Effect of Starch Sources on Properties of Extrudates Containing DDGS

Abstract

Various levels of DDGS (20, 40, and 60% wb) were blended with starch sources (cassava, corn, and potato), and other ingredients to produce an iso-nitrogenous feed (28% protein) at varied moisture contents (15, 20, and 25% wb). The feed blends were extruded in a single-screw extruder at a preset screw speed of 130 rpm (13.6 rad/s) with three temperatures profiles 90–100–100°C, 90–120–120°C, and 90–140–140°C. The effect of these variables on processing conditions (extruder torque and die pressure) and other extrudate properties (expansion ratio (ER), unit density (UD), color (L*, a*, and b*), sinking velocity (SV), water absorption, water solubility, and pellet durability indices (PDI)) were analyzed. For all the three starch extrudates, changing the levels of DDGS, feed moisture content, and extruder barrel temperature had a significant effect on SV, PDI, a*, and b* values at α = 0.05.

Keywords:

• DDGS
• Extrusion
• Physical properties
• Processing parameters
• Starch

INTRODUCTION

Starch is a major functional ingredient that is responsible for the expansion of the extruded products. In extrusion industry, starch ingredients are used to produce puffed products, while other ingredients-namely proteins, fats, and fiber-act as diluents. Additionally, during the process of extrusion, the starch is plasticized with water, which is subjected to a specific mechanical and thermal energy treatment.[1] For better expansion, the minimum starch content was found to be 60–70%.[2] The expansion of starch mainly depends on its degree of gelatinization—temperature and moisture content are the two primary factors that influence gelatinization.[3] The amylose-amylopectin ratio is found to be more critical in determining the properties of starch-based extruded products. The higher the amylopectin content in the starch, the higher the expansion of the starch, and hence the expansion of product.[4]

Extrusion processing is used globally for the production, modification, and improvement of quality of various products such as: ready-to-eat cereals, snacks, pet foods, and aqua feed.[5] Extrusion cooking is a high-temperature, short-time process that performs multivariable unit operations; specifically, these are mixing, shearing, cooking, puffing, and drying in one energy efficient rapid continuous process.[6] A very small variation in processing conditions can affect the process variables, as well as the product quality.[7] During the process of extrusion cooking, the ingredient mixture undergoes numerous structural and chemical transformations, such as protein denaturation, starch gelatinization, degradation reactions of vitamins, pigments, and complex formation between amylase and lipids.[8] Heat treatment of food product inside an extruder enhances the digestibility due to the inactivation of enzymes, microbes, and many antinutritional factors present in feed materials to render the product sterile. The ingredient mixture generally becomes plastic during the process of extrusion cooking and expands or puffs while leaving the die as moisture in the material flashes out as steam.[6]

The expansion of the extrudates and its texture depends mainly on the interaction of shear, heat, and the moisture present in the extruder.[9] Moisture has a great influence on the extrudates' quality by affecting the cell structure and thus influencing the fragility of expanded products.[10] The quality of the extruded products is determined by the chemical and structural transformations in foods during extrusion cooking.[11] Many extrusion-processing variables can be controlled to obtain certain final product characteristics. Manipulation of independent variables (e.g., screw speed, screw configuration, die design, feed moisture content, temperature, and feed rate) produces changes to one or more of the dependent variables (e.g., residence time, mechanical energy input to the extruder, and thermal energy into the extruder), and consequently, affects the functional and physicochemical properties of the melt.[12]

Most of the ethanol produced in the USA today (≈5 billion gallons) is via dry grinding, with distillers dried grains with solubles (DDGS) as the main co-product, which is relatively high in protein and fiber but low in starch.[13] DDGS—a co-product of dry grind ethanol production—is mainly used as livestock feed, primarily in beef and dairy rations, but to a lesser extent in swine and poultry diets, as well. Extensive research has been conducted to determine the effect of moisture content on the extrudates properties with various kinds of starch and protein based feed materials.[14–16] Chemical modifications of starch in the extruder, namely substitution and stabilization have been studied by Mauro.[17]

Few studies have been conducted on extrusion processing with DDGS as a base material and production of DDGS based aquaculture feeds.[16,18–22] Chevanan etal.[20] conducted extensive research on extrudate physical properties and extruder processing parameters by varying the feed moisture content, extruder barrel temperature, and die dimensions, and they determined that the parameters mentioned had a significant impact on the extrudate and extrusion properties studied. In another study conducted by Chevanan etal.[21] presented the effect of changing the DDGS levels, feed moisture content, and extruder screw speed on the physical and nutritional properties of the extrudates, and the results from this industrial twin-screw extrusion study indicated that floating aquaculture feeds could be successfully produced using ingredient mixes containing ≤60% DDGS for the processing conditions studied. Another interesting point is that most of the studies reported in the literature[16,18–21] were based on cornstarch. Cornstarch had lesser amount of amylopectin (72%), which is responsible for the expansion of the extruded products compared to other starches sources.[23] Other starch sources like cassava (82–85%) and potato (79%) containing higher amount of amylopectin could result in extrudates with higher expansion and functional properties, when incorporated with DDGS and other ingredients. Furthermore, to date, a negligible attempt has been made to study the effects of various starch sources on the extrudate properties with DDGS as a base material with varying feed and extrusion parameters. Therefore, this study was undertaken with the objectives: (1) studying the effect of various starch sources (cassava, corn, and potato) on extrudates properties containing DDGS extruded in a single-screw extruder and thereby; and (2) determining the effect of various levels of DDGS, feed moisture content, and extruder barrel temperature on the resulting extrudates properties and processing parameters.

MATERIALS AND METHODS

Raw Ingredients and Blend Formulation

DDGS, supplied by Dakota Ethanol LLC (Wentworth, SD) was ground to fine powder of ≈425 μm size using a laboratory grinder (s500 disc mill, Genmills Inc., Clifton, NJ, USA). The cornstarch was procured from Cargill dry ingredients (Paris, IL, USA), potato and cassava starch came from AKFP (American Key Food Products, Closter, NJ, USA), and soy flour from Cargill soy protein solution (Cedar Rapids, IA, USA). The vitamin mix, mineral mix (Vitapak, Land O' Lakes Feed, St. Paul, MN, USA), Menhaden fish meal (Consumer Supply Distribution Company, Sioux City, IA, USA), and whey (Bongards Creameries, Perham, MN, USA) were incorporated into the ingredient blends at proportions of 1, 2, 5, and 5%, respectively, on a wet basis of feed. All these ingredients were blended in a laboratory scale Hobart mixer (model D300 Hobart Corporation, Troy, OH, USA), at the lowest speed (82 rpm inner rotation and 36 rpm outer rotation) for 15 min and conditioned to the target feed moisture content (15, 20, and 25% wb). These ingredient blends were formulated to a total energy content of 14, 464 kJ/kg, and a protein content of 28% wb. Next, the blends were placed in polyethylene bags and stored at room temperature for 24 h to allow moisture uniformity throughout the blends prior to extrusion processing.

Experimental Design and Analysis

Twenty-seven isocaloric ingredient blends (14, 644 kJ/kg) were formulated to a protein content of 28% (wb). Three starch sources (cassava, corn, and potato) were blended with other common feed ingredients namely soy flour, fish meal, whey, vitamin and mineral mix (Table 1). Experimental factors and the levels of each included three starch sources (cassava, corn, and potato), three proportions of DDGS (20, 40, and 60% wb), feed moisture content (15, 20, and 25% wb), and extruder barrel temperature profiles (90–100–100°C, 90–120–120°C, and 90–140–140°C) as summarized in Table 2. These factors and their subsequent levels resulted in a total of 81 treatment combinations (3 × 3 × 3 × 3 full-factorial design) for the study. Formal statistical analyses on all the collected data were performed using SAS software[24] (SAS Institute, Cary, NC, USA) using type I error rate (α) of 0.05, which included Least Significance Difference (LSD), Duncan's Multiple Range Test for mean comparison, and correlation analysis.

Table 1 Ingredient components in different feed blends

	Cassava starch blend			Cornstarch blend			Potato starch blend
Feed Ingredient	Blend 1	Blend 2	Blend 3	Blend 1	Blend 2	Blend 3	Blend 1	Blend 2	Blend 3
DDGS	20.0	40.0	60.0	20.0	40.0	60.0	20.0	40.0	60.0
Starch	35.0	24.9	14.5	31.5	22.5	13.2	31.0	22.0	13.0
Soy flour	31.0	21.1	11.5	34.5	23.5	12.7	35.0	24.0	13.0
Fish meal	6.0	6.0	6.0	6.0	6.0	6.0	6.0	6.0	6.0
Whey protein	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0	5.0
Vitamin mix	1.0	1.0	1.0	1.0	1.0	1.0	1.0	1.0	1.0
Mineral mix	2.0	2.0	2.0	2.0	2.0	2.0	2.0	2.0	2.0
Total	100.0	100.0	100.0	100.0	100.0	100.0	100.0	100.0	100.0

Table 2 Experimental design used in the study. ^a

Starch type	Cassava			Corn			Potato
DDGS levels (% wb)	20	40	60	20	40	60	20	40	60
Feed moisture content (% wb)	15	20	25	15	20	25	15	20	25
Extruder barrel temperature (°C)	100	120	140	100	120	140	100	120	140
^a{3 (starch sources) × 3 (DDGS levels) × 3 (moisture contents) × 3 (temperatures)} = 81 total treatment combinations.

Extrusion Processing

After mixing and preconditioning, the DDGS based blends were randomly extruded using a laboratory scale single-screw extruder (Brabender Plasti-corder extruder PL 2000, South Hackensack, NJ, USA), which is shown in Fig. 1. The single-screw extruder powered by: a 7.5-hp motor had a 3:1 compression ratio; a screw length to diameter (L/D) ratio of 20:1; and a 317.5-mm barrel length. The die assembly had an internal conical section and a 101.6-mm length. We used a screw with a 9.05-mm uniform pitch, which had variable flute depth, with the depth of 9.05 feed depth and a 3.81-mm die portion. The speed of the screw, which had a range of 0 to 210 rpm (0–22 rad/sec), and the temperature inside the extruder barrel were run by a computer-control system. The extruder was operated at a preset screw speed of 130 rpm (13.6 rad/s), and the temperature profiles were maintained at 90–100–100°C, 90–120–120°C, and 90–140–140°C in the feed, transition, and metering zones to produce expanded product. The response variables measured included extrudate physical properties like expansion ratio, unit density, color (L*, a*, and b*), sinking velocity, water absorption, solubility, and pellet durability indices and extruder processing parameters (extruder motor torque and die pressure).

Figure 1 Schematic diagram of the single-screw extruder used in the study.

Measurement of Extrudates Properties and Processing Parameters

At least 2 kg of extrudates were extruded for each treatment combination (81 total treatment combinations), which resulted in approximately 162 kg of DDGS based extrudates. The physical properties on the collected extrudates were determined by the method as outlined by Chevanan etal.[20]

Expansion Ratio

Radial expansion ratio was measured as the ratio of diameter of the extrudates to the diameter of the die. The diameter of the extrudates was measured using a digital caliper (Digimatic Series No. 293, Mitutoyo Co., Tokyo, Japan).

Unit Density

The sample extrudates were cut into pieces of lengths ≈25.4 mm, each piece weighed for its mass using a weighing scale (Mettler Instrument Corporation, Hightstown, NJ, USA), height and diameter using Vernier caliper (Mitutoyo Corporation, Japan). The unit or apparent density was then determined as the ratio of calculated mass to the calculated volume of the extrudates (considering the extrudates to be a right circular cylinder).

Color

The color of ground extrudates (≈150 μm) was measured using a Minolta Chromameter (model CM 2500d, Minolta, Japan). A Hunter Lab color space was used to measure the brightness/luminosity (L*, where L = 0 for black; L = 100 for white), redness/greenness (+a for red and –a for green), and yellowness/blueness (+b for yellow and –b for blue). We completed three measurements for each sample.

Sinking Velocity

The sample extrudates were cut into pieces of length ≈25.4 mm and then dropped into a 2 L measuring cylinder, which was filled with distilled water at 25 ± 1°C. The time taken for the extrudate to reach the bottom of the cylinder (total distance = 0.415 m) was recorded using a timer. Sinking velocity was then determined as the ratio of the distance travelled to the time taken by the extrudates to hit the bottom of the cylinder.

Water Absorption and Solubility Indices

Extrudates were ground to fine powders of ≈150 μm using a coffee grinder (Black & Decker® Corporation, Towson, ML, USA). The ground extrudates (2.5 g) was suspended in distilled water (30 mL) in a tared 60 mL centrifuge tube. The suspension was placed in a conventional oven at 30°C for 30 min, which was stirred intermittently and centrifuged at 3000g for 10 min. The supernatant was carefully poured into a tared aluminium cup and dried at 135°C for 2 h.[25] The weight of the gel remaining in the centrifuge tube was measured. Three replicate measurements were recorded. The WAI and WSI were calculated by:

(1)

where WAI: water absorption index (−); W_g : the weight of gel (g); and W_ds : the weight of dry sample (g).

(2)

where WSI: the water solubility index (%); W_ss : the weight of dry solids of supernatant (g); and W_ds : the weight of dry sample (g).

Pellet Durability Index

About 200 g of extrudates were broken into pieces of lengths ≈25.4 mm, which were then divided into two batches of 100 g each. Each batch was placed in a pellet durability tester (model PDT-110, Seedburo equipment company, Chicago, IL, USA) for tumbling over a period of 10 min. The extrudates were sieved through sieve no. 6 (3.36 mm) before and after tumbling, and then they were measured for its underflow and overflow. The pellet durability Index was then calculated using equation (3)

(3)

where PDI: the pellet durability Index (%); M_at : the mass of the pellets after tumbling (g); and M_bt : the mass of the pellets before tumbling (g).

Extruder Motor Torque and Die Pressure

Extruder motor torque required to rotate the screw and the die pressure were recorded every 20 sec using a torque and pressure transducer, respectively, after the extruder was operating under steady-state conditions. The torque and pressure transducers had a range of 0 to 40, 000 mG (0 to 400 N m) and 0 to 10,000 psi (0 to 68.9 MPa), respectively. The torque and die pressure values at the end of the barrel for each treatment were averaged from 10 values.

RESULTS AND DISCUSSION

The effect of changing the levels of DDGS, feed moisture content, and extruder barrel temperatures on the physical properties of the resulting extrudates is summarized in Table 3. For cassava starch extrudates, changing the levels of DDGS significantly affected most of the extrudate properties studied except SV and WSI. In contrast, feed moisture content and extruder barrel temperature affected all the extrudate properties except for WSI with changes in feed moisture content. For corn- and potato-starch extrudates, changing the levels of DDGS had a significant effect on all the extrudate properties with an exception for UD. Changes in feed moisture content did not affect L* values significantly for corn and potato starch extrudates. Furthermore, for cornstarch extrudates, increasing the extruder barrel temperature from 100 to 140°C had a significant impact on all the extrudate properties studied except for L* values (Table 3).

Table 3 Main effect of physical properties of cassava, corn, and potato starch extrudates. ^†

		Physical properties of DDGS extrudates
Starch source	Parameter levels	ER (−)	UD (kg/m^3)	L* (−)	a* (−)	b* (−)	SV (m/s)	WAI (−)	WSI (%)	PDI (%)
Cassava	DDGS (% wb)
	20	1.360a	874.9b	63.2a	4.91b	30.5c	0.069a	3.33a	16.3a	86.6a
		(0.04)	(26.80)	(0.49)	(0.17)	(0.34)	(0.01)	(0.14)	(0.26)	(1.24)
	40	1.215b	897.4ab	58.8b	7.86a	31.5b	0.073a	3.04b	16.3a	71.0b
		(0.02)	(21.40)	(0.56)	(0.13)	(0.25)	(0.01)	(0.11)	(0.27)	(5.76)
	60	1.126c	924.4a	52.9c	8.04a	38.4a	0.071a	2.90c	17.1a	70.3b
		(0.01)	(23.70)	(0.33)	(0.19)	(0.60)	(0.01)	(0.05)	(0.77)	(4.76)
	MC (% wb)
	15	1.103c	989.8a	59.2a	6.319c	33.0b	0.094a	2.70c	16.8a	76.6b
		(0.01)	(21.60)	(0.68)	(0.26)	(0.90)	(0.00)	(0.03)	(0.76)	(3.19)
	20	1.244b	898.2b	58.0b	6.976b	33.7a	0.080b	3.05b	16.7a	80.6a
		(0.03)	(20.10)	(0.96)	(0.34)	(0.80)	(0.00)	(0.08)	(0.24)	(2.10)
	25	1.354a	808.5c	57.6b	7.528a	33.8a	0.039c	3.52a	16.2a	70.6c
		(0.03)	(16.80)	(1.14)	(0.33)	(0.69)	(0.01)	(0.14)	(0.32)	(7.05)
	T (°C)
	100	1.171b	964.1a	59.6a	6.78b	32.5b	0.088a	2.70c	17.7a	67.8c
		(0.02)	(21.30)	(1.05)	(0.36)	(0.59)	(0.00)	(0.05)	(0.75)	(6.97)
	120	1.252a	870.3b	58.4b	6.83b	33.8a	0.071b	3.10b	16.6ab	77.1b
		(0.03)	(15.60)	(0.91)	(0.26)	(1.00)	(0.01)	(0.07)	(0.13)	(2.21)
	140	1.278a	862.2b	56.8c	7.21a	34.2a	0.056c	3.47a	15.3b	83.1a
		(0.04)	(28.90)	(0.81)	(0.33)	(0.73)	(0.01)	(0.14)	(0.23)	(2.63)
Corn	DDGS (% wb)
	20	1.155a	947.4a	61.0a	5.72c	30.9c	0.084a	3.11a	14.9c	84.5a
		(0.02)	(24.00)	(2.15)	(0.19)	(0.32)	(0.00)	(0.09)	(0.23)	(1.58)
	40	1.103b	973.9a	57.4a	7.32b	34.4b	0.082a	2.88c	16.4b	83.3b
		(0.01)	(14.90)	(0.58)	(0.19)	(0.35)	(0.00)	(0.06)	(0.15)	(1.28)
	60	1.057c	969.4a	52.6b	8.56a	40.7a	0.079b	2.99b	16.8a	65.5c
		(0.01)	(18.90)	(0.53)	(0.15)	(0.83)	(0.00)	(0.03)	(0.14)	(3.39)
	MC (% wb)
	15	1.059c	1034.3a	56.7a	7.24b	34.9b	0.093a	2.88b	16.0b	79.9a
		(0.01)	(10.00)	(2.13)	(0.28)	(1.00)	(0.00)	(0.05)	(0.17)	(2.30)
	20	1.104b	976.6b	57.7a	6.59c	35.9a	0.086a	2.91b	16.2a	77.6ab
		(0.01)	(16.20)	(0.95)	(0.25)	(0.92)	(0.00)	(0.06)	(0.23)	(4.27)
	25	1.150a	879.8c	56.5a	7.77a	35.1b	0.067a	3.18a	15.8c	76.6b
		(0.02)	(18.50)	(1.07)	(0.29)	(0.98)	(0.00)	(0.08)	(0.29)	(3.02)
	Temperature (°C)
	100	1.112a	1006.7a	58.9a	7.41a	34.7b	0.091a	2.79c	16.8a	78.0b
		(0.01)	(7.61)	(1.03)	0.24()	(0.98)	(0.00)	(0.05)	(0.15)	(3.75)
	120	1.122a	935.2b	56.7a	7.22b	34.8b	0.076c	2.99b	15.9b	73.5c
		(0.02)	(15.70)	(2.12)	(0.33)	(0.93)	(0.00)	(0.05)	(0.13)	(3.53)
	140	1.080b	948.7b	55.4a	6.97c	36.5a	0.080b	3.20a	15.3c	82.6a
		(0.02)	(27.50)	(0.91)	(0.29)	(0.97)	(0.00)	(0.08)	(0.29)	(2.01)
	DDGS (% wb)
	20	1.310a	936.7a	63.0a	4.95c	31.3c	0.066b	3.56a	16.5b	86.5a
		(0.03)	(31.50)	(0.50)	(0.14)	(0.22)	(0.01)	(0.79)	(0.27)	(2.03)
	40	1.185b	943.1a	57.1b	6.72b	35.6b	0.086a	3.33b	15.6c	85.0b
		(0.02)	(19.60)	(0.53)	(0.16)	(0.50)	(0.00)	(0.09)	(0.24)	(1.35)
	60	1.104c	946.4a	51.4c	8.26a	40.4a	0.081a	2.84c	17.3a	74.4c
		(0.01)	(19.90)	(0.50)	(0.16)	(0.56)	(0.00)	(0.03)	(0.22)	(3.41)
Potato	MC (% wb)
	15	1.098c	1053.5a	57.3a	6.91a	34.4b	0.098a	2.99b	16.2b	81.9a
		(0.01)	(19.40)	(0.78)	(0.32)	(0.76)	(0.00)	(0.06)	(0.23)	(3.51)
	20	1.209b	929.2b	57.2a	6.25b	36.5a	0.080b	3.36a	16.1b	82.2a
		(0.03)	(15.60)	(1.13)	(0.27)	(0.84)	(0.00)	(0.09)	(0.23)	(2.23)
	25	1.291a	843.5c	57.0a	6.76a	36.4a	0.054c	3.38a	17.0a	81.8a
		(0.03)	(17.30)	(1.20)	(0.31)	(0.91)	(0.01)	(0.10)	(0.33)	(2.34)
	T (°C)
	100	1.154c	981.6a	57.5a	6.44b	36.3a	0.091a	3.03c	16.9a	80.1b
		(0.01)	(20.30)	(1.18)	(0.34)	(0.82)	(0.00)	(0.06)	(0.34)	(3.60)
	120	1.203b	920.2b	57.9a	6.71a	34.6b	0.069b	3.21b	16.8a	77.0c
		(0.02)	(16.60)	(0.93)	(0.31)	(0.82)	(0.01)	(0.06)	(0.18)	(2.21)
	140	1.240a	924.5b	56.1b	6.78a	36.4a	0.072b	3.50a	15.7b	88.87a
		(0.04)	(31.60)	(1.00)	(0.26)	(0.90)	(0.01)	(0.12)	(0.23)	(0.69)
^†Means followed by similar letters within each parameter are not significantly different for a given property at α = 0.05. Values in the parentheses are standard error. Also, ER: expansion ratio; UD: unit density, L*: brightness/luminosity; a*: redness/greenness; b*: blueness/yellowness; SV: sinking velocity; WAI: water absorption index; WSI: water solubility index; and PDI: pellet durability index.

Expansion ratio

The degree of expansion is an important factor to be monitored that affects the density, fragility, and softness of the extruded products.[26] Also, the degree of expansion of the extrudate is closely related to the size, number, and distribution of air cells surrounded by the cooked material.[27] For all the three starch bases (cassava, corn, and potato), increasing the DDGS content within the blend from 20 to 60% wb, resulted in a pronounced decrease in ER values by 17.2, 8.48, and 15.7%, respectively (Table 3). This decrease in ER values could account for the decrease in starch content, which is primarily responsible for the expansion of the extruded product, in the respective ingredient blend with the increase in DDGS level (Table 1). Present results are in agreement with findings postulated by Chevanan etal.[19,21] In addition, for all the three starch bases, increasing the feed moisture content from 15 to 25% wb and processing temperature from 100 to 140°C (cassava, corn, and potato), resulted in a proportional increase in ER values of the extrudates (Table 3).

Investigations performed by Gomez and Aguilera[28] and Paton and Spratt[29] revealed that temperature had a direct relation to the expansion ratio, which are in acceptance with the results of these experimental studies. At the higher temperature range, it is likely that more chemical bonds between starch molecules were being broken, and that the dispersion of starch molecules was greater hence less resistance to expansion, better puffing, which resulted in extrudates with higher ER values.[30] Research conducted by Chinnasamy and Hanna,[5] Faubion and Hoseney,[31] and Antila etal.[32] postulated an inverse relation between ER and feed-moisture content. The findings in this study seem to contradict their previous research done. This is a somewhat unexpected finding and a finer analysis needs to be made. In addition, the contradiction in the results might be due to the difference in the ingredient composition and extruder processing parameters used in the studies. On the other hand, results of Sun and Muthukumarappan[33] showed the existence of direct relation between feed moisture content and expansion ratio, which was seconded by the results of this current study for all the three starch extrudates.

The experimental ER data ranged from 0.947 to 1.667 with 1.67 and 0.95 as the highest and the lowest, observed with cassava and cornstarch extrudates, respectively. The highest (1.667) and the lowest ER (0.947) values were found with cassava and cornstarch extrudates, respectively (Tables 4 and 6). In addition, it has been proven that particle size of the ingredient blend, amylose content, degree of gelatinization, and fat level also have effect on the expansion ratio values.[4,34] Research conducted by Bhatnagar and Hanna[35] proved decreased ER values with a corresponding increase in amylose content. In agreement to the statement,[35] in this present study, cassava starch and cornstarch possessed the minimum and maximum amylose content, which resulted in the highest and the lowest ER values.

Table 4 Treatment combination effect of cassava starch extrudates. ^†

		Feed Moisture Content (% wb)
		15			20			25
		Extruder barrel temperature (°C)
Property	DDGS (% wb)	100	120	140	100	120	140	100	120	140
ER (−)	20	1.12d-g	1.19d-g	1.05h-i	1.19d-g	1.10e-h	1.05h-i	1.08g-i	1.10f-i	1.00i
	40	1.15d-h	1.10f-i	1.14d-g	1.46b-c	1.21d-e	1.13d-g	1.62a	1.22d	1.14d-g
	60	1.42b-c	1.15d-h	1.18d-g	1.50b	1.38c	1.20d-f	1.66a	1.45b-c	1.21d-e
UD (kg/m^3)	20	1073.8a-b	939.8d-f	1108.4a	912.7e-h	935.7d-f	988.1b-e	692.9l	812.1h-k	1077.5a-b
	40	963.6b-f	1049.1a-d	882.4e-j	788.7i-l	903.6e-i	889.5e-j	1060.4a-c	956.5c-f	924.3e-h
	60	861.4f-j	972.2b-f	826.1g-k	793.7i-l	785.1j-l	835.72g-k	726.6k-l	722.2k-l	787.2i-l
L* (−)	20	63.2b-e	60.6f-g	54.3h-j	61.4d-f	62.0c-f	54.3h-j	61.2e-g	60.8e-g	55.3h-i
	40	65.2b	58.9g	53.5i-k	63.7b-d	60.0f-g	52.7j-l	61.8c-f	54.9h-j	51.4k-l
	60	68.3a	60.9e-g	51.8k-l	63.8b-c	56.1h	51.7k-l	60.6f-g	54.7h-j	50.9l
a* (−)	20	4.31k	7.56d-g	7.13f-h	4.26k	7.17f-h	7.56d-g	4.93i-j	6.98g-h	6.98g-h
	40	4.21k	7.69d-f	9.09b	4.77i-k	7.57d-g	7.24e-h	5.31i	9.02b	7.88d-e
	60	4.37j-k	7.88d-e	8.78b-c	6.85h	8.18c-d	7.87d-e	5.25i	8.75b-c	9.82a
b* (−)	20	29.5k	31.1h-i	39.1c-d	28.9k	31.0h-j	40.6b	29.6j-k	29.0k	38.0d
	40	29.0k	31.0h-i	34.5e-f	30.4i-k	31.6g-i	40.1b-c	32.8g	32.3g-h	41.3a-b
	60	31.9g-i	31.9g-i	34.4e-f	29.5j-k	32.7g	42.6a	33.2f-g	33.1f-g	34.9e
SV (m/s)	20	0.117a	0.09a-c	0.09a-d	0.09a-c	0.09a-c	0.08c-e	0.10a-b	0.09a-c	0.10a-c
	40	0.09a-c	0.107a-c	0.09a-e	0.08c-e	0.08c-f	0.08b-e	0.06d-g	0.06e-g	0.08c-f
	60	0.06f-g	0.09a-c	0.08c-e	0.02h	0.05g	0.06f-g	0.00i	0.000i	0.00i
WAI (−)	20	2.57o-p	2.45q	2.65m-o	2.66l-o	2.80j-k	2.76k-l	3.03i	2.72k-n	2.64n-p
	40	2.55p	2.75k-m	2.79j-k	3.22g	2.82j-k	3.13g-h	3.98c	3.37f	2.87j
	60	3.36f	2.55p	2.63n-p	3.73d	3.68d	3.08h-i	4.90a	4.18b	3.53e
WSI (%)	20	16.9a	16.5a	15.7a	17.6a	16.7a	15.2a	17.1a	17.6a	13.3a
	40	17.6a	17.2a	15.8a	18.0a	16.8a	14.4a	17.5a	15.6a	14.0a
	60	18.5a	16.1a	16.6a	17.8a	17.0a	16.9a	18.6a	16.3a	15.9a
PDI (%)	20	95.0a	88.6b-c	66.0i-j	75.3g	67.2i	66.0i-j	87.7b-d	54.3k	77.1f-g
	40	89.6b-c	78.2f-g	71.9h	87.8b-d	84.4d-e	63.3j	86.0c-d	89.3b-c	87.7b-d
	60	87.0b-d	11.2m	22.4l	82.4e	78.9f	88.2b-c	88.3b-c	87.0b-d	90.2b
^†Means followed with the similar letters within each parameter are not significantly different at α = 0.05. Also, ER: expansion ratio; UD: unit density, L*: brightness/luminosity; a*: redness/greenness; b*: blueness/yellowness; SV: sinking velocity; WAI: water absorption index; WSI: water solubility index; NdPDI: pellet durability index.

Table 5 Treatment combination effect of cornstarch extrudates. ^†

		Feed moisture content (% wb)
		15			20			25
		Extruder barrel temperature (°C)
Property	DDGS (% wb)	100	120	140	100	120	140	100	120	140
ER (−)	20	1.14c-f	1.15c-d	1.08f-i	1.08g-i	1.07h-i	1.05i	0.94k	1.00j	1.00j
	40	1.13c-g	1.11d-h	1.10d-i	1.13c-f	1.10d-i	1.09e-i	1.20b	1.08g-i	0.96j
	60	1.11d-h	1.09e-i	1.07h-i	1.32a	1.17b-c	1.06h-i	1.31a	1.14c-e	1.06h-i
UD(kg/m^3)	20	1058.9a-d	1010.1a-f	1033.3a-e	1026.7a-e	1003.4a-g	978.9c-g	1060.9a-c	1088.8a-b	1047.8a-e
	40	1038.5a-e	1012.3a-f	948.9e-i	958.9d-g	981.0c-g	905.9g-i	875.8h-k	966.1c-h	1101.7a
	60	983.0c-g	990.9b-g	984.5c-g	780.3l-m	859.4i-l	922.6f-j	743.6m	852.8j-l	800.7k-m
L* (−)	20	64.9a	58.3a-c	52.2a-c	48.8b-c	59.0a-c	53.3a-c	60.9a-c	57.4a-c	55.5a-c
	40	64.6a	60.2a-c	53.0a-c	62.8a-b	60.9a-c	52.8a-c	60.1a-c	53.7a-c	51.4a-c
	60	65.7a	58.8a-c	52.4a-c	61.1a-b	55.6a-c	55.7a-c	60.4a-c	52.2a-c	46.7c
a* (−)	20	6.36g-h	8.03d-f	8.11c-f	5.69i	8.17c-e	8.23c-e	4.74j	6.49g-h	9.33a
	40	6.53g-h	5.69i	7.75e-f	4.44j	7.83d-f	7.96d-f	5.05j	6.18h-i	7.88d-f
	60	6.24g-i	8.21c-e	9.82a	4.96j	8.46c-d	9.25a-b	7.48f	6.84g	8.71b-c
b* (−)	20	29.8j-k	33.1g-i	43.9b	28.9k	32.9g-i	43.0b-c	31.8h-j	36.7d	34.6e-g
	40	29.9j-k	35.2d-f	41.9c	31.3i-j	32.5h-i	42.9b-c	32.7g-i	35.5d-e	41.2c
	60	29.2k	33.5f-h	35.8d-e	33.2g-i	33.0g-i	35.9d-e	31.3i-j	36.9d	47.5a
SV (m/s)	20	0.109a	0.09c-f	0.09d-g	0.10a-b	0.09c-f	0.09c-f	0.10a	0.10a	0.09b-e
	40	0.09a-c	0.09a-c	0.08g-h	0.09c-f	0.08h-i	0.08h-i	0.09a-d	0.08e-h	0.07i
	60	0.09a-d	0.09a-d	0.08f-h	0.04m	0.05l	0.07j	0.05k	0.07j	0.05k
WAI (−)	20	2.63k-l	3.08e-g	2.91f-j	2.80h-l	2.83g-l	3.01e-h	3.23c-e	2.65k-l	2.82g-l
	40	2.63k-l	2.61l	2.89f-k	2.67j-l	2.71i-l	3.12d-f	3.59b	3.00e-h	3.04e-h
	60	2.93f-i	2.66j-l	2.78h-l	3.47b-c	2.94f-i	3.34c-d	4.09a	3.42b-c	2.97f-i
WSI (%)	20	16.1h-i	17.0d-e	17.4b	15.5j-k	15.8i-j	15.6j-k	14.7l	15.0l	16.8e-f
	40	16.1h-i	17.0c-e	17.4b-c	15.7j-k	17.3b-d	16.3g-h	13.2m	16.1h-i	16.8e-f
	60	15.3k	17.1b-e	17.8a	14.8l	16.6f-g	15.7i-j	12.6n	15.6j-k	17.0c-e
PDI (%)	20	96.4a	91.6a-c	74.8g-h	82.3e-f	78.2f-g	68.0i	81.7e-f	80.2e-g	66.2i
	40	93.3a-b	80.4e-g	56.4j	84.9d-e	82.7e-f	37.2l	91.5a-c	85.8c-e	86.2c-e
	60	84.9d-e	75.2g-h	49.0k	74.7g-h	85.7c-e	68.1i	89.9b-d	90.1b-d	71.6h-i
^†Means followed with the similar letters within each parameter are not significantly different at α = 0.05. ER: expansion ratio; UD: unit density, L*: brightness/luminosity; a*: redness/greenness; b*: blueness/yellowness; SV: sinking velocity; WAI: water absorption index; WSI: water solubility index; and PDI: pellet durability index.

Table 6 Treatment combination effect of potato starch extrudates. ^†

		Feed moisture content (% wb)
		15			20			25
		Extruder barrel temperature (°C)
Property	DDGS (% wb)	100	120	140	100	120	140	100	120	140
ER (−)	20	1.08h-j	1.13f-i	1.10g-j	1.14e-i	1.10g-j	1.07i-k	1.19d-f	1.03j-k	1.01k
	40	1.20d-e	1.13f-i	1.15e-h	1.30c	1.23d	1.07h-j	1.55a	1.14e-i	1.08h-j
	60	1.32c	1.12f-i	1.12f-i	1.41b	1.33c	1.14e-i	1.56a	1.41b	1.16d-g
UD (kg/m^3)	20	1195.7a	1037.7c-d	951.0d-g	1029.1c-d	980.7d-g	950.7d-g	1097.3b-c	1100.7b-c	1138.9a-b
	40	975.6d-g	977.3d-g	879.4g-j	877.1g-j	918.4e-h	1012.9c-e	785.0j-l	942.3d-g	994.9d-f
	60	965.3d-g	937.7d-g	914.9e-h	801.6i-l	819.0h-k	892.1f-i	703.9l	774.5k-l	782.6j-l
L* (−)	20	62.2b	56.7d-e	54.4e-g	62.2b	59.4c	51.9h-j	59.0c	58.7c-d	51.1i-j
	40	68.5a	56.1e-f	52.3g-i	61.9b	53.7f-h	52.4g-i	62.3b	56.3e	48.8j-k
	60	64.3b	55.0e-f	48.3k	63.6b	62.3b	53.7f-h	63.1b	55.5e-f	48.7k
a* (−)	20	4.72h-i	6.55c-d	9.71a	4.41i-j	7.76b	7.60b	5.54f-g	7.91b	8.04b
	40	3.90j	6.15d-f	7.69b	4.87h-i	6.31c-e	7.57b	5.25g-h	6.40c-d	8.07b
	60	4.68h-i	6.86c	7.74b	6.45c-d	5.70e-g	9.70a	4.72h-i	6.83c	8.21b
b* (−)	20	31.7f-g	36.9d-e	35.7e	29.6i	31.5f-g	40.4c	32.2f-g	31.2f-i	40.4c
	40	30.7g-i	36.4d-e	41.6b-c	31.3f-h	36.4d-e	41.2b-c	32.6f	36.0e	42.3a-b
	60	31.9f-g	38.0d	43.4a	29.7h-i	35.9e	35.4e	31.9f-g	37.9d	43.4a
SV (m/s)	20	0.10a-c	0.10a-d	0.09b-e	0.09c-f	0.10a-d	0.09a-d	0.11a-b	0.11a	0.09a-d
	40	0.10a-d	0.10a-c	0.09c-g	0.07g-i	0.07h-j	0.08d-h	0.05j	0.10a-d	0.07e-i
	60	0.07f-i	0.09c-g	0.09c-g	0.00k	0.06i-j	0.06h-j	0.00k	0.06i-j	0.06i-j
WAI (−)	20	3.13e-f	2.91f-i	2.75i-j	3.36d	3.11e-f	2.58j	3.62c	2.70i-j	2.81h-i
	40	3.14e	3.00e-h	2.88g-i	3.50c-d	3.60c	3.13e-f	4.24a	3.90b	2.81h-i
	60	3.50c-d	3.43c-d	2.71i-j	3.42c-d	3.08e-g	2.89g-i	4.15a	4.21a	3.01e-h
WSI (%)	20	14.5k-l	16.1f-i	18.4a-b	16.5d-g	16.1f-i	17.2c-f	15.0i-l	16.9c-f	15.1h-l
	40	17.4b-e	15.7g-j	17.6b-d	17.0c-f	14.8j-l	17.1c-f	15.3g-k	14.1l	16.3e-h
	60	17.8b-c	14.8j-l	19.3a	17.9b-c	17.6b-d	17.1c-f	17.1c-f	14.2k-l	17.3b-e
PDI (%)	20	95.5a	93.3a-b	54.7m	89.6c-e	76.6h	59.1l	94.0a-b	89.3c-e	84.7g
	40	92.7a-c	86.5e-g	64.6k	70.9j	84.9f-g	75.0h-i	86.8e-g	87.1e-g	91.4b-d
	60	88.2d-g	85.1f-g	60.5l	72.8i-j	75.4h-i	88.4d-f	88.4d-f	86.8e-g	90.8b-d
^†Means followed with the similar letters within each parameter are not significantly different at α = 0.05. Also, ER: expansion ratio; UD: unit density; L*: brightness/luminosity; a*: redness/greenness; b*: blueness/yellowness; SV: sinking velocity; WAI: water absorption index; WSI: water solubility index; and PDI: pellet durability index.

Color

Color is one of the physical properties used by feed customers to access the quality of pellets.[36] Additionally, processing conditions during extrusion favor non-enzymatic browning by Maillard reaction between proteins and reducing sugars.[37] For all the three starch bases studied, increasing the DDGS levels from 20 to 60% wb resulted in a proportional decrease in L* values (Table 3). In contrast, a* and b* values were found to increase significantly with the increase in DDGS content. This is quite logical because DDGS used in the study was yellowish brown in color and hence the proportional increase in DDGS content in the blend, resulted in a marginal decrease in L* values and marked increase in a* and b* values. Similar results were reported by Chevanan etal.,[19,21] For cassava starch extrudates, an increase in feed moisture content from 15 to 25% wb, resulted in decreased L* values by 18.3% (Table 3).

Moisture content did not have a significant effect on L* values of corn and potato starch extrudates. No clear patterns emerged on color values (L*, a*, and b*) due to changes in temperature profile of the extruder for all the three starch extrudates. The highest L* value (68.5) was found with potato starch extrudates, which signifies that the potato starch used in the study was brighter in comparison with other starch sources used. Cornstarch used in the current investigation was yellowish in color, which possessed the highest b* value (47.5).

Unit Density

Unit density is an important quality parameter, which determines whether the extrudates possess either floating or sinking nature.[20] For cassava starch extrudates, increase in DDGS levels from 20 to 60% wb considerably increased the unit density values by 5.66%, where as any changes in the DDGS levels did not have significant impact on unit density values of corn and potato starch extrudates (Table 3). Similar trend was reported by Chevanan etal.[20] with an increase in DDGS levels in the ingredient blends. Furthermore, for all the three starch bases (cassava, corn, and potato), increasing the feed moisture content from 15 to 25% wb resulted in a pronounced decrease in unit density values of the extrudates by 18.3, 14.9, and 19.9%, respectively (Table 3). Generally, expansion of the extrudates decreases substantially when the moisture content of the feed ingredient blend increases.[31,32,38] In contrast, results in this study showed an increased expansion for corresponding increase in feed moisture content, which resulted in decreased unit density values. This inconsistency in the experimental results could be attributed to the difference in the feed and extruder processing conditions used in the studies.

Unit density values markedly decreased with the increase in extruder barrel temperature from 100 to 140°C. Results proved that temperature has an inverse relationship with the apparent viscosity of the ingredient melt inside the barrel and die,[39] and ultimately higher temperatures resulted in lower apparent viscosity of the melt. Hence, as the ingredient melt with low viscosities exits through the die, the extrudates are subjected to higher expansion, and thus reduced unit density values. Similar reports were pointed out by Case etal.,[40] and Chevanan etal.[20]

The lowest and the highest unit density values were found to be 692.9 kg/m³ and 1195.7 kg/m³ for cassava and potato starch extrudates, respectively (Tables 4 and 6). The highest expansion ratio and the lowest unit density values were observed with cassava starch extrudates, which seems to be quite logical due to the presence of higher amount of amylopectin in comparison with the other starch sources used in the study.

Sinking Velocity

Extrudates sinking velocity is an important property that determines the stability of the extrudates in water and is more closely related to the absorption of water during the feed on the water surface.[19] Increasing the DDGS levels from 20 to 60% wb resulted in decreased and increased sinking velocity values for corn and potato starch extrudates by 5.95 and 22.7%, respectively (Table 3). However, no significant difference was found to exist on cassava starch extrudates for the change in DDGS levels. Furthermore, for cassava and potato starch extrudates, increasing the feed moisture content from 15 to 25% wb resulted in a sharp reduction in sinking velocity values by 58.5 and 44.9%, respectively (Table 3). Similar results were reported by Chevanan etal.[20] One possible way of explanation could be attributed to the increase in extrudates' expansion with corresponding increase in feed moisture content, which resulted in decreased unit density of the extrudates, which in turn led to decreased sinking velocity values. For all the three-starch extrudates (cassava, corn, and potato) sinking velocity value decreased by 36.4, 12.1, and 20.9%, respectively, with an increase in extruder barrel temperature from 100 to 140°C (Table 3). Results of this study strongly support and challenges the findings postulated by Chevanan etal.,[20] At higher processing temperatures, the products expanded well; this resulted in a less dense product, which in turn had poor sinking velocity values. The highest sinking velocity (0.11 m/s) was observed for potato starch extrudates, which had the highest unit density values too (Table 6). Furthermore, for corn and potato starch extrudates, the statistical analyses on all collected data showed that a significant difference did exist among all possible treatment combinations (Table 7).

Table 7 Interaction effects of DDGS, moisture content, and temperature on the physical properties of cassava, corn, and potato starch extrudates. ^†

		Physical properties of DDGS extrudates
Starch source	Interactions	ER (−)	UD (kg/m^3)	L* (−)	a* (−)	b* (−)	SV (m/s)	WAI (−)	WSI (%)	PDI (%)
	DDGS	*	*	*	*	*	ns	*	ns	*
	MC	*	*	*	*	*	*	*	ns	*
	Temp	ns	*	*	*	*	*	*	*	*
Cassava	DDGS*MC	ns	*	*	*	*	*	*	ns	*
	DDGS*temp	ns	*	*	*	*	ns	*	ns	*
	MC*temp	ns	ns	*	*	*	*	*	ns	*
	DDGS*MC*temp	ns	*	*	*	*	ns	*	ns	*
	DDGS	*	ns	*	*	*	*	*	*	*
	MC	*	*	ns	*	*	*	*	*	*
	Temp	*	*	ns	*	*	*	*	*	*
Corn	DDGS*MC	*	ns	ns	*	*	*	*	*	*
	DDGS*temp	*	*	ns	*	*	*	*	*	*
	MC*temp	*	*	ns	*	*	*	*	*	*
	DDGS*MC*temp	*	*	ns	*	*	*	*	*	*
	DDGS	*	ns	*	*	*	*	*	*	*
	MC	*	*	ns	*	*	*	*	*	ns
	Temp	*	*	*	*	*	*	*	*	*
Potato	DDGS*MC	*	*	*	*	*	*	*	*	*
	DDGS*temp	*	*	*	*	*	*	*	*	*
	MC*temp	*	*	*	*	*	*	*	ns	*
	DDGS*MC*temp	*	ns	*	*	*	*	*	*	*
*Significant at α = 0.05; ns not significant at α = 0.05. ^†-ER: expansion ratio; UD: unit density, L*: brightness/luminosity; a*: redness/greenness; b*: blueness/yellowness; SV: sinking velocity; WAI: water absorption index; WSI: water solubility index; and PDI: pellet durability index.

Table 8 Main effect of DDGS, feed moisture content, and barrel temperature on the extruder processing parameters of cassava, corn, and potato starch extrudates ^†

		Cassava starch extrudates		Cornstarch extrudates		Potato starch extrudates
Parameters	Levels	Torque (N-m)	Die pressure (MPa)	Torque (N-m)	Die pressure (MPa)	Torque (N-m)	Die pressure (MPa)
DDGS (% wb)	20	41.2b	9.88a	50.5a	9.20a	44.2c	9.01a
		(2.14)	(0.60)	(2.70)	(0.55)	(2.00)	(0.53)
	40	30.7c	6.77b	46.3b	7.08b	53.3b	6.62b
		(2.17)	(0.46)	(3.21)	(0.42)	(2.98)	(0.31)
	60	50.0a	4.70c	33.6c	3.76c	61.1a	4.29c
		(4.03)	(0.39)	(2.93)	(0.33)	(4.99)	(0.36)
Feed moisture content (% wb)	15	36.7b	11.6a	34.9b	11.2a	44.8b	10.6a
		(3.08)	(0.47)	(1.92)	(0.35)	(4.47)	(0.32)
	20	55.6a	6.69b	63.3a	6.19b	69.3a	6.54b
		(2.65)	(0.37)	(3.25)	(0.34)	(2.86)	(0.36)
	25	29.5c	3.07c	32.1b	2.63c	44.5b	2.76c
		(2.65)	(0.34)	(2.65)	(0.31)	(2.59)	(0.23)
Extruder barrel temperature (°C)	100	51.2a	9.58a	51.3a	8.79a	70.4a	8.15a
		(3.34)	(0.44)	(3.66)	(0.47)	(4.58)	(0.38)
	120	36.7b	7.73b	48.3a	6.51b	50.3b	7.42b
		(2.92)	(0.61)	(2.95)	(0.44)	(2.78)	(0.48)
	140	33.9b	4.04c	30.7b	4.75c	37.9c	4.35c
		(2.45)	(0.37)	(1.74)	(0.50)	(2.16)	(0.41)
^†Means followed by similar letters within each parameter are non significantly different for a given property at ∞=0.05. Values in parentheses are standard error.

Water Absorption and Solubility Indices

The water absorption and solubility indices characterize the extrudate products and are often crucial in predicting how the extruded materials may behave if further processed. In general, any increase in WSI corresponds to decrease in WAI of the extrudates, which has been proved by several researchers.[41,42] Changing the levels of DDGS, feed moisture content, and extruder barrel temperature had a significant effect on the WAI of the resulting extrudates for all the three starch extrudates (Table 7). In this present experiment, for all the three starch bases (cassava, corn, and potato), increasing the DDGS levels from 20 to 60% wb resulted in a significant decrease in WAI values by 12.9, 3.85, and 20.2%, respectively (Table 3). Similar reports were reported by Chevanan etal.,[21] and Kannadhason etal.,[43] in their study using DDGS as a base material. Increasing the feed moisture content from 15 to 25% wb significantly increased the WAI values for all the three starch extrudates (Table 3). The findings in this study are consistent with previous observations of Gomez and Aguilera,[28] and Chang etal.[44] According to Anderson,[45] higher moisture contents and temperatures lead to greater breakdown of starch and hence an increased formation of expansible matrix, which results in higher water holding capacity, which in turn could resulted to an increased WAI. Additionally, an increase in extruder barrel temperature resulted in significantly increased WAI values for all the three starch extrudates (Table 3). Similar results were pointed out by Sun and Muthukumarappan,[33] in their extrusion studies with defatted soy flour by varying the feed moisture content, barrel temperature. This increase in WAI values with increase in processing temperature could be accounted for the structural modifications involving fiber and these modifications might have promoted interactions between fiber, and the starch in the feed ingredient blends, reducing the solubility.[15] Moreover, Gujral and Singh,[46] proposed increase in WAI values with corresponding increase in starch content in the blend, which was in agreement with the observations of this study. This could be attributed to the decrease in the proportion of starch content in the blend with proportional increase in DDGS levels, which resulted in decreased WAI values.

Changing the levels of DDGS, feed moisture content, and extruder barrel temperature had a significant effect on WSI values of corn and potato starch extrudates. However, changes in DDGS levels and feed moisture content did not have any effect on WSI values of cassava starch extrudates. For corn and potato starch extrudates, increasing the DDGS levels from 20 to 60% wb resulted in a noticeable increase by 12.7 and 4.85%, respectively, whereas no significant effect could be noticed with cassava starch extrudates for any changes in DDGS (Table 3). Similar results were reported by Chevanan etal.[21] For all the three starch extrudates, significant decrease in WSI values was noticed with an increase in extruder barrel temperature from 100 to 140°C (Table 3). Furthermore, for potato starch extrudates, increasing the feed moisture content from 15 to 25% wb resulted in an increased WSI values by 4.94% (Table 3). This increase in WSI is related to the damage caused to the starch structure, which indicates starch decomposition or dextrinization. On the other hand, cornstarch extrudates resulted in a significant decrease by 1.25% for changes in feed moisture content from 15 to 25% wb. This is in agreement with the findings postulated by Chevanan etal.,[20] who reported that feed moisture content and WSI were directly proportional to each other for cornstarch extrudates.

Pellet Durability Index

Pellet durability index is a direct measurement of the pellet's quality to withstand breakage and disintegration of the particles.[47] Moreover, durable pellets alleviate the problem of dust and fines formation.[48] For all the three starch extrudates, increasing the DDGS levels from 20 to 60% wb resulted in a significant decrease in PDI values by 18.8, 22.5, and 14.0%, respectively (Table 3). Similar findings were reported by Chevanan etal.[18,21] This substantial decrease in PDI could be attributed to the decrease in starch content in the ingredient blend as the DDGS levels increased, which resulted in poor starch gelatinization, reduced cohesion, and hence durability.[20] Moreover, ingredients such as fat decreases the pellet quality or durability,[49] and as the DDGS levels increased in the ingredient blends, fat content also increased, which resulted in poor durability of the extrudates. Increasing the feed moisture content from 15 to 25% wb did not have a significant effect on the PDI values of potato starch extrudates. On the other hand, for cassava and cornstarch extrudates, the PDI values decreased by 3.57 and 4.13%, respectively with an increase in feed moisture content (Table 3), which strongly supported and challenged the findings of Chevanan etal.[18,21] Moreover, the mechanical strength of the extrudates depends on the extent of heat treatment and the extent of starch gelatinization that resulted in higher PDI.[4] It was observed that changing the extruder barrel temperature from 100 to 140°C, resulted in a significant increase in PDI values for all the three starch extrudates. Surprisingly, results in this study contradict the findings postulated by Chevanan etal.[20] This is an unexpected finding and a finer analysis is to be made in the consecutive studies.

The lowest (11.2%) and the highest (96.4%) PDI values were observed for corn and cassava starch extrudates (Tables 4 and 5). The lowest PDI value (11.2%) was noticed with the blend that contained 60% DDGS, which is perhaps due to the low starch and high fiber contents. Also, another explanation could be attributed to the presence of highly altered materials that do not interact; hence the extrudates become non-cohesive, resulting in extrudates with poor durability.[4]

Extruder Torque

The torque required to rotate the extrusion screw is a function of its speed, fill, and the viscosity of the feed ingredient blend in the screw channel.[50] Changing the levels of DDGS, feed moisture content, and extruder barrel temperature had a significantly affected the extruder torque at ∞ = 0.05 (Table 10). Increasing the DDGS levels from 20 to 60% wb resulted in a proportional increase in torque values by 29.3 and 42.0% for cassava and potato starch extrudates, respectively (Table 8). For cornstarch extrudates, extruder torque decreased with an increase in DDGS levels, however. Increasing trend in torque values with corresponding increase in DDGS level was postulated by Chevanan etal.,[19] which was strongly supported and challenged by the results of this present study (for cassava and potato starch extrudates).

Table 9 Treatment combination effects of extruder processing parameters of cassava, corn, and potato starch extrudates. ^†

			Feed moisture content (% wb)
			15			20			25
			Extruder barrel temperature (°C)
Starch source	Property	DDGS levels (% wb)	100	120	140	100	120	140	100	120	140
		20	28.2e-f	29.9e	15.8f-h	51.9c-d	46.2c-d	51.8c-d	74.2b	26.6e-f	45.9d
	Extruder Torque	40	23.0e-f	20.9e-f	7.77g-h	43.8d	26.7e-f	56.1c-d	55.6c-d	19.4e-g	24.9e-f
	(N-m)	60	103.8a	59.2c	43.0d	76.4b	95.9a	52.8c-d	6.38h	5.28h	7.48g-h
			15.4b	20.0a	11.6d	13.1c	8.39f	5.96h-i	6.01h-i	6.98g-h	1.36l
Cassava	Die pressure	20
	(MPa)	40	13.5c	11.6d	5.20i-j	10.9d-e	5.28i-j	3.73k	7.55f-g	1.90l	0.92l
		60	10.9d-e	9.75e	6.04h-i	6.39g-i	4.36j-k	2.05l	2.20l	1.20l	−0.51m
			38.5e-h	23.6i-k	21.5j-l	82.1b	81.5b	43.9d-f	59.9c	51.3c-e	52.3c-d
	Extruder	20
	Torque (N-m)	40	33.5f-j	32.3f-j	18.6k-l	115.2a	80.6b	32.5f-j	51.1c-e	31.7f-j	29.4g-k
Corn		60	60.9c	50.0c-e	41.6d-g	26.1h-k	73.4b	36.6f-i	2.77m	10.7l-m	−0.12m
	Die pressure	20	16.9a	13.4c	14.3b	12.2d	9.44f-g	5.06k	4.50j	4.14l-m	0.89o-p
		40	12.1d	11.3e	10.2f	8.98g	6.15j	3.93l-m	7.52h-i	2.19n	1.02o-p
	(MPa)	60	8.10h	7.77h	6.66i-j	4.73k-l	3.62m	1.59n-o	1.71n-o	0.64p	−0.93q
			24.3h-i	23.6h-i	15.9i-j	50.5f-g	51.3f-g	43.1g	69.2d	62.1d-e	57.8e-f
	Extruder	20
	Torque	40	46.6g	30.5h	17.9i-j	110.0b	64.3d-e	43.1g	67.1d-e	62.1d-e	44.2g
Potato	(N-m)	60	157.2a	47.6g	46.7g	95.1c	97.9c	66.7d-e	19.4i-j	12.9j-k	6.18k
	Die pressure	20	13.1b	16.9a	11.9c	13.4b	9.61e	5.52i-j	4.98j-k	4.64k-l	0.84p
	(MPa)	40	9.64e	10.9d	8.17f	8.59f	6.46g	3.87m	5.69h-i	4.64k-l	1.47o
		60	9.85e	8.63f	6.19g-h	5.37i-j	4.11l-m	1.90o	2.59n	0.80p	−0.79
^†Means followed with the similar letters within each parameter are not significantly different at α = 0.05.

Table 10 Interaction effects of DDGS, moisture content, and temperature on the extruder processing parameters of cassava, corn, and potato starch extrudates.*

	Cassava starch extrudates		Cornstarch extrudates		Potato starch extrudates
Interactions	Torque (N-m)	Die pressure (MPa)	Torque (N-m)	Die pressure (MPa)	Torque (N-m)	Die pressure (MPa)
DDGS	*	*	*	*	*	*
MC	*	*	*	*	*	*
Temp	*	*	*	*	*	*
DDGS*MC	*	*	*	*	*	*
DDGS*Temp	*	*	*	*	*	*
MC*Temp	*	*	*	*	*	*
DDGS*MC*Temp	*	*	*	*	*	*
*Significant at α = 0.05.

Furthermore, increasing the feed moisture content from 15 to 25% wb decreased the torque values for all the three starch extrudates (Table 8). This finding contradicts with the outcomes of Chevanan etal.,[19] who found a direct relation that existed between extruder torque and feed moisture content. Also, in the research conducted by Chevanan etal.[19,20] the authors proved that torque value increases with an increase in extruder barrel temperature. Results showed that for cassava, corn, and potato starch extrudates, the torque values at 100°C were significantly higher by 33.8, 41.5, and 38.5%, respectively than when extruded at 140°C (Table 8). In other words, torque values showed an inverse relation with the extruder barrel temperature. One explanation could be the reduction in viscosity of the ingredient melt in an Arrhenius fashion with an increase in processing temperature, which might have required less torque to rotate the extruder screw. The torque values of various starch extrudates ranged from −0.12 to 157.2 N-m with the lowest (−0.12 N-m) and the highest (157.2 N-m) torque values observed for corn and potato starch extrudates, respectively (Table 9).

