Rainbow trout Rearing Oncorhynchus mykiss on corn fermented protein product
by Professor Simon J Davies, International Aquaculture Editor, Emeritus Professor, Harper Adams University.
& Derek Balk and Melissa Jolly-Breithaupt, Flint Hills Resources, USA
Rainbow trout production contributes significantly to the global salmonid industry and is an iconic species of high value and acceptability. It is reared extensively in many temperate regions of the world such as in the USA, Canada, Norway, Denmark, and UK and in most areas of Europe and regions of Latin America like Mexico and, Chile as well as in parts of Australia.
The global rainbow trout market was estimated to be valued at US$3,524.08 million in 2018 and projected to reach US$4,998.19 million by 2025, at a CAGR of 5.14 percent during the 2018 to 2025 period with attaining well over one million tons production.
Rainbow trout is a carnivorous fish and requires diets containing a high level of protein and energy as oils (typically 45 and 25 percent) in commercial feeds. Consequently, aquafeed production must keep expanding in order to meet demand. Feed formulations for salmonids have traditionally relied on fishmeal to provide the bulk of dietary protein. Although total fishmeal use in aquaculture feeds increased each year until 2007-08, the percentage of fishmeal in feed formulations has decreased for most species by between 35 and 50 percent (Tacon and Metian, 2008). Alternative feed ingredients such as soybean meal, soy protein concentrate, canola meal, canola protein concentrate, corn gluten meal, cottonseed meal, peas, and wheat gluten meal have been investigated to replace fishmeal and reduce the cost of fish production (Gatlin et al, 2007).
Notably, many plant-derived protein ingredients contain anti-nutrients such as phytic acid and protease inhibitors that interfere with nutrient assimilation. Plant protein-based diets may also contain lower levels of limiting amino acids such as methionine, lysine and threonine than fishmeal-based diets. However, supplementing the limiting EAA's with crystalline sources can restore growth rates in fish, to some extent (Cheng et al, 2003). Consumer's today have raised serious ethical concerns about the sustainability of soybean meal, primarily driven by the issue of deforestation.
From 2000 to 2020, the US biorefinery industry has grown from 56 to 209 large scale fermentation facilities for the production of alcohol also known as ethanol from different cereal grains. The dry milling ethanol production resulted in 3.3 billion pounds of corn distillers' oil and 29.4 million metric tons of dried distillers' grains in 2020 (Renewable Fuels Association, 2020).
Among all the DDG products from different grains, corn DDGS is the predominant one. Corn DDGS is produced in ethanol plants by using a dry-grind method (Overland et al, 2013). Conventional DDGS contains a moderate level of crude protein (24 -32 percent) compared to fishmeal and soy protein products and has less phosphorus than fishmeal (Gatlin et al, 2007).
Use of DDGS has been studied in the diets of many aquaculture species including rainbow trout (Cheng et al., 2003; Cheng and Hardy, 2004; Barnes et al, 2012). Also, diets containing 10 or 20 percent DDGS appeared to reduce growth in rainbow trout even with supplementation of essential amino acids and phytase due to high fiber content (Barnes et al, 2012a).
In another study, DDGS was included in rainbow trout diets up to 22.5 percent without affecting growth when lysine and methionine were supplemented (Cheng and Hardy, 2004). Stone et al (2005) stated that if the crude protein content could be increased and indigestible fiber decreased, the inclusion level of DDGS may be increased in fish feeds. This can be achieved by fractionating and removing non-fermentable fractions before or after ethanol production. Pre-fermentation fractionation creating a high protein DDG (HPDDG, 42 and 45 percent crude protein) has been evaluated in rainbow trout with contrasting results (Barnes et al, 2012; Overland et al, 2013).
The result of post-fermentation mechanical separation
NexPro fermented protein, a product of U.S. based Flint Hills Resources, is the result of post-fermentation mechanical separation of the DDG product utilising a patented technology called Maximized Stillage Co-Products. Fractionating the product post-fermentation allows the fermentation process to assist with separation and weaken the cellular wall structure of the fibrous fractions and concentration of inactive Saccharomyces cerevisae yeast, which is utilised for the production of alcohol.
NexPro® has a superior crude protein (~50 vs ~28 percent), lower crude fiber levels and improved nutritional composition compared to traditional DDGS. As a result, (NexPro®) will likely compete with soy protein concentrate, corn protein concentrate, corn gluten meal and brewer's yeast as an ingredient in fish feed formulations.
