Investigation into feed preparation for regulatory fish metabolism studies

43

8

Research ArticleReceived: 25 February 2013 Revised: 5 June 2013 Accepted article published: 12 June 2013 Published online in Wiley Online Library: 26 July 2013

(wileyonlinelibrary.com) DOI 10.1002/jsfa.6262

Investigation into feed preparation forregulatory fish metabolism studies†

Ina Goeritz,a Cornelia Atorf,a Paul Whalley,b Paul Seymour,b Michael Kleina

and Christian Schlechtriema∗

Abstract

BACKGROUND: Test diets used in fish metabolism studies for regulatory purposes must be homogenously fortified with theradiolabelled test substance and stable with respect to leaching. Standard fish food, as used in commercial fish farming, shouldalso be used in fish metabolism studies. Therefore, suitable spiking and coating procedures are required to ensure the correctdosing of the fish during the experiment.

RESULTS: Methods for the homogeneous, safe and efficient application of radiolabelled test items to the surface of commercialfeeding pellets were developed. Leaching studies showed that test items of low lipophilicity applied to feeding pellets need tobe stabilised, to reduce the risk of significant losses prior to ingestion by the fish. Coating of solvent-spiked pellets with calciumalginate was shown to reduce leaching losses of water-soluble test items significantly. Alternatively, commercial feeding pelletscan be coated with alginate or vegetable oil fortified with a radiolabelled test item also leading to sufficiently stabilised testdiets. Experimental conditions, such as the water temperature, may have a significant effect on the leaching behaviour of thetest items.

CONCLUSIONS: A detailed description of different spiking and coating procedures for the preparation of experimental dietssuitable for fish metabolism studies is provided.c© 2013 Society of Chemical Industry

Keywords: pesticide regulation; fish metabolism studies; experimental diet; coating; alginate; vegetable oil

INTRODUCTIONInformation provided for the authorisation of plant protectionproducts must be sufficient to permit an evaluation of the risks,to man and (food-producing) animals, arising from residues of theactive substance and relevant metabolites remaining in food andfeed.1 EU Directive EC 1107/2009 requires a livestock metabolismstudy whenever residues of plant protection products are likely.Up to now, only poultry (represented by laying hens), lactatingruminants (represented by lactating goats or alternatively cows)and pigs were tested. Metabolism studies on freshwater fish (rain-bow trout or common carp) will be necessary in the future for theauthorisation of plant protection products used in crops fed tofarmed fish.2 Metabolism studies are commonly carried out withradiolabelled test items applied either as capsules or via spiked ani-mal feed. For fish metabolism studies a standard food of suitablecomposition (as used in commercial fish farming) should be usedand further spiked by means of a suitable top-dressing method.However, as previously shown by investigations on medicated fishfeed, significant leaching losses of drug into the water may occurwhen surface-spiked pellets are not sufficiently stabilised.3,4 Thus,when testing moderately lipophilic active ingredients of plantprotection products (log POW < 5, where POW is octanol-water par-tition coefficient) in fish metabolism studies, appropriate measuresare recommended to prevent the test substance leaching from thesurface of the spiked pellets, prior to ingestion by the fish. Accord-ing to the working document on fish metabolism studies, leachinglosses from fortified diets should not exceed 10% prior to ingestion

by the test animals.5 Generally, care must also be taken to achieve ahomogeneous distribution of the test item throughout the spikedfeed, as this is one of the main difficulties associated with surfacespiking methods.4,6,7 Therefore, a robust and safe feed prepara-tion protocol is required, to ensure the production of water-stableand homogeneously fortified test diets. When a test substance issoluble in triglycerides, one suitable top-dressing method for fishfeed pellets is to dissolve the test substance in a small amount offish oil or vegetable oil before mixing with the feed.8 Alternatively,gelatin can be used as binding agent. This method is commonlyapplied for the preparation of medicated artificial diets used totreat fish diseases.4,9 Alternatively, the feed can be spray spikedwith the test substance using a suitable organic solvent.8 The lat-ter is advantageous because solvent-spiked pellets can be coatedafter fortification with different organic matrices, such as sodiumalginate3 to obtain sufficient water stability. Currently, there isno standard method to determine the stability of spiked feed in

∗ Correspondence to: Christian Schlechtriem, Fraunhofer IME, 57392Schmallenberg, Germany. E-mail: [email protected]

