Comparison of growth performance of Nile tilapia (Oreochromis niloticus) fingerlings fed different inclusion levels of black soldier fly (Hermetia illucens) pre-pupae meal diets and its effect on the physical characteristics of the feed

Groenewald, Niklaas Jakobus (2018-03)

Thesis (MSc)--Stellenbosch University, 2018.

Thesis

ENGLISH ABSTRACT: The nutritional composition of fish and the health benefit associated with the consumption thereof makes it a valuable food and feed source for both humans and animals. However, with the increasing size of the world’s population and the demand for fish and other fishery products, an increased amount of pressure is being placed on wild catch fisheries. As a result, the latter are becoming stagnant, or declining, in many regions around the world. Aquaculture has, to some extent, been able to relieve the pressure placed on wild catch fisheries. Yet, a major downfall is that it still relies on wild caught fish to be used as a protein source for many aquaculture fish species. Various alternative protein sources has been investigated for use in aquaculture feeds to replace conventional and unsustainable protein sources, such as fishmeal. Among the various alternative animal protein sources, the use of insects are gaining popularity as a promising and sustainable solution. Insects are a natural food source for many fish species and have relatively balanced amino- and fatty acid profiles, while it is also highly effective in integrated waste management (IWM) systems. The objective of the study was to investigate the effects of different inclusion levels of black soldier fly (BSF) (Hermetia illucens) pre-pupae meal in the diets of Nile tilapia (Oreochromis niloticus) fingerlings (n = 630) on the growth performance, fillet yield and feed quality. Six diets were compared: BSF0 (0 %; the control diet), BSF5 (5 %), BSF10 (10 %), BSF15.6 (15.6 %), BSF20 (20 %) and BSF25 (25 %; the summit diet). There were no significant differences in the final average body weight, feed intake, average daily gain (ADG) and conditioning factor (CF) between the different treatments. There was no significant difference for the feed conversion ratio (FCR) of BSF0 (1.35) and BSF5 (1.50), however FCR was lower (P≤0.05) for BSF10 (1.63) relative to BSF0, and worsened with increasing inclusion levels. The FCR was not significantly different between BSF10 and BSF15.6 (1.79), but BSF10 did have a significantly lower FCR than BSF20 (1.97) and BSF25 (1.91), whilst the latter two did not differ (P>0.05) from each other. During the preparation of the experimental feed, the inclusion of a binder was required in BSF20 and BSF25 in order to prevent complete crumbling with the extrusion process. The use of a binder changed the hardness and floating characteristics of diets BSF20 and BSF25, which may have had an influence on differences in FCR. There was no significant difference observed for specific growth rate (SGR) between BSF0 (2.05), BSF5 (1.91) and BSF10 (1.73), or between diets BSF10 through to BSF25 (1.62). Treatment BSF10 was thus not significantly different from any treatment diets. Treatment BSF0 was the only diet with a SGR significantly higher than BSF15 (1.68), BSF20 (1.60) and BSF25 (1.62). The protein efficiency ratio (PER) of BSF0 (2.02) and BSF5 (1.82) did not differ significantly from each other. The PER of BSF0 differed significantly from BSF10 (1.68), BSF15.6 (1.55), BSF20 (1.40) and BSF25 (1.43). The difference could be linked to the increasing lipid content of the feeds which is known to adverse effect the digestibility of the feeds. The carcass yield was not significantly different between any of the treatments, where all the yields (percentage body weight) ranged from 81.63 g (BSF20) to 83.5 g (BSF0). The fillet yield did show some variation, but the results are not very accurate (R2=0.094) due to the methodology of filleting. Furthermore, no significant differences were observed for the proximate composition (moisture, crude protein, lipid and ash) of the fillets. The inclusion level of the BSF pre-pupae meal showed a statistically significant influence on the unit density, sinking velocity, water uptake and leaching rate of the pellets. However, the effect of binder used in BSF20 and BSF25 may have had an influence on these results. All unit densities differed significantly from each other, increasing with higher BSF pre-pupae inclusion levels. The control (BSF0) and BSF5 had a sinking velocity of 0 (floating pellets), while the feed only started sinking with inclusion levels of 10 % BSF pre-pupae meal (BSF10, 4.44 cm/s). Faster sinking velocities were observed with increasing inclusion levels of BSF pre-pupae meal. Water uptake was generally higher in BSF0 compared to other treatments for all the timeframes, except for the shortest submersion time (5 min), where BSF15.6 (288.60 %) had a significantly higher water uptake than all the other diets. Due to the disintegration of the pellets over time, the water uptake was expressed as a percentage of feed remaining rather than initial quantity used. The feed remaining was used to determine disintegration rate. Factors such as lipid content and interactions between ingredients may have indirectly influenced feed quality parameters. The study did not generate enough evidence to verify the claim that BSF pre-pupae meal can be used as a viable alternative protein source to conventional sources in Nile tilapia feeds, due to variability in water stability and inclusion of a binder that were not accounted for. However, it was found that inclusion levels of up to 25 % can be used without any effect on the body composition, and up to 5 % without compromising the growth parameters. Therefore, it is recommended to use an inclusion level of 5 % BSF pre-pupae meal to maintain the growth performance. For future studies, it is suggested to use defatted BSF pre-pupae meal for potentially higher inclusion levels of the meal for comparative growth results relative to conventional protein sources – as the higher lipid content may adversely affect the binding ability and the feed quality.