This study evaluated NexPro® as a sustainable protein source in feeds for rainbow trout by replacement of soy protein concentrate (SPC) in a balanced series of diets, including other ingredients and fishmeal. The chosen parameters of the study include growth performance, feed efficiency, digestibility and nutrient retention, the latter of which is important from the perspective of reducing nutrient losses from fish farms that cause environmental impact (such as phosphorus and nitrogen).
Flint Hills Resources supplied NexPro® corn fermented protein to Bozeman Fish Technology Center (BFTC), Bozeman, Montana, for experimental feed production as described below. First, the digestibility trial and then the growth trial were conducted by the University of Idaho's Aquaculture Research Institute, specifically the Hagerman Fish Culture Experiment Station (HFCES) in Hagerman, Idaho. The product was analyzed at HFCES for nutrient composition.
Diet composition and application
Experimental feeds: In vivo apparent nutrient digestibility of NexPro® was determined by feeding separate groups of sub-adult rainbow trout a diet containing the product at 30 percent. A reference diet (10-kg batch) containing practical ingredients and 0.1 percent indigestible inert marker (yttrium oxide) was prepared at the HFCES. Test diets containing 30 percent NexPro® and 70 percent reference diet mash on dry-matter basis were prepared. Both diets were cold-pelleted with a California pellet mill fitted with a four millimetre die. Pellets were dried in a forced-air dryer at 35 °C for 48 hrs. Samples of each diet were taken for proximate composition and mineral analyses, including yttrium analysis.
Fish maintenance and feeding regime: Rainbow trout from the HFCES in-house broodstock (House Creek strain) were used for the study. Twenty-five fish (~250 g) were stocked in four 145-L tanks, each supplied with 12 L min-1 of constant temperature (15 °C) spring water supplied by gravity to the fish rearing laboratory.
Each of the reference and test diets was randomly assigned to two tanks of fish. Fish were fed their respective diets twice daily, at 0830 to 0900 and 1530 to 1600h to apparent satiation for one week. On day 4 and 8, fish in each tank were lightly anaesthetized using tricaine - methanesulfonate (MS-222, 100 mg L-1, buffered to pH 7.0), removed from water for 30 to 60 seconds, and feces gently expelled using light pressure on the abdomen near the vent, a process called 'stripping'.
Experimental feeds: All experimental diets for the growth trial were formulated with a feed formulation software (WinFeed 2.8, Cambridge, UK) after the nutrient digestibility data were available for NexPro®.
A control diet plus five experimental feeds were formulated to contain 40 percent digestible protein and 17.2 MJ/kg digestible energy, three percent lysine and ~0.8 percent digestible phosphorus (as-is basis).
The feeds were formulated as follows:
Diet 1: Control – standard level of fishmeal in commercial trout feeds: Diets 2 to 5 (25 to 100 percent incremental replacement of SCP with NexPro®); Diet 6: 25 percent replacement of SPC with dried brewer's yeast (BY) on a crude protein basis.
All diets met or exceeded the minimum nutrient requirements of rainbow trout (NRC, 2011). Dried brewer's yeast (Saccharomyces cerevisiae) was also tested by replacing 25 percent SPC on a crude protein basis so as to compare it with the control and the diet with NexPro® replacing 25 percent SPC on a crude protein basis. Diets were produced by extrusion pelleting similar to commercial fish feed production technology. The nutritional composition the test products are presented in Table 1. Crude protein content of NexPro® (50.87 percent) was higher than that of DDGS (28.36 percent) whereas crude fat was lower in NexPro® (4%) than that in DDGS (11.6%). Energy content was higher in DDGS than in NexPro®.
The approximate composition and energy content of diets used in the growth trial are presented in Table 4, whilst the mineral composition of the diets is presented on as-fed basis in Table 5.
Fish and feeding: Rainbow trout fingerlings, hatched from eggs purchased from a commercial source (TroutLodge, Sumner, WA) were used in the study. Thirty fish (initial average weight: 15.6 g) were stocked into each of 18, 145-L tanks. Each tank was supplied with 10 to 12 L/min of constant temperature (15 °C) spring water supplied by gravity to the fish rearing laboratory.
In a completely randomized design, each of the six experimental diets was randomly assigned to triplicate tanks within the laboratory system to account for any tank position effects. Each diet was fed by hand to respective tanks of fish to apparent satiation, three times per day and six days per week for 12 weeks. Photoperiod was held constant at 14 h light: 10 h dark.
Proximate composition (moisture, protein, fat and ash) of feed, whole-body fish and fecal samples were determined using AOAC (2002) procedures.
Calculations of apparent digestibility coefficients of diets
Apparent digestibility coefficients (ADC), for both diets and NexPro®, for dry matter, organic matter, protein, lipid, energy and minerals, (including phosphorus), were calculated using the formula described by Bureau et al. (2002): Employing the live-weight and feed consumption data, and indices calculated as by Hardy and Barrows (2002)
Statistical analysis of data: Data were tested for normality and homogeneity of variance prior to one-way analysis of variance (ANOVA). When required, data were transformed to achieve normal distribution and subjected to Tukey's HSD test to separate the means at a significance level of P<0.05. In case of non-homogeneous variance, Welch's ANOVA was performed. If significant differences were found, Tukey's test, which corresponded to the Games-Howell test, was conducted to separate the means. In case of non-normal distribution, the non-parametric Kruskal-Wallis test was performed. All statistical tests were performed with SAS 9.3 software.
Feed conversion ratios were very good
In the digestibility trial the nutrient composition of the reference diet used for digestibility trial is presented in Table 2. The diet contained 44.7% crude protein and 17.8% crude fat which are typical of diets used for a growth trial in our laboratory. Apparent digestibility coefficients of nutrients for DDGS in rainbow trout are presented in Table 3. Even though dry matter digestibility (50.5 percent) was lower, crude protein digestibility was very good (86.4 percent). ADC for energy was slightly low (59.6%). Minerals especially Mg, P, K, Cu and Zn were highly digestible.
In the growth trial the rainbow trout juveniles were fed diets containing graded levels of NexPro® (NXP) and a single level of brewer's yeast for 12 weeks. Fish readily accepted the experimental diets. Overall, fish were robust with no abnormalities or deformities. Growth and feed utilization indices of the fish are presented in Table 6.
Mean final weight of fish was significantly different among the dietary groups (P<0.05). Fish fed the 75NXP (240 g) diet had significantly higher final weight than that fed BY diet (217 g). However, there was no significant difference in final weight among the fish fed NexPro® diets.
Weight gain was highest in fish fed 75NXP diet (224 g/fish) than in fish fed BY diet (202 g/fish) and they were significantly different. There were no significant differences in percent weight gain, specific growth rate, daily growth index, survival, feed consumption of fish or FCR among the dietary treatment groups after 12 weeks of feeding (P>0.05).
Mean percent weight gain was the highest in the 75NXP (1421) and the lowest in the BY (1296). Specific growth rate ranged from 3.14 percent/day (BY) to 3.24 percent/day (75NXP). Survival was high in all dietary treatment groups (93.3 to 100 percent) at the end of 12 weeks. Feed consumption per fish varied from 179 g (BY) to 209 g (100NXP) whereas daily feed consumption ranged from 1.83% body weight/day (75NXP and BY) to 2.02 percent bodyweight/day (100NXP).
Feed conversion ratios were very good for all the feeds (0.87 to 0.97). Protein efficiency ratio was significantly lower in fish fed 100NXP diet (2.20) than in fish fed other diets (2.33 to 2.39). Condition factors of fish were high across the diets (1.55 to 1.63) and were not significantly different among the dietary treatments.
Whole body proximate composition and mineral composition of the fish fed the experimental diets are presented in Table 7. This did not vary significantly among the dietary treatments (P>0.05). Phosphorus ranged from 0.365 percent (75NXP) to 0.40 percent (Control) and decreased as HP 330 level increased in the diets. Whole-body iron level varied from 14.5 ppm to 18.3 ppm. Zinc level varied from 21.3 ppm (25NXP) to 31.0 ppm (100NXP).
Nutrient retention of juvenile rainbow trout fed the experimental diets for 12 weeks are presented in Table 8. There was no significant difference among the dietary groups for fat and protein retention (P>0.05). Fat retention ranged from 71.62 percent (100NXP) to 82.3 percent (Control).
Protein retention in rainbow trout ranged from 37.8% (100NXP) to 40.8% (50NXP). Energy retention was significantly higher in Control (45.3%) and 75NXP (46%) groups than in 100NXP group (40.8%). Retention values for calcium (18.6 to 23.1 percent) and phosphorus (26.4 to 29.8 percent) were not significantly different among the dietary treatments (P>0.05).