† This paper was presented at the 6th SETAC World Congress/SETAC Europe 22ndAnnual Meeting, 20–24 May 2012, Berlin, Germany

a Fraunhofer IME, 57392, Schmallenberg, Germany

b Syngenta, Jealott’s Hill International Research Centre, Bracknell, RG42 6EY, UK

J Sci Food Agric 2014; 94: 438–444 www.soci.org c© 2013 Society of Chemical Industry

43

9

Preparation of experimental diets for fish metabolism studies www.soci.org

water; however, principally, a certain amount of feed needs to beincubated in a beaker of water and allowed to sit for a variablelength of time with occasional shaking. Leaching losses are mea-sured in water samples taken from the test vessel after definedtime intervals and expressed in terms of test item concentration (in%) of the immersed feed sample. The aim of the present study wasto develop a solvent spiking procedure, ensuring the productionof water stable and homogeneously fortified test diets for fishmetabolism studies. The second aim was to investigate how theleaching rate in water of typical test items, characterised by lowlipophilicity, can be reduced by different pellet-coating materials.The third aim was to investigate the influence of water tempera-ture on the extent of leaching losses from coated pellets. Finally, analternative spiking procedure in which fortified coating materialwas applied to the feed pellets was tested. Fish metabolism studiesare recommended to be carried out on rainbow trout of 400–600 gbody weight.5 Therefore, in this study investigations were per-formed with commercial trout feed of a suitable pellet size (4 mm).

EXPERIMENTALFeed preparationStandards and reagentsA 14C-labelled plant protection product of log POW = 3.1 wasobtained from Syngenta. The specific radioactivity of test item was5.028 MBq mg−1 (feed in trials 1, 2 and 3) and 5.169 MBq mg−1

(feed in trial 4).All solvents used were high-performance liquid chromatogra-

phy grade and purchased from J.T. Baker (Deventer, Netherlands).

Experimental dietThe radiolabelled test item was applied to the surface of extrudedstandard trout feed (Milivit Type F-2P B40; Trouw Nutrition,Trouw Nutrition) with a pellet diameter of 4 mm. This feedstuff,which is based on a mixture of wheat, soy and fish meal, is used asa complete diet in commercial fish farming.

Feed preparation trials 1–3: Solvent spiking of fish feed pellets andsubsequential coatingSpiking apparatus and procedure. A rotary evaporator (Buchi R-210) was connected to a stainless steel capillary solvent-inlet tubeto obtain a simple vacuum spiking apparatus (Fig. 1). A 2 L pear-shaped drying flask (Buchi) was used as a rotating flask for sprayapplication. Two hundred grams of fish feed were estimated as agood sized batch for processing by the spiking apparatus. Duringspiking, the pressure of the vacuum system was set to 700 mbarand a rotation speed was set so that it caused an even distributionof the pellets on the inner surface of the flask. The amount ofsolvent (acetone) used to dissolve the test item was reduced to amaximum volume of 20 mL per application to prevent excessivemoistening of the pellets. The rotating drying flask was kept in awater bath at 40 ◦C during the spiking procedure to allow for thequick evaporation of the solvent. After dosing of the test item,the capillary inlet tube was sealed with Parafilm and the pressurewas reduced to 450 mbar for another 30 min of rotation, to ensurecomplete evaporation of the solvent. The spiked pellets werespread on aluminium foil and left for at least 24 h in the fume hoodbefore further use, in order to remove potential solvent residuesin spiked feed.

Figure 1. Rotary evaporator spiking apparatus.

Trial 1: calcium alginate coating versus corn oil coating. Test item(199 µg, equivalent to 1 MBq) was dissolved in 17 mL of acetoneand applied to 200 g of fish feed pellets as described above.The dosing vessel and solvent-inlet capillary were washed threetimes with 1 mL of acetone after the dose application, to ensurecomplete dosing of the test substance.

Four aliquots of 2 g of the spiked pellets were extractedthree-fold with 3 mL of acetone–dichloromethane (1:1, v/v). Thecombined extracts were centrifuged for 10 min at 1500 rpm andtwo 100 µL aliquots taken from the resulting supernatants. Each100 µL aliquot was mixed with 4 mL of Ultima Gold Cocktail(Perkin Elmer, Groningen, Netherlands) and analysed by liquidscintillation counting (LSC) (Tri-carb; Packard, Downers Grove,USA) to estimate the concentration of the test item.

Residues of the test substance attached to the rotary evaporator,the capillary inlet tube, flask and application vial were carefullyrinsed with acetone and the collected material was processedas the other samples were prior to LSC analysis. The recoverypercentage of the test item was calculated.

