Material Selection for Extruded Shrimp Feed Under Moisture and Shear Constraints

Extruded shrimp feed production requires physical characteristics fundamentally different from those of floating freshwater fish feed. The pellet must maintain structural integrity when submerged in marine or brackish water for 60–120 min, because shrimp are slow benthic feeders that manipulate feed particles over extended periods. Shrimp feed is manufactured as sinking or slowly sinking dense pellets, with bulk density often in the range 550–700 g/L and water stability index above 85% after 60 min in 15 ppt salinity seawater at 28 °C when measured by static immersion retention on a 500 µm sieve. These requirements place constraints on moisture management, shear input, and raw material selection that are more severe than those for low-density floating feed. The extruder barrel transforms a high-moisture, high-starch mass into a continuous viscoelastic melt; specific mechanical energy input, measured in kJ/kg dry feed, determines the extent of starch gelatinization, protein denaturation, and lipid dispersion. Excessive SME accelerates wear and disperses lipids to free surfaces, while insufficient SME fails to develop the continuous cooked starch matrix necessary for water-stable pellets. Material selection therefore cannot be separated from the narrow processing window created by the intersection of moisture, shear, and temperature.

What Moisture Levels Shift Torque and Specific Mechanical Energy Demand During Preconditioning and Melt Conveying?

In extrusion of shrimp feed, moisture content at conditioner discharge is adjusted to 24–30 g/100 g wet basis for formulations containing 30–42 g/100 g total starch. At these moisture levels, the melt viscosity in the metering zone of a co-rotating twin-screw extruder with L/D 25:1 and screw diameter 52 mm is reduced sufficiently to maintain SME between 110 kJ/kg and 180 kJ/kg. When preconditioner moisture falls below 22 g/100 g, the dynamic viscosity of the melt can increase by a factor of 1.4–2.0, and SME rises above 210 kJ/kg. Such conditions generate melt temperatures above 135 °C and cause measurable loss of available lysine in fishmeal-based formulas. Conversely, moisture above 32 g/100 g reduces SME below 85 kJ/kg and delays starch gelatinization because the excess water competes for thermal energy and reduces die pressure to 12–18 bar, producing weak pellets with poor water stability. The operational window is therefore only 6–8 g/100 g wide in many formulations. In the preconditioner, steam injection at a flow rate of 180–220 kg/h into a 1,500 kg/h dry feed line adds both heat and water; the actual water uptake depends on particle porosity and initial moisture. For soybean meal with initial moisture 11 g/100 g, steam condensation can be non-uniform, creating local moisture pockets that reduce melt homogeneity. Optical microscopy of unstained extrudate cross-sections from a Bühler preconditioner model DNDL-44 has shown residual starch granules located in regions with moisture below 20 g/100 g. Published data for this specific configuration is limited, but the practical consequence is that particle size distribution must be controlled to d50 250–450 µm to ensure uniform moisture penetration.

Starch gelatinization in low-moisture extrusion does not follow the same kinetic path as atmospheric cooking. In a closed extruder barrel, moisture availability, shear rate, and pressure act simultaneously on starch granules. Wheat and tapioca starches are preferred because their gelatinization onset temperatures at moisture content 25 g/100 g fall between 58 °C and 68 °C when measured by differential scanning calorimetry at a heating rate of 10 °C/min according to AACC Method 76-21.01. High-amylose maize starch, with onset temperature above 75 °C under the same moisture level, may fail to fully gelatinize when die temperature is capped at 120 °C to protect shrimp attractants and heat-sensitive amino acids. The degree of gelatinization in extruded shrimp feed is 85–95% when SME reaches 140–180 kJ/kg and barrel moisture is 25–28 g/100 g. Below 75% gelatinization, pellet water stability measured by immersion in 15 ppt seawater at 28 °C for 60 min drops below 80% retention on a 500 µm sieve. The amylose-to-amylopectin ratio also controls retrogradation and bonding; a wheat starch with amylose content 22–26 g/100 g provides adequate binding without producing brittle pellets after drying. Tapioca starch, with amylose 17–20 g/100 g, produces high cohesiveness but can form a sticky melt that increases torque if added above 12 g/100 g of formula. Starch damage created by pre-grinding to d50 160–220 µm further reduces gelatinization onset temperature but also increases water absorption; this interaction is beneficial when preconditioner residence time is short.