An excellent candidate as a protein source in fish feed
Corn fermented protein, being a combination of recovered corn protein and spent yeast, is rich in protein and provides a better amino acid profile than traditional corn gluten meal, especially lysine. NexPro® has lower carbohydrate and crude fiber content than DDGS. Moreover, levels of minerals such as phosphorus, iron and zinc are substantially higher in NexPro® than in DDGS. NexPro® is also different than the high protein dried distiller's grain in that it's produced after ethanol production whereas HPDDG is produced via a fractionation before ethanol production. NexPro has higher crude protein (50 vs. 45 percent) than HPDDG but has slightly lower lysine (1.93 and 2.1 percent) and similar methionine (0.83 and 0.89 percent) levels. All these favorable characteristics of NexPro® make it an excellent candidate as a protein source in fish feed.
The ADC values obtained were similar or higher than the values obtained by Cheng and Hardy (2004) for different proximate categories of DDGS and were used for the formulation of diets used for growth trial later. The present study evaluated NexPro® as a replacement for soy protein concentrate in rainbow trout diets while levels of fishmeal, other animal proteins, and soybean meal were kept constant to avoid any confounding effects of variable levels of other protein sources.
Fish grew well with low FCRs (0.87 to 0.97) similar to what we generally observe with good commercial diets used in our laboratory. Even though significant differences in final weight or weight gain per fish existed among the dietary treatments, percent weight gain or specific growth rates were not significantly different. Also, there was no difference between the control and NexPro® containing diets in terms of growth performance. Similar findings were also obtained by Cheng and Hardy (2004), when 22.5 percent DDGS was included in rainbow trout diets with lysine and methionine supplementation and replaced 75 percent of fishmeal.
This study also corroborates with the findings of Overland et al (2013) who successfully replaced a mixture of plant proteins such as SPC, sunflower meal and rapeseed meal with 22.5 and 45 an excellent candidate as a protein source in fish feed high protein dried distiller's grain (HPDDG) in the diets of 143 g rainbow trout.
In their study, just like in the present study, fishmeal level was constant across the diets (~21 percent). In another study with 34 g rainbow trout, similar results were obtained when 10% or 20% HPDDG was included by replacing fishmeal in the diets (30-40 percent fishmeal) but supplementing essential amino acids including lysine and methionine (Barnes et al, 2012).
Rainbow trout fed the highest level of NexPro (24 percent, 100NXP) tended to consume more feed even though not significantly more than the control group. As the level of NexPro® increased (0-24 percent) in the diet, feed intake appeared to increase marginally indicating no palatability issue associated with tested levels of NexPro®. However, protein efficiency ratio (weight gain per unit protein consumed) of fish was significantly lower in the 100NXP group than the other dietary groups.
Protein retention was numerically lower and energy retention was significantly lower in the 100NXP group than in the control group. The results suggested that when fish were fed the highest level of NexPro® (24%), they tended to eat more but not utilise the nutrients as efficiently as the control group. This is in contrast to the findings of Overland et al. (2013) who observed no differences in feed intake with 22.5 or 45 percent HPDDG in the diets of rainbow trout.
Also, they did not see significant differences in protein or energy retention among the dietary treatments despite a decrease in protein digestibility and an increase in energy digestibility. In general, protein and phosphorus retention were higher across the treatments in that study than in the present study due to lower dietary crude protein and phosphorus levels in the earlier study. In that study, HPDDG replaced a mixture of SPC, sunflower meal and rapeseed meal whereas NexPro® replaced only SPC in the present study. Soy protein concentrate is a highly digestible protein (90-95 percent crude protein digestibility).
Even though NexPro® replaced SPC in terms of digestible protein incrementally in the diets in the present investigation, the values used were actually of DDGS in the absence of digestibility values for NexPro®. Corn co-products' quality and nutrient profile vary widely due to the grain source and processing methods employed (Liu, 2011; Welker et al, 2014) but NexPro® has consistently uniform quality control and specifications.
Brewers' yeast in the diet (6.9 percent) reduced the weight but not the growth rate of fish as compared to control diet. It might have slightly reduced the palatability of the diet causing marginally lower feed intake that was not apparent during feeding of fish.
In summary, NexPro® corn fermented protein can effectively replace SPC up to 100 percent in a rainbow trout diet without significantly affecting growth performance or feed efficiency.
An inclusion at 18 to 24 percent in the diet in the presence of other good quality protein sources was deemed to be optimal under the trial conditions. NexPro® is a viable solution for mitigating the 'protein gap' in advanced trout feeds for a sustainable trout industry. The relative prices of NexPro® and SPC and effects of NexPro® on feed conversion ratio will likely dictate decisions by feed formulators as to appropriate levels of SPC and/or NexPro® in rainbow trout diet formulations.