For the corn oil coating, 50 g of spiked feed pellets werethoroughly mixed with 0.5 g of corn oil (Sigma–Aldrich, Steinheim,Germany) and incubated overnight on a shaking table. An evenlydistributed thin oil film on the pellets (10 g kg−1 corn oil coating)was produced.

For the calcium alginate coating, 64 g of spiked feed wasthoroughly mixed with 10.2 g of a 20 g kg−1 solution of sodiumalginate (Sigma) in distilled water according to the method of Duiset al.3 This was mixed with 2 g of calcium chloride powder leadingto a hardening of the alginate film, which comprised 190 g kg−1 ofthe uncoated feedstuff.

The homogeneity of the test item in the spiked feed pellets wasverified to confirm the quality of the coated pellets. Samples of

J Sci Food Agric 2014; 94: 438–444 c© 2013 Society of Chemical Industry wileyonlinelibrary.com/jsfa

44

0

www.soci.org I Goeritz et al.

pellets coated with corn oil (3 × 200 mg) were pulverised and thencombusted (Oxidizer; Perkin Elmer, Groningen, Netherlands), priorto LSC analysis. Analysis of the calcium alginate coated pellets,in contrast, was carried out as described above by extractionfollowed by LSC.

Trial 2: different levels of calcium alginate coating. Solvent-spikedfeed was prepared using a spay application as described above. Thespiked pellets were later coated with different levels of calciumalginate. Apart from the amount of alginate coating (hardenedwith calcium chloride) per gram of pellets recommended byDuis et al.3 (190 g kg−1), lower levels representing 50% and 25%of the original amount (95 g kg−1 and 48 g kg−1, respectively)were tested. Samples of uncoated feed were analysed by LSC,concentration of the test item in coated feed was calculated basedon the theoretical dilution of test item in the pellets resulting fromthe calcium alginate coating.

Trial 3: calcium alginate coating for investigating the effects oftemperature on leaching. An additional batch of solvent-spikedfeed (1.99 mg of test item, equivalent to 10 MBq per 200 g feed)coated with 48 g kg−1 calcium alginate was prepared to investigateleaching losses in water of different temperatures. Feed samples(6 × 200 mg) were analysed by LSC after combustion as describedabove.

Feed preparation trial 4: simultaneous spiking and coating of fishfeed pellets with fortified alginate and corn oilAn alternative feed preparation method, where feed pelletswere spiked with the radiolabelled test item during the coatingprocedure, was tested.

For the corn oil coating (10 g kg−1), the test item (28.36 µg,equivalent to 147 kBq) was suspended in 0.5 g of corn oil beforemixing with 50 g of untreated pellets, as described for solvent-spiked pellets.

For the calcium alginate coating (190 g kg−1), alginate (7.96 gof a 20 g kg− 1 solution, prepared according to Duis et al.3),was thoroughly mixed with the test item (48.37 µg, equivalentto 250 kBq). This was applied to 50 g of untreated pellets andhardened with 1.56 g of calcium chloride, as described for solvent-spiked pellets.

The concentration and homogeneity of the test substance infeed samples of each approach were determined by LSC analysisafter extraction as described for solvent-spiked pellets.

Leaching experimentsLeaching trials 1–3: leaching of solvent-spiked pellets coated withcorn oil and calcium alginateDuring the first leaching trial, samples of solvent-spiked pelletswith 10 g kg−1 corn oil coating and 190 g kg−1 calcium alginatecoating (feed preparation trial 1) were tested in long-term leachingexperiments (up to 1 h) to calculate and compare the loss of thetest substance after immersion in water. Two grams of coatedpellets were placed in 200 mL of tap water (room temperature) ina wide neck bottle and gently shaken at 75 rpm using a horizontallaboratory shaker. Shaking was performed to imitate the watermovement generated by the fish and the flow-through systemunder real study conditions. After 5, 10, 15, 20, 25, 30 and 60 min,two water samples of 1 mL were removed from the bottle. Toensure homogeneity of the aqueous phase during sampling, justbefore the samples were taken, the water was gently stirred using

a scoop. Concentration of the test item in the water sampleswas measured by LSC after mixing with 4 mL of Ultima Gold LLTCocktail (Perkin Elmer). Leaching losses, expressed as a percentageof the test item concentration from the feed samples, werecalculated for each sampling point.