Representative gelatinization and water stability data for starch sources at 25 g/100 g moisture, with SME controlled to 160 kJ/kg:

Protein Shear Sensitivity and Maillard Degradation in High-Torque Extrusion of Fishmeal-Based Formulas

Shrimp feed formulas commonly contain 300–450 g/kg fishmeal, 150–250 g/kg soybean meal, and 30–80 g/kg squid or krill meal. The fishmeal fraction is already denatured to some extent during rendering, but residual protein solubility in 0.2% potassium hydroxide measured by ISO 5983-1:2005 can decline from 70–85 g/100 g to 40–55 g/100 g when SME exceeds 200 kJ/kg and melt temperature exceeds 130 °C. Soluble protein loss above this threshold correlates with reduced pellet durability after drying. Soybean meal introduces trypsin inhibitors that are adequately inactivated only when barrel temperature exceeds 121 °C for 15–20 s and moisture exceeds 23 g/100 g. However, the same temperature range promotes Maillard reactions between lysine and reducing sugars present in wheat flour or molasses, reducing available lysine by 8–15% when reducing sugar content exceeds 2 g/100 g of formula. The shear sensitivity of squid meal is particularly pronounced because its protein fraction consists largely of myofibrillar proteins that orient under high shear and form a fibrous network. This orientation can improve water stability if SME is held between 120 kJ/kg and 160 kJ/kg and moisture is 24–27 g/100 g; above 180 kJ/kg, the same network fragments and the pellet surface becomes rough, increasing fines by 1.5–2.5 percentage points after top-dressing. Operational data from a Clextral BC-45 twin-screw extruder with L/D 27:1 indicate that torque spikes of ±8–12% occur when unground squid meal particles above 800 µm enter the mixing zone; milling to d50 300 µm reduces these fluctuations.

The selection of an aqueous binder is not governed solely by water stability. Wheat gluten added at 20–50 g/kg increases wet melt extensibility but also raises low-moisture viscosity; at 25 g/100 g moisture, a formula with 40 g/kg wheat gluten can increase die pressure from 22 bar to 34 bar in a 2.0 mm die insert, while SME rises by 15–20 kJ/kg. Carboxymethyl cellulose sodium salt with degree of substitution 0.7–0.9 and viscosity 1,500–2,500 mPa·s at 20 g/L and 25 °C forms a cold-swelling hydrocolloid network that reduces oil migration to the pellet surface during vacuum coating. However, at addition levels above 10 g/kg, the same polymer creates a slippery melt layer on the screw elements, lowering frictional heat transfer and producing a pellet with a glossy but weak outer layer. Sodium alginate at 5–15 g/kg requires calcium ions to crosslink; in marine shrimp feed formulas, calcium carbonate at 20–40 g/kg provides sufficient Ca²⁺ in the hydrated melt to form alginate gel networks that raise water stability retention from 72% to 88% after 120 min in 15 ppt seawater. The limitation is that alginate crosslinking increases die swell by 6–10% and requires die land length to be extended from 12 mm to 18 mm for a 2.0 mm die to maintain diameter tolerance. Gelatinized starch from pregelatinized tapioca at 30–60 g/kg functions as an instant binder but can increase barrel fill in the first screw zone; feed intake must be reduced by 5–8% to prevent over-torque.

When Lipid Addition Reduces Die Pressure but Creates Post-Expansion Oil Migration and Variable Bulk Density

Shrimp feed lipid content is frequently 60–100 g/kg in the pre-extrusion mixture, with additional 20–40 g/kg applied post-extrusion by vacuum coating. Lipids act as internal lubricants in the melt, lowering apparent viscosity and reducing SME by 10–20% when pre-extrusion fat increases from 40 g/kg to 90 g/kg. In a Wenger TX-52 twin-screw extruder with 52 mm screws and a 2.5 mm die, an increase from 60 g/kg to 90 g/kg fish oil at 26 g/100 g moisture can lower die pressure from 28 bar to 18 bar and reduce pellet bulk density from 620 g/L to 580 g/L. This pressure drop is not linear; above 90 g/kg fat, the melt loses cohesive strength and starch-lipid complexes form at the starch granule surface, reducing water stability retention below 75% after 60 min. The melting point and free fatty acid content of the lipid also matter. Fish oil with free fatty acid content above 5 g/100 g and moisture in the melt above 28 g/100 g promotes hydrolysis during extrusion, releasing short-chain aldehydes that impart rancid off-notes and reduce shrimp palatability. Saturated fat sources such as poultry fat with slip melting point 30–34 °C reduce die pressure less than fish oil but are less likely to migrate; partial replacement of fish oil with fully hydrogenated soybean oil at 10–20 g/kg can stabilize die pressure fluctuations to ±1.5 bar. Vacuum coating after extrusion at 0.7–0.8 bar vacuum and 60–70 °C pellet temperature is preferred when total lipid exceeds 100 g/kg, because it preserves water stability and avoids the shear-induced lipid smearing observed in high-fat melts.

Mineral Filler Effects on Abrasion, Melt pH, and Pellet Density Under High-Shear Conditions

Calcium carbonate, magnesium carbonate, dicalcium phosphate, and zeolite are common mineral fractions in shrimp feed. At addition levels of 15–30 g/kg, calcium carbonate with median particle size 10–20 µm acts as a nucleating agent during bubble growth, increasing pellet density by 5–8% and reducing expansion. Above 40 g/kg, it becomes an abrasive filler that raises screw and barrel wear rates. Published equipment maintenance data for a Bühler twin-screw extruder with L/D 24:1 indicates that barrel liner replacement intervals can decrease from 8,000 operating hours to 4,500–5,000 operating hours when calcium carbonate content exceeds 40 g/kg and screw speed is above 300 rpm. Zeolite clinoptilolite at 10–20 g/kg absorbs water and buffers ammonia; it can raise the melt pH from 6.2 to 6.6 and reduce the severity of Maillard reactions, but it also increases apparent viscosity at a given moisture content, requiring an additional 1–2 g/100 g water to maintain SME within 120–160 kJ/kg. Dicalcium phosphate anhydrous with d50 50–80 µm provides phosphorus but has low water-binding capacity; its inclusion above 15 g/kg can create local shear peaks in the mixing zones of the extruder, increasing torque standard deviation by 10–15% relative to a calcium carbonate control. The effect of mineral filler on water stability is indirect: fine insoluble particles fill micropores in the gelatinized starch matrix and reduce water ingress, but only when particle size is below 50 µm and dispersion is complete. Poor dispersion due to inadequate mixing leads to mineral-rich zones that become fracture initiation sites during drying and subsequent handling.

Process capability is also bounded by raw material moisture sorption. Soybean meal at 25 °C and 60% RH equilibrates to 11–12 g/100 g moisture, while fishmeal at 60% RH can equilibrate to 8–10 g/100 g depending on salt content. If pre-extrusion storage is not humidity-controlled, batch-to-batch moisture variation of ±1.5 g/100 g is enough to shift SME by ±10 kJ/kg and alter pellet water stability by ±5 percentage points. Pre-drying of fishmeal to below 9 g/100 g is required before fine grinding when ambient RH exceeds 60%, otherwise the material clogs a 250 µm hammermill screen and produces oversized particles that generate torque spikes in the extruder. The same raw materials must also be screened for mycotoxins and biogenic amines, because heat-stable contaminants are not destroyed by extrusion. For shrimp feed specifically, histamine levels above 50 mg/kg in fishmeal have been associated with reduced feed intake in controlled feeding trials, and extrusion at 120 °C does not degrade histamine. These operational boundaries define the material selection envelope as tightly as the rheological constraints.

Analytical methods and regulatory acceptance thresholds for shrimp feed raw material control: