Supporting Information
The addition of transglutaminase improves the physical quality of extruded fish feed
Julia Wolskaa,b,*, Jan Jonkersa, OlleHolstb, Patrick Adlercreutzb
aSkretting Aquaculture Research Centre, PO Box 48, 4001 Stavanger, Norway
bBiotechnology,Center for Chemistry and Chemical Engineering, Lund University, POBox 124, SE-221 00 Lund, Sweden
*Corresponding author at: Skretting ARC, PO Box 48, 4001 Stavanger, Norway. Tel.: +47 51 82 55 22; fax: +47 51 82 55 01. E-mail:
Methods
Feed production
All feeds were produced at Skretting Aquaculture Research Centre (ARC) Feed Technology Plant (Stavanger, Norway). One diet was used for the entire study (Supplementary Table 1). All dry ingredients were prepared in batches of 150 kg, premixed in a vertical Nauta type mixer (Skretting ARC, Stavanger, Norway) and ground in a Dinnissen hammer mill (Dinnissen BV, Sevenum, The Netherlands) with a screen size of 1.0 mm. Next, the ingredients were mixed in the main mixer, a Dinnissen horizontal ribbon mixer (500LTR, Dinnissen BV, Sevenum, the Netherlands), for 7 min. Subsequently, the feed mash was conditioned with water and steam in a double-shaft, differential diameter conditioner (Skretting ARC, Stavanger, Norway) and extruded using a twin screw extruder (TSE 36 HC Thermo Scientific, Thermo Fisher, UK) with a feeding rate of 75 kg h−1 and 6-mm diameter die hole. The temperature of the first extruder section was set at 50 C, and sections 2–7 at 85 C. The screw speed was adjusted to between 450 and 500 rpm in order to obtain the correct bulk density of each product. Fluids added to the process are listed in Supplementary Table 2.
Extrudates were dried in a batch drier (Skretting ARC, Stavanger, Norway) at 90 C to the same moisture content (90 g kg−1). The dry kernels were filled with oil in a 6-L rotating vacuum coater (6RRVC Forberg, Oslo, Norway), where the exact amount of oil mix prepared according to Supplementary Table 1 was infused into the pellets further on in the process, referred to as the coating step. They were then cooled in a custom-designed cooler (Skretting ARC, Stavanger, Norway) and packed.
Experimental design
Experiment 1: the effect of the mode of application and the processing conditions on the physical quality of feed modified with TGase
Two different modes of enzyme application were evaluated: addition into the main material mixer before extrusion or into the vacuum coater.
For the first application, six batches of dry material mix were produced; for three batches, TGase was added to the dry ingredients in the main mixer at a dosage of 20 g kg−1 dry material mix (w/w); no enzyme was added to the other three batches (Supplementary Table 2). For both enzyme levels, three different moisture levels were used in the production process (Supplementary Table 2), which resulted in six different products. These feeds were dried and coated as described in the previous section. However, some of the dry kernels without added enzyme were used for the enzyme treatment in the coater as described below.
For the second application, the enzyme was dispersed in oil and added to the dry kernels during the coating step. Forty grams of TGase was dispersed in 857 g of oil mixture (Supplementary Table 1) with intensive stirring and infused into 2000 g of dry kernels, resulting in three different products.
Experiment 2: finding the optimal dosage of TGase
Further optimization was done by modifying the dosage of TGase in the coating process. For this purpose, kernels without enzyme, produced under all three processing conditions, were coated with oil containing three levels of dispersed enzyme preparation (2.5, 5 and 10 g of TGase per 1 kg of kernels), resulting in 9 more products.
Analysis of amino acid composition
The amino acid profile of the raw materials and feed was measured according to European Directive 98/64/EC.
Test of water stability
The method measures the water stability of wet pellets after they have been soaked in water for 24 h. Thirty grams of sample were soaked in 200 mL of water for 24 h at room temperature. The sample was then stirred on a magnetic stirrer at 450 rpm for 20 min with a 40-mm stirring bar. The content of the beaker was placed on a sieve and immediately rinsed with 1.0 L of clean cold tap water. All the retained material was then transferred onto an aluminium tray and dried in an oven at 105 C for 12 h. Water stability was calculated as the percentage of dry matter remaining after the test (dry matter basis).
Supplementary Table 1
Diet formulation.
Raw material / Supplier / Inclusion (g kg−1)Wheat / Felleskjøpet AS, Stavanger, Norway / 100
Wheat gluten / Cargill Nordic A/S, Charlottenlund, Denmark / 50
Sunflower meal / Linas Agro, Panevezys, Lithuania / 40
Low temperature dried fish meal / Welcon, Egersund, Norway / 150
Faba beans dehulled / Negoce Soufflet, Nogent sur Seine, France / 60
Soy protein concentrate / Imcopa SA, Araucaria, Brazil / 300
Rapeseed oil / Scanmills A/S, Kolding, Denmark / 100
Fish oil / Vinnslustöðin, Vestmannaeyjar, Iceland / 200
Supplementary Table 2
Content of glutamic acid and lysine in different raw materials and final feed.
Raw material / Glutamic acid (g kg−1) / Lysine(g kg−1)
Fish meal / 91 / 50
Soy protein concentrate / 119 / 39
Wheat gluten / 289 / 14
Corn gluten / 134 / 10
Sunflower meal / 78 / 15
Final feed / 68 / 21
Supplementary Table 3
Experimental design for addition of enzyme to the dry mix and vacuum coater
Processing conditions / Steam added to preconditioner (kg h−1) / Water added to extruder (kg h−1)Control feed (no enzyme added)
High moisture / 6.0 / 9.8
Medium moisture / 6.0 / 7.5
Low moisture / 0.0 / 7.5
Enzyme added to the dry mix (20 g kg−1 w/w)
High moisture / 6.0 / 9.7
Medium moisture / 6.0 / 7.5
Low moisture / 0.0 / 7.5
Enzyme added to the vacuum coater (20 g kg−1 of the kernels)
High moisture / 6.0 / 9.8
Medium moisture / 6.0 / 7.5
Low moisture / 0.0 / 7.5
For all the diets, water was added to the preconditioner at 9.8 kg h−1.
SupplementaryFig. 1Breaking distance of feed produced without TGase (solid bars), with 20 g kg−1 (w/w) TGase added to the dry mix (dotted bars) and with 20 g kg−1 (w/w) TGase added to the coater (dashed bars). Each type of feed was produced under three different processing conditions: high moisture, medium moisture and low moisture (Supplementary Table 3). Means (n=3) with the same letter within one treatment are not significantly different (p>0.05) according to the Tukey HSD test
Supplementary Fig. 1