In the second trial the pellets coated with 190, 95 and 48 g kg−1

calcium alginate, as well as uncoated pellets (feed preparationtrial 2) were tested to estimate the trajectory of leaching lossesover a shorter incubation period of 10 min. Four grams of eachtype of calcium alginate coated pellets were placed in separatebeakers and filled with 400 mL of tap water (room temperature).As described in leaching trial 1, beakers were gently shakenthroughout the experiment. After 15, 30, 45 and 60 s, and 2, 3, 4, 5and 10 min, two samples of 1 mL were taken from every beaker andanalysed by LSC. Prior to sampling, water surface was briefly stirred.Each experiment was carried out twice to duplicate the sample size.

Finally, the effect of temperature on leaching rate was inves-tigated in the third leaching trial. Pellets coated with 48 g kg−1

calcium alginate (feed preparation trial 3) were immersed in 15 ◦C,20 ◦C and 25 ◦C water. In each case 2 g of feed pellets were placed ina wide neck bottle filled with 200 mL of tap water. Throughout theentire experiment the reaction vessels were placed in a horizontal-shaking water bath (75 rpm), to maintain a constant water temper-ature and movement. At each sampling point, e.g. 2.5, 5, 7.5 and10 min after the immersion of pellets in water, four water samplesof 1 mL were taken and analysed as described above by LSC.Leaching rates were evaluated as done for experiments 1 and 2.

Leaching trial 4: leaching of pellets simultaneously spiked andcoated with fortified corn oil and calcium alginateIn the fourth leaching trial, losses of the test item from feedcoated with fortified alginate and corn oil (feed preparation trial4) into water were investigated. Apart from a different samplingsize (4 × 1 mL instead of 2 × 1 mL of water), leaching experimentswere carried out and evaluated exactly as described in the firstleaching trial.

Statistical analysisLeaching rates, expressed in terms of percentage of test itemconcentration in food, prior to immersion in water, weretransformed {arcsin[(x/100)ˆ(0.5)] × 180/Pi()}, where x is percentleaching loss, and Pi is the ratio of a circle’s circumference to itsdiameter, equal to 3.14159, to satisfy the assumption of normality,and were then compared by two way ANOVA (SigmaStat3.5;Systat Software, Richmond, USA) followed by post hoc comparisonaccording to Holm–Sidak. The overall significance level was 0.05.

Leaching losses of pesticide from feed were described by doublefirst-order in parallel (DFOP) kinetics to calculate the time when10% of the substance has leached. DFOP is a bi-exponential modelthat describes degradation by the sum of two parallel normal first-order degradation rates in different part of the compartment (e.g.the feed). The kinetics needs a total of three parameters, namelytwo different first-order rate constants. The parameter g describesthe fraction of contaminant leaching according to the first rate con-stant. The equation of the DFOP model is presented in Equation 1:

M = M0[

g × exp (−k1t) + (1 − g) × exp (−k2t)]

(1)

where M is the mass (mg) at time t (min); M0 is the mass at timet = 0; g is the fraction of substance leaching according to the firstrate constant; k1 is the first leaching rate constant (min−1); and k2

is the second leaching rate constant (min−1).

wileyonlinelibrary.com/jsfa c© 2013 Society of Chemical Industry J Sci Food Agric 2014; 94: 438–444

44

1


RESULTS AND DISCUSSIONSolvent spiking procedureThe vacuum spiking apparatus (Fig. 1) allowed the gentle mixingand drying of the radiolabelled test item–solvent solution thatwas sprayed onto small batches of fish feed pellets. The testsubstance was homogeneously distributed on the surface of thespiked feed pellets as shown by the low coefficients of variationof 1.52% and 9.01% between the concentrations of test itemestimated in feed samples taken for chemical analysis (Table 1,trials 1 and 2). Almost the entire amount of test item applied wasrecovered (94–112%) within the feed (i.e. 1119 and 932 µg kg−1

of the nominal concentration of 995 µg kg−1 for trial 1 and trial2, respectively) with only minor residues (approximately 1.4% ofthe nominal concentration, data not presented) attached to thedifferent parts of the spiking apparatus detected. The use of thevacuum spiking apparatus thus ensured the safe and efficient useof the radiolabelled test item. The preparation of homogeneouslyfortified test diets for fish metabolism studies is an essential stepensuring that the test animals are exposed to the intended doseof a radiolabelled test item. However, the distribution of test itemon the surface of pellets can lead to high leaching losses into thewater prior to ingestion by the experimental animals.4 Apart fromthat, surface spiking may affect the palatability of the experimentaldiet3 leading to a reduced uptake or even rejection of feed bythe fish. Coating the surface of spray-spiked pellets with ediblevegetable oil or a settable gel material such as sodium alginatemay improve both the water stability and palatability of the testfeed.6 Therefore, to allow for these potential effects differentpellet-coating materials were investigated.

Trial 1: homogeneity and water stability of solvent-spikedpellets coated with corn oil and calcium alginateCoating of solvent-spiked feed pellets with 10 g kg−1 of cornoil caused a 6% decrease (from 1119 to 1051 µg kg−1) on theconcentration of test item (Table 1). Coating of the spiked feedwith 190 g kg−1 calcium alginate, however, resulted in a gel-likefilm surrounding the pellets that caused a 22% dilution (from1119 to 875 µg kg−1) of the test item in the pellets (Table 1). Thehomogeneity of the spiked pellets was only slightly reduced bythe coating procedures as demonstrated by the small coefficientsof variation of 4.69% and 5.42% between the concentrations oftest item measured in representative samples taken from the samebatches of feed.

Feed pellets coated with corn oil (10 g kg−1) and calcium alginate(190 g kg−1) were immersed in water for up to 60 min. Figure 2aillustrates that within 5 min spray-spiked pellets released 7–8% ofthe test item into the water, independent of the type of coatingapplied. However, after 10–60 min leaching losses from corn oiland calcium alginate coated feed differed significantly resultingin an accumulated leaching loss of 21% for corn oil coated feedand 12% for calcium alginate coated feed. The threshold valuefor leaching loss, as recommended in the guidance document forfish metabolism studies,5 is 10%. After immersion in water of the10 g kg−1 corn oil coated feed this threshold was calculated tobe exceeded after 6.4 min (Table 2). In contrast, the 190 g kg−1

calcium alginate feed coating slowed the time to reach the 10%threshold to 15.1 min. The relevant time period for feed uptake,for healthy fish and under a suitable feeding regime, is usually lessthan 5 min, even for slow feeding fish like carp.

The small amount of corn oil (10 g kg−1) used during the firstfeed preparation trial was sufficient to seal the spiked pellets with a

thin homogenous oil layer. This can be assumed, as 20 g kg−1 cornoil tested during our preliminary studies did not further reduceleaching losses in quantifiable amounts (data not published).However, considerable disintegration of corn oil coated feedpellets was observed during the leaching experiment. This mayexplain the higher leaching losses of corn oil coated feed comparedto calcium alginate coated feed.

Coating the feed with calcium alginate reduced leaching lossesand disintegration of the pellets, which shows that a calcium algi-nate coating provides a better seal of the surface of spiked pelletsthan an oil coating. This is in accordance with the results of Duiset al.3 who compared leaching from calcium alginate (190 g kg−1)and cod liver oil (40 g kg−1) coated feed, spiked with antibacterials.

Oil coating of pellets for larger fish is commonly used for thepreparation of medicated artificial diets.3,9 Most frequently, thedesired amount of drug is mixed with the feed and then oil(5–40 g kg−1) is added to bind the drug to the pellet. However,preliminary tests in our laboratory showed that the use of oil as anadhesive should be kept as low as possible and that no proportionhigher than 20 g kg−1 of corn oil coating is recommended. Thisis because coating with 30–40 g kg−1 of corn oil clearly affectedthe consistency of the spiked feed pellets, leading to a drippingoily surface of the pellets and to a rapidly formed oil film onthe water surface when transferred to the fish tank. Apart fromconsistency aspects, excessive use of oil can also significantlyimpair the nutritional value of the diet.

Alginate gel, is formed when divalent cations are added to analginate solution and does not dissolve in water. The gel acts asa diffusion barrier leading to a decreased migration of chemicalsfrom feed.10,11 Alginates are naturally occurring biopolymers thatare widely used in the pharmaceutical, cosmetic, and food industryand are regarded as generally non-toxic and biocompatible.12,13

However, there are indications that biopolymers such as alginatemay have detrimental effects on nutrient digestibility in fish, asthey accelerate gastrointestinal transit time.14,15 A feeding studycarried out by the first author16 showed that rainbow trout canbe easily adapted to calcium alginate coated test diets. Growthperformance of the animals was usually acceptable. Little isknown about possible effects of alginate on the bioavailability oftest items in coated feed. Since the extent of swelling increasesas hydrogels, such as alginate, pass down the intestinal tractdue to an increase of pH,12 it can be assumed that the chemicalis fully released from the diet. However, depending on thechemical:alginate ratio and the calcium chloride concentration,the chemical release may be prolonged or retarded.11

Gelatin is a commonly used binding agent for surface spikingof feed pellets with medical products9,17 and may therefore beconsidered as an alternative to calcium alginate. Gelatin, which is amixture of water-soluble proteins derived from collagen by hydrol-ysis, is practically odourless and tasteless.10 However, gelatin wasnot investigated in this current study, as in a previous study com-paring the palatability of different binding agents used as adhesivefor the antibiotic amoxycillin, gelatin was clearly unpalatable torainbow trout.4 Also, in contrast to calcium alginate, gelatin swellsand softens in water and may thus not resist leakage of the test item.

Improved water stability and palatability of surface-spiked feedmay be achieved, without further coating, when feed pellets aredrained in the chemical–solvent solution leading to a deeperabsorption of the test item into the pellets.18 However, this maylead to a significant extraction of nutrients from the diet andnamable solvent residues in the diet, as the complete evaporationof the solvent from the inner core of the pellets cannot be ensured.


44

2


Table 1. Concentrationa, b of test item (log Pow = 3.1) in 4 mm trout feed pellets in (µg kg−1) ± standard deviation (n = 4)

Leaching trial

Feed preparation 1 2 3 4

Uncoated 1119 ± 17 (1.52) 932 ± 84 (9.01) — —

Corn oil, 10 g kg−1 1051c ± 57 (5.42) — — 558e ± 11 (1.97)

Alginate, 190 g kg−1 875 ± 41 (4.69) 783d ± ND — 767e ± 29 (3.78)

Alginate, 95 g kg−1 — 851d ± ND — —

Alginate, 48 g kg−1 — 889d ± ND 11 228c ± 99 (0.88) —

The coefficient of variation (%) is shown in parentheses.a Concentrations of test items in the trout feed presented in this table were taken as a basis for the calculation of leaching losses of test items fromthe feeding pellets.b Nominal concentration for uncoated feed is 995 µg kg−1 in trial 1 and 2. Nominal concentration for corn oil and alginate coated feed is 985 and836 µg kg−1, respectively in trial 1, and 566 and 812 µg kg−1, respectively, in trial 4. Values for nominal concentrations consider dilution by weightgain due to add-on of coating material.c Concentration of test item in feed samples was determined via liquid scintillation counting after combustion with a sample size of n = 3 for corn oilcoated and n = 6 for alginate coated feed.d Calculated values, as coated feed had not been measured. Calculation is based on theoretical dilution of test item in feed due to the add-on of 190,95 and 48 g kg−1 alginate coating, respectively.e Feed was spiked via the coating material instead of using the vacuum spraying apparatus.n is the sample size; ND, not determined.

Figure 2. Leaching losses of pesticide from solvent-spiked trout feed of 4 mm pellet size (a) with 190 g kg−1 calcium alginate ( ) and 10 g kg−1 corn oilcoating ( ) after immersion in water; (c) without coating ( ), with 190 g kg−1 ( ), 95 g kg−1 ( ) and 48 g kg−1 ( ) calcium alginate coating after immersionin water; (d) with 48 g kg−1 calcium alginate ( ) after immersion in water at 15 ◦C ( ), 20 ◦C ( ) and 25 ◦C ( ). (b) Leaching rates of feed spikedvia 190 g kg−1 calcium alginate (diamonds) and 10 g kg−1 corn oil (circles) coating. Leaching losses are expressed as a percentage of the initial pesticideconcentration in feed. The 10% leaching threshold is indicated by a dotted line ( ). The points of intersection with the timeline axis were determinedby DFOP fitting (see Table 2). Statistics: Significant differences between the treatments are marked using different superscripted letters. To facilitatereading, in Fig. 2c differences are marked only from the 2 min interval onwards. If the letters appear only once, at the end of the curve (Fig. 2b), differencesaccount for each single time interval along the curve.

Trial 2: stability of solvent-spiked pellets coated withdifferent levels of calcium alginateThe second leaching trial showed that the level of calcium alginatefor pellet coating can be significantly reduced compared to thecoating level suggested by Duis et al.3 without causing higherleaching losses. Surprisingly, even lower leaching losses were

observed when the suggested calcium alginate level was reducedby 25% or even 50% (Fig. 2c). For uncoated and 190 g kg−1 calciumalginate coated pellets the 10% leaching level was calculated to beexceeded after 0.8 and 2.7 min, respectively (Table 2). However, forpellets coated with 95 and 48 g kg−1 calcium alginate, a leachingloss of 10% was only reached after immersion in water for 11.7


44

3


Table 2. Time of immersion in water (min) until 10% of the test itemhas leached from the feed*

Leaching trial

Feed preparation 1 2 3 4

Uncoated — 0.8 — —

Corn oil, 10 g kg−1 6.4 — — 32.9

Alginate, 190 g kg−1 15.1 2.7 – 53.2

Alginate, 95 g kg−1 — 11.7 — —

Alginate, 48 g kg−1 — 17.2 >31 —

∗Values in the table are the results of double first-order in parallel(DFOP) analysis.

and 17.2 min, respectively. Therefore, pellets coated with 95 and48 g kg−1 calcium alginate are sufficiently stable to be used in fishmetabolism studies.

Trial 3: effect of water temperature on leaching losses fromcalcium alginate coated pelletsThe optimal water temperature for rainbow trout and commoncarp is 15 ◦C and 25◦, respectively. Therefore, the influenceof a range of water temperatures (15, 20 and 25 ◦C) on theextent of leaching losses from solvent-spiked pellets coated with48 g kg−1 calcium alginate was determined. Leaching losses inall treatments increased significantly over time amounting to3–5% after 10 min (Fig. 2d). The 10% leaching threshold was notexceeded before 31 min, as estimated by DFOP modelling (Table2). The significantly lowest losses occurred at 15 ◦C, except for the5 min time point. Leaching losses measured at 20 ◦C and 25 ◦Cwere not different until at 10 min, after which, significantly higherlosses were observed at 25 ◦C (Fig. 2d). The small differencescan be explained by the temperature dependent kinetics of theleaching process. The accelerated rehydration of the alginate gelinduced by the higher temperature may also have contributed toan increased release of entrapped substances.10 The results showthat experimental conditions such as the water temperature canhave a significant effect on the stability of an experimental diet.Therefore, the leaching potential of a test diet should always beestimated under conditions comparable to the experiment.

Trial 4: homogeneity and water stability of pelletssimultaneously spiked and coated with fortified calciumalginate and corn oilSimultaneous spiking and coating of the feed with calcium alginate(190 g kg−1) and corn oil (10 g kg−1) was comparable to the solventspiking procedure followed by coating. The recovery of test itemfrom the simultaneously spiked corn oil coated feed was 99%(558 µg kg−1) compared to 108% (1051 µg kg−1). The recovery oftest item from the simultaneously spiked calcium alginate coatedfeed was 94% (767 µg kg−1) compared to 105% (875 µg kg−1).Homogeneity was demonstrated for both coatings by the lowcoefficient of variation of 1.97–3.78% (Table 1).

Leaching losses from pellets coated with fortified alginate hard-ened with calcium chloride (190 g kg−1) were significantly lowerthan the losses estimated for feed coated with fortified cornoil (10 g kg−1) (Fig. 2b), further indicating calcium alginate asthe preferential coating material. The 10% leaching thresholdfor 190 g kg−1 calcium alginate coated feed was not exceededuntil 53.2 min of immersion in water, whereas for 10 g kg−1

corn oil coated feed 10% leaching was observed after 32.9 min(Table 2). Comparing these findings with the results of the firstleaching experiment indicates that leaching, from both calciumalginate and corn oil coated feed, occurs significantly slower whenthe test item is mixed with the coating material prior to applyingto the feed. However, comparing the results of different test trialscannot be done without restrictions, as different batches of feedpellets were used for feed preparation. Generally, the differentspiking procedures (solvent spiking followed by coating or spikingwith fortified coating material) seem to be equally suitable for thepreparation of test diets for fish metabolism studies with regard toefficiency, homogeneity and leaching. However, when test itemswith a log POW = 3.1 are applied, the use of a gelling polymer suchas calcium alginate could potentially reduce leaching losses com-pared to edible oil coating. As calcium alginate can influence thechemical composition and thus the nutritional value of the test diet,the amount of calcium alginate applied to the pellets should bekept as low as possible. An amount of 48 g kg−1 of calcium alginatemay be sufficient to reduce leaching losses to less than 10%.

CONCLUSIONS

• The solvent spiking procedure described in this study providesa suitable method for the homogeneous, safe and efficientapplication of radiolabelled test items to the surface ofcommercial feeding pellets.

• When test items of low lipophilicity are applied to feedingpellets that are not further stabilised, there is a high risk that thebulk of the dosed test item will leach from the pellets surfaceprior to ingestion by the fish. Coating of spiked pellets is thusrequired.

• Coating of solvent-spiked pellets (4 mm) with calcium alginatewas shown to reduce leaching losses of water soluble testitems significantly. Alternatively, commercial feeding pelletscan be coated with alginate or vegetable oil fortified with aradiolabelled test item to obtain sufficiently stabilised test diets.

• Experimental conditions such as the water temperature mayhave a significant effect on the leaching behaviour of the testitems. Therefore, spiked diets should be always tested prior totheir use in fish metabolism studies to prove their stability inwater.

ACKNOWLEDGEMENTSThis study was jointly funded by Syngenta and Fraunhofer IME.

REFERENCES1 European Commission, Regulation (EC) No 1107/2009 of the European

parliament and of the Council of 21 October 2009. Off J EuropeanUnion L309(24.11.09): 1 (2009).

2 European Commission, Regulation (EU) No 283/2013 of 1 March2013 setting out the data requirements for active substances, inaccordance with Regulation (EC) No 1107/2009 of the EuropeanParliament and of the Council concerning the placing of plantprotection products on the market. Off J European Union L93(3.4.13):1 (2013).

3 Duis K, Inglis V, Beveridge MCM and Hammer C, Leaching of fourdifferent antibacterials form oil- and alginate-coated fish-feedpellets. Aquacult Res 26:549–556 (1995).

4 Treves-Brown KM, Applied Fish Pharmacology. Kluwer AcademicPublishers, Dordrecht, pp. 6–12 (2000).


44

4


5 European Commission, Working document on ‘Nature of Residues inFish’ SANCO/11187/2013, Jan 2013, http://ec.europa.eu/food/plant/pesticides/guidance_documents/docs/app-j_en.pdf (visited on17th of July 2013)

6 Daniel P, Drugs and chemicals in aquafeeds: The problems andsolutions. Options Mediterraneennes 86(A):85–94 (2009).

7 Prescott JF, Baggot JD and Walker RD, Antimicrobial Therapy inVeterinary Medicine, 3rd edition. Iowa State University Press, Ames,p. 701 (2000).

8 Organisation for Economic Co-operation and Development (OECD),Bioaccumulation in Fish: Aqueous and dietary exposure. (OECDguidelines for testing of chemicals, TG 305.) OECD, Paris (2012).

9 Noga EJ, Fish Disease, Diagnosis and Treatment, 2nd edition. Wiley-Blackwell, Ames, pp.366–367 (2010).

10 Paolucci M, Fabbrocini A, Volpe MG, Varricchio E and Coccia E,Development of biopolymers as binders for feed for farmed aquaticorganisms, in Aquaculture, ed. by Zainal Abidin Muchlisin. InTechPublishers, New York, pp. 3–34 (2012).

11 Tonnesen HH and Karlsen J, Alginate in drug delivery systems. DrugDev Ind Pharm 28:621–630 (2002).

12 George M and Abraham E, Polyionic hydrocolloids for the intestinaldelivery of protein drugs: Alginate and chitosan – A review. JControlled Release 114:1–14 (2006).

13 Prabaharan M, Prospective of guar gum and its derivatives as controlleddrug delivery systems. Int J Biol Macromol 49:117–124 (2011).

14 Rosas C, Tut J, Baeza J, Sanchez A, Sosa V, Pascual C, et al, Effect of typeof binder on growth, digestibility, and energetic balance of Octopusmaya. Aquaculture 275:291–297 (2008).

15 Storebakken T, Binders in fish feeds I. Effect of alginate and guargum on growth, digestibiblity, feed intake and passage through thegastrointestinal tract of rainbow trout. Aquaculture 47:11–26 (1985).

16 Goeritz I, Falk S, Stahl T, Schafers C and Schlechtriem C, Biomagnificationand tissue distribution of perfluoroalkyl substances (PFASs) inmarket-size rainbow trout (Oncorhynchus mykiss). Environ ToxicolChem DOI: 10.1002/etc.2279.

17 Duborow RM, Medicated Feed for Food Fish. [Online]. Available: https://srac.tamu.edu/index.cfm/event/searchFactSheets/?searchterm=473&quickSearchSubmit=Search+Factsheets [25 February 2013].

18 Szoke A, Oral veterinary drug delivery system and/or feed additivedelivery system, process for the preparation and use thereof. WorldIntellectual Property Organization, WO2004/039172, A1 (2004).


Documents

Investigation into feed preparation for regulatory fish metabolism studies