Die Pressure

The pressure developed inside the die is a function of various parameters such as rheological properties of the feed ingredient blend and pumping characteristics, in addition to the die dimensions used in the extruder.[20] Changing the levels of DDGS, feed moisture content, and extruder barrel temperature had a significant impact on the die pressure values at ∞ = 0.05 (Table 10). For all changes in the independent variables studied, a sharp decrease in die pressure was noted for all the three starch extrudates. For all the three starch extrudates, increasing the DDGS levels from 20 to 60% wb markedly decreased the die pressure values by 56.8, 53.3, and 48.0%, respectively (Table 8). Similar findings were reported by Chevanan etal.,[19] which holds good for all the three starch sources used in this current investigation. Furthermore, Gujral and Singh[46] pointed out that starch content possessed a direct relation with die pressure, which was consistent with the observations of this study for all the three starch extrudates. This could be attributed to the decrease in the starch content as the ratio of DDGS in the ingredient blend increases, which resulted in lower die pressure values. In another research conducted by Chevanan etal.,[20] the authors found that an inverse relation existed between die pressure and feed moisture content. Results also showed that increasing the feed moisture content substantially decreased the die pressure values for all the three starch extrudates.

In addition, for cassava, corn, and potato starch extrudates, the die pressure of extrudates found at 100°C were significantly higher by 57.2, 43.7, and 43.6%, respectively, when compared the die pressure values of products extruded at 140°C (Table 8). Chevanan etal.,[20] and Lam and Flores,[51] also found decreased die pressure values with corresponding increase in extruder barrel temperature during the extrusion of fish feed. This decrease in die pressure values with increase in processing temperature could be attributed to the decrease in the viscosity of the ingredient melt at higher temperatures, which could have resulted in reduction in die pressure values. The highest and lowest die pressure values were found to be −0.79 MPa and 20.0 MPa for potato and cassava starch extrudates, respectively (Table 9). The negative die pressure values observed in this current study seems to be inapt and hence a closer examination of these data will be undergone in the future studies of extrusion processing of starch and DDGS.

Correlation Analysis

After a closer observation of main treatment, individual treatment combination, and interaction effects between variables, the multivariate data were subjected to Pearson correlation analysis to deepen the understanding of the strength of correlation that existed among the various extrudate properties in the extrusion. Tables 11, 12, and 13 represent the Pearson correlation coefficients (absolute r values shown) of various response variables of cassava, corn, and potato starch extrudates, respectively. A correlation describes the strength of an association between response variables.

Table 11 Pearson Correlation Coefficients of cassava starch extrudates.†

	ER	UD	L*	a*	b*	SV	WAI	WSI	PDI
ER	1
UD	−0.791	1
L*	0.016	0.314	1
a*	−0.105	−0.219	−0.766	1
b*	0.083	−0.243	−0.740	0.434	1
SV	0.618	−0.604	0.178	−0.261	−0.092	1
WAI	−0.679	0.830	0.024	−0.012	−0.022	−0.807	1
WSI	0.102	−0.209	−0.007	0.030	0.037	0.250	−0.375	1
PDI	−0.056	0.238	0.142	−0.310	−0.101	−0.240	0.340	−0.171	1
^†ER: Expansion ratio; UD: Unit density; L*: Brightness/luminosity; a*:redness/greenness; b*: blueness/ yellowness; SV: Sinking velocity; WAI: Water absorption index; WSI: Water solubility index; and PDI: Pellet durability index.

Table 12 Pearson correlation coefficients of cornstarch extrudates. ^†

	ER	UD	L*	a*	b*	SV	WAI	WSI	PDI
ER	1
UD	−0.722	1
L*	0.040	0.231	1
a*	−0.036	−0.227	−0.312	1
b*	−0.117	−0.289	−0.457	0.473	1
SV	0.703	−0.501	0.188	−0.195	−0.220	1
WAI	−0.603	0.467	−0.031	−0.082	0.015	−0.519	1
WSI	0.359	−0.412	−0.235	0.512	0.382	0.132	−0.664	1
PDI	0.059	0.249	0.478	−0.462	−0.535	0.083	0.094	−0.368	1
^†ER: expansion ratio; UD: unit density; L*: brightness/luminosity; a*:redness/greenness; b*: blueness/yellowness; SV: sinking velocity; WAI: water absorption index; and WSI: water solubility index; and PDI: Pellet durability index.

Table 13 Pearson Correlation Coefficients of potato starch extrudates. ^†

	ER	UD	L*	a*	b*	SV	WAI	WSI	PDI
ER	1
UD	−0.962	1
L*	−0.751	0.740	1
a*	0.762	−0.763	−0.762	1
b*	0.683	−0.705	−0.870	0.601	1
SV	0.357	−0.165	−0.219	0.121	0.106	1
WAI	−0.518	0.402	0.430	−0.560	−0.379	−0.453	1
WSI	−0.046	−0.011	−0.025	0.194	0.086	−0.193	−0.484	1
PDI	−0.278	0.324	0.309	−0.394	−0.313	0.049	0.367	−0.535	1
^†ER: expansion ratio; UD: unit density; L*: brightness/luminosity; a*:redness/greenness; b*: blueness/ yellowness; SV: sinking velocity; WAI: water absorption index; WSI: water solubility index; and PDI: pellet durability index.

From this study, we could observe high correlation coefficients that occurred between some of the response variables. As anticipated, there was a high negative correlation between ER and UD with absolute r values >0.70 for cassava and cornstarch extrudates (Tables 11 and 12), where as >0.90 for potato starch extrudates (Table 13). This was highly expected because ER and UD are dependent on the degree of expansion; inversely related, however; the higher the ER, the lower is the UD of the resulting extrudates. Additionally, ER and SV had a positive correlation with absolute r values >0.60 and >0.70 for cassava and cornstarch extrudates, respectively (Tables 11 and 12). This was quite acceptable because ER and SV possessed a direct relation, which means that higher the ER, the better is the ability of the extrudates to float (reduced sinking velocity value). Potato starch extrudates did not possess high correlation between ER and SV, however (Table 13). Furthermore, from Tables 11 and 13, we could notice that a high correlation was exhibited among the color values (a* and L*, b* and L*) of cassava and potato starch extrudates, respectively.

In the overall analysis, potato starch extrudates possessed higher correlation with absolute r values (0.68 ≤ r ≤ 0.96) among most of the response variables studied (UD and ER; L* and ER, L* and UD; a* and ER, a* and UD, a* and L*; b* and ER, b* and UD, b* and L*, b* and a*) in comparison with cassava and cornstarch extrudates (Table 13).

CONCLUSIONS

An attempt to determine the effect of various starch sources on the extrudate properties and extruder processing parameters by varying the levels of DDGS, feed moisture content, and extruder barrel temperature was made. All of the factors (DDGS levels, feed moisture content, and extruder barrel temperature) significantly affected the UD, L*, a*, b*, WAI, and PDI values of cassava starch extrudates. For corn and potato starch extrudates, DDGS levels, feed moisture content, and extruder barrel temperature had a significant impact on ER, a*, b*, SV, WAI, WSI, and PDI values at α = 0.05. Interestingly, all the factors studied had a significant effect on the extruder processing parameters like extruder torque and die pressure values. For all the three starch extrudates, increasing the DDGS levels from 20 to 60% wb significantly decreased the ER, L*, WAI, PDI, and die pressure values whereas increasing the feed moisture content from 15 to 25% wb significantly increased the ER, b*, and WAI values. Furthermore, UD, L*, SV, WSI, torque, and die pressure values significantly decreased for all the three starch extrudates with a temperature increase from 100 to 140°C. Also, correlation analysis on the collected data proved that for all the three starch extrudates, a strong negative correlation existed between ER and UD. In the overall analysis, it was observed that cassava starch and potato starch extrudates were more suitable for floating and sinking aqua feed, whereas cornstarch extrudates were more durable. Further analysis on DDGS based extrusion considering several other parameters (extruder screw speed, net protein content) will be conducted in the near future, which would be more appropriate for tilapia, channel catfish, and rainbow trout.

ACKNOWLEDGMENTS

The financial support from Agricultural Experiment Station, South Dakota State University, Brookings, SD is greatly appreciated. The author's appreciation also extends to Nehru Chevanan, Vykundeshwari Ganesan, and Rumela Bhadra in performing the extrusion.