AFRIKAANSE OPSOMMING: Die voedingswaarde van vis en die gesondheidsvoordele wat met die verbruik daarvan gepaard gaan, maak dit 'n waardevolle voedsel- en voedingsbron vir beide mens en dier. Alhoewel, met die toename in die wêreld se bevolking en die vraag na vis en ander visseryprodukte word daar toenemend druk op wildvang vissery geplaas. Gevolglik stagneer of verminder wildvang vissery in baie streke regoor die wêreld. Akwakultuur het tot 'n mate die druk op wildvang vissery ligter gemaak. 'n Groot tekortkoming is egter dat akwakultuur nogsteeds staatmaak op die gebruik van wilde vis as 'n proteïenbron vir baie akwakultuurvisse. Navorsing word voordurend gedoen op die gebruik van verskeie alternatiewe proteïenbronne in akwakultuurvoere om konvensionele en nie-volhoubare proteïenbronne, soos vismeel, te vervang. Die gebruik van insekte as 'n belowende en volhoubare oplossing raak egter meer gewild in vergelyking met verskeie ander alternatiewe dierlike proteïenbronne. Insekte is uiteraard 'n natuurlike voedselbron vir baie visspesies en het 'n gebalanseerde amino- en vetsuurprofiel, terwyl dit ook hoogs effektief is in geïntegreerde afvalbestuurstelsels. Die doel van die studie was om die effek van verskillende Swart Soldaat Vlieg (SSV) (Hermetia illucens) pre-papie meel insluitingsvlakke in die diëte van Nyl tilapia (Oreochromis niloticus) vingervissies (n = 630) op die groeiprestasie, filet opbrengs en voergehalte te ondersoek. Ses diëte is met mekaar vergelyk: SSV0 (0 %; die kontrole dieet), SSV5 (5 %), SSV10 (10 %), SSV15.6 (15.6 %), SSV20 (20 %) en SSV25 (25 %; die toppunt dieet). Daar was geen betekenisvolle verskille in die finale gemiddelde liggaamsgewig, voerinname, gemiddelde daaglikse toename en kondisioneringsfaktor tussen die verskillende behandelings nie. Daar was geen beduidende verskil vir die voer omskakeling verhouding (VOV) van SSV0 (1.35) en SSV5 (1.50) nie, maar die VOV was laer (P≤0.05) vir SSV10 (1.63) relatief tot SSV0, en het versleg met die toename in insluitingsvlak. Die VOV het nie betekenisvol tussen SSV10 en SSV15.6 (1.79) verskil nie, maar SSV10 het 'n beduidende laer VOV as SSV20 (1.97) en SSV25 (1.91) gehad, terwyl die laaste twee nie van mekaar verskil (P>0.05) het nie. Tydens die voorbereiding van die eksperimentele voer is die insluiting van 'n bindmiddel benodig in BSF20 en BSF25 om te verhoed dat die ekstrusieproses volledig verkrummel. Die gebruik van bindmiddel het egter die hardheid en drywende eienskappe van diëte BSF20 en BSF25 verander, wat die verskille van FCR beïnvloed het. Daar was geen betekenisvolle verskille waargeneem vir die spesifieke groeikoers tussen SSV0 (2.05), SSV5 (1.91) en SSV10 (173) nie, of tussen diëte SSV10 tot SSV25 (1.62) nie. Behandeling SSV10 was dus nie betekenisvol anders as enige ander behandelingsdieet nie. SSV0 was die engiste dieet wat ‘n spesifieke groeikoers beduidend hoër as SSV15 (1.68), SSV20 (1.60) en SSV25 (1.62) gehad het. Die proteïen doeltreffendheid verhouding van SSV0 (2.02) en SSV5 (1.82) het nie betekenisvol van mekaar verskil nie. Die proteïen doeltreffendheid verhouding van SSV0 het betekenisvol