally

Pitlochry

Perthshire PH16 5LB

Tel: 01796 472 060 Fax: 01796 472 856 Email: majames@frmltd.com