van SSV10 (1.68), SSV15.6 (1.55), SSV20 (1.40) en SSV25 (1.43) verskil. Hierdie verskil kan egter aan die toenemende lipiedinhoud van die voer gekoppel word, aangesien dit bekend is dat die lipiedinhoud die verteerbaarheid van die voer nadelig kan beïnvloed. Die karkasopbrengs het nie betekenisvol tussen enige van die behandelings verskil nie, waar al die opbrengste (persentasie liggaamsgewig) van 81.63 g (SSV20) tot 83.5 g (SSV0) gewissel het. Die opbrengs van die filet het 'n mate van variasie getoon, maar die resultate is nie baie akkuraat nie (R2=0.094), moontlik as gevolg van metodologie van filetering. Daar is verder geen betekenisvolle verskille vir die proksimale samestelling (vog, ruwe proteïen, lipied en as) waargeneem nie. Die insluitingsvlakke van SSV pre-papie meel het 'n statistiese betekenisvolle invloed op die eenheidsdigtheid, sinksnelheid, wateropname en uitlogsyfer van die voer gehad. Die effek van die bindmiddel wat in BSF20 en BSF25 gebruik word, kon egter die resultate beïnvloed het. Al die eenheidsdigthede het betekenisvol van mekaar verskil, waar digtheid met elke hoër insluitingsvlak van SSV pre-papie toegeneem het. Die kontrole (SSV0) en SSV5 het ‘n sinksnelheid van 0 (drywende korrels) gehad. Die voer het egter eers begin sink met insluitingsvlakke vanaf 10 % SSV pre-papie meel (SSV10, 4.44 cm/s). Vinniger sinkingsnelhede was waargeneem met toenemende insluiting vlakke. Wateropname was oor die algemeen hoër in SSV0 in vergelyking met ander behandelings van al die tye, behalwe vir die kortste onderdompelingstyd (5 min), waar SSV15.6 (288.60 %) ‘n betekenisvolle hoër wateropname as al die ander diëte gehad het. As gevolg van die disintegrasie van korrels met die verloop van tyd, was die wateropname uitgedruk as 'n persentasie van die oorblywende voer eerder as die aanvanklike hoeveelheid wat gebruik is. Die oorblywende voer was gebruik om die desintegrasietempo te bepaal. Faktore soos die lipiedinhoud en interaksies tussen bestanddele kon indirek die voergehalteparameters beïnvloed het. Die studie het nie genoeg bewyse gelewer om die eis te verifieer dat SSV pre-papie meel as 'n lewensvatbare alternatiewe proteïenbron gebruik kan word vir konvensionele bronne in Nyl tilapia voer nie,as gevolg van veranderlikheid in waterstabiliteit en insluiting van 'n bindmiddel wat nie in ag geneem is nie. Daar is egter gevind dat insluitingsvlakke van tot 25% gebruik kan word sonder enige effek op die liggaamsamestelling en tot 5 % sonder om die groeiparameters in gedrang te bring. Daarom kan die aanbeveling gemaak word om 'n insluiting vlak van 5 % SSV pre-papie meel te gebruik om die groeiprestasie te handhaaf. Vir toekomstige studies word dit aanbeveel om ontvette SSV pre-papie meel te gebruik vir moontlike hoër insluitingvlakke van die meel vir vergelykbare groeiverslae relatief tot konvensionele proteïenbronne – aangesien die hoër lipiedinhoud die bindingsvermoë en gehalte van die voer nadelig kan beïnvloed.

Please refer to this item in SUNScholar by using the following persistent URL: http://hdl.handle.net/10019.1/103820
This item appears in the following collections: