Impact of Particle Size, Thermal Processing, Fat Inclusion and Moisture Addition on Pellet Quality and Protein Solubility of Broiler Feeds

The present study evaluated the effect of feed particle size, thermal processing, several levels of fat inclusion and moisture addition on pellet quality and protein solubility in potassium hydroxide (KOH) in a corn, soybean meal and animal by products based broiler diets. The different processing factors were combined in a 2 x 4 x 4 x 2 factorial arrangement in an eight randomized block consisting of eight production series: two particle sizes (coarse: 1,041 microns and medium: 743 microns), four fat inclusion levels at the mixer (15, 25, 35 and 45 g/kg of feed), four moisture addition levels in the conditioner (0, 7, 14 and 21 g/kg of feed) and two thermal processing treatments (conditioner-pellet press treatment or conditioner-expander-pellet treatment) which resulted in 64 different processed feeds. For the determination of the pellet durability index (PDI), the amount of intact pellets and protein solubility determinations, eight feed samples (replicates) were collected for each treatment. The data were transformed using a variation of Box-Cox transformation in order to fit a normal distribution (p > 0.05). Adding moisture up to 21 g/kg of feed in the conditioner improved pellet quality of the diets (p < 0.05). Expansion of diets before pelleting improved (p < 0.05) PDI and amount of intact pellets by 26% and 31%, respectively, as compared to a simple conditioning-pelleting feed processing. Expander treatment (at 110 °C) decreased (p < 0.05) protein solubility in KOH from 686 g/kg to 643 g/kg total protein as compared to pelleting process (at 80-82 °C). The amount of intact pellets reduced from 773 g/kg to 746 g/kg of feed (p < 0.05) as particle size increased from medium to coarse grinding. Pellet quality was significantly reduced with fat inclusion levels higher than 35 g/kg of diet.

The aim of a pellet mill is to combine heat, moisture and pressure to make a high percentage of cylindrical and solid structures, the pellets, from mash feed. Pellets should be sufficiently durable to resist compression, impact and abrasion forces during conveyance and storage in feed mill and transportation to the farms [1, 2]. Pellet quality can be measured by the means of intact pellets’ amount/kg of feed and pellet durability index (PDI).

Many factors can affect pellet quality in a feed mill. Particle size plays an important role in pellet press operation. Particles larger than 1,000 microns or 1,500 microns may cause fracture of pellets [3, 4]. For this reason, it is recommended that in a corn-soybean meal diets for broilers, particle size should be approximately 650-700 microns to achieve the highest PDI [5]. Increasing levels of added fat in feed also reduce pellet quality [6-8]. Fat addition reduces the friction between die wall and feed ingredients, and decreases the compression made upon the feed particles inside the die holes [9, 10]. In a corn-soybean meal diet as fat is added to mash feed prior to pelleting, the fat inclusion should not exceed 20 g/kg of feed in order to limit the negative effect on the pellet quality [11].

When water was added to the mixer in a proportion of 25 g/kg and 50 g/kg, it showed a positive effect of adding water on PDI of pelleted diets [8]. In another study the addition of 24 g of moisture/kg of feed conditioned at 60 °C, increased the PDI from 56.5% to 67.2% [12]. A survey of North American feed mills indicated that diets submitted to expander treatment before pelleting as compared to conventional condition-pelleting improved pellet quality by 15%-25% [13].

The protein digestibility of feed can be improved with processing (i.e., temperature, moisture, pressure and time); however, excessive processing can also reduce digestibility or in vitro assessment of the feed’s protein solubility [14]. After an expander treatment at a temperature of 130-136 °C, soybean meal protein solubility in potassium hydroxide (KOH) decreased from 805.7 g/kg to 641.2 g/kg of total protein content [15].

Some strategies can be implemented to enhance pellet quality. The productivity and revenue of broiler industry may be improved by correct application of these strategies. The objective of the present study is to evaluate particle size, added moisture during conditioning, dietary fat level and the use of two different thermal processing methods on pellet quality and protein solubility. 

The experimental diets were manufactured in a broiler integrator feed mill located in Rio Grande do Sul-State, Brazil. The feed was a corn-soybean meal and animal by-products based broiler grower diets with varying levels of fat inclusion (15, 25, 35 and 45 g/kg of feed) (Table 1). 

Table 1 Composition of the experimental diets.

 

All ingredients were weighed and the batch were ground in a post-mill hammer mill (Buhler VHM with 16 hammers-10 t/h output/unit x 6 units) with 5.0 mm diameter hole sieves to obtain the medium and coarse particle sizes of the final diet. Different diet particle sizes were attained by changing the hammer tip speed, as the hammer mill was equipped with a variable hammer rotation control (3,600 rpm for medium size grinding and 1,800 rpm for coarse size grinding). All the diet components were blended in a paddle type mixer (Buhler DFML 8000L). Mixing time was divided into three phases: dry mixing (45 s), liquid addition (60-90 s) and wet mixing (25 s). The different fat inclusions in the diet were obtained by spraying fat into mash feed during the liquid addition phase of the mixing time.

For diets submitted to conditioner-expander-pelleting treatment, feed were steam-conditioned for 15 s at 80-82 °C under a steam pressure of 1.5-2.0 bar in a standard barrel type conditioner. In sequence, feed was directed to an expander (Kahl Expander, Model 38.2 with annular gap, 2,000 mm length and 400 mm, with an energy consumption of 11-13 kW/t·h and 65 bar of cone pressure), with an average feed retention time of 5 s, in a mean temperature of 110 °C and to pellet press (Buhler, Model DPAS) with die specifications: 660 mm diameter, die holes with 60 mm depths and 4.5 mm diameter and with no relief. For simple conditioner-pelleting treatment, diets were steam-conditioner and in sequence these diets were directed to pellet press using the same operation parameters that were used for conditioner-expander-pelleting treatment. The control of the different levels of moisture addition in the conditioner were controlled by means of a water proportioning system (Kahl WD-GLI15, water temperature of 60 °C and water pressure of 3-6 bar). The output of the conditioner was used to calculate the amount of moisture added to mash feed and moisture sensors set in the end of the conditioner were used to monitor this addition. Feed just prior to entering pellet press were sampled (a composite sample for each moisture addition level) for moisture content analysis.

2.2 Production and Sampling of Feed The experiment was based on an eight randomized block design consisting of eight production series repeated along the eight weeks of experimental period. The four main factors (two particle size, four moisture addition levels, four fat inclusion levels and two thermal processing) were combined in a 2 x 4 x 4 x 2 factorial arrangement and totaling 64 different combinations of processed diets which were entirely replicated eight times (eight production series). A total of 1,024 t of feed were manufactured along this study. Ingredients were firstly ground in two different particle sizes (coarse and medium) and then combined in four different fat inclusion levels in the mixer (fat added at a rate of 15 g/kg to 45 g/kg of feed—Table 1). These combinations of mash feed were dispatched to the conditioner where the batches were submitted to different moisture additions (0, 7, 14 and 21 g/kg of feed) and thermal processing methods (conditioner-pelleting and conditioner-expander-pelleting treatments).

The sampling points were: (1) in the mixer discharge: where a composite sample of mash feed, for each treatment and production series, were collected for particle size, moisture content and chemical composition check; (2) prior to pellet press: where a composite sample of conditioned and conditioned-expanded feed for each treatment and production series, were collected for moisture content monitoring; (3) between pellet press and the cooler: where eight composite samples of pelleted feed (one sample in each production serie), corresponding to eight replicates, were analyzed for pellet quality, moisture content, chemical composition and protein solubility in KOH. A stainless scoop with a 40 cm rod was used for feed collection and samples were packaged in an identified plastic sample bags. The average environment temperature and relative humidity recorded during the experimental period (November to December 2012) were 24 °C and 62%, respectively.

The feed from different treatments were analysed for the following physical and chemical parameters:

PDI: determined accordingly to Method S269.4 from ASAE Standards [2];

Amount of intact pellets: 200 g of feed is sieved in a screen of 3.0 mm round holes screen and the portion of feed that is retained in this screen is considered as intact pellets;

Protein solubility in KOH: determined according to method described by Parsons et al. [16];

Moisture content: determined accordingly to the Method 930.15 referenced at the Association of Official Analytical Chemists International [17];

Particles size: determined accordingly to the method referenced by the American Society of Agricultural and Biological Engineers [18].

The statistical model included the replicates, particle size, type of thermal processing, levels of fat and moisture addition and interactions between the factors: Yijklm = μ + PRh+ PSi + TPj + FIk + MAl + (PS x TP) ij + (PS x FI) ik + (PS x MA) il +(TP x FI) jk + (TP x MA) jl + (PS x TP x FI) ijk + (PS x TP x MA) ijl + (PS x FI x MA) ikl + (TP x FI x MA) jkl + (PS x TP x FI x MA) ijkl + εhijklm

where: Yhijklm = the characteristic (PDI, amount of intact pellet, protein solubility in KOH) of the feed sample, μ = population mean, PRh = production series effect (h = production series ranging from 1 to 8), SE PSi = particle size effect (i = medium or coarse particle size), TPj = thermal processing effect (j = conditioner-pelleting or conditioner-expander-pelleting treatmentes), FIk = fat inclusion effect (k = 15, 25, 35 and 45 g/kg of feed), MAl = moisture addition effect (l = 0, 7, 14 and 21 g/kg of feed), and εijklm = residual error (assumed to be normally distributed with mean zero).

A general linear model was employed to analyze the effects of categorical and quantitative factors present in the statistical model. The statistical package Statistica 8.0 version (StatSoft, Inc.) was used to perform the analysis of variance of collected data. The Anderson-Darling test was used to check the normality of the residuals of the estimated model for the amount of intact pellets, PDI and protein solubility in KOH. According to Razali and Wah [19], Shapiro-Wilk test and Anderson-Darling test are the most powerful normality test followed by Lilliefors test and Kolgomorov-Smirnov test. The Hartley F-max test was used to check the homoscedasticity of the categorical factors data (thermal processing and particle size). Linear and quadratic effects due to moisture and fat addition to the feed were determined by linear regression. The significance of the performed tests were accepted if p ≤ 0.05. 

The residuals for the amount of intact pellets, PDI and protein solubility did not follow a normal distribution (Anderson-Darling test p < 0.05). To fit these residuals in a Gaussian distribution, these data were submitted to Johnson transformations (variation of Box-Cox transformations) : (1) amount of intact pellets = -0.4984 + 0.6016 x Ln ((amount of intact pellets g/kg – 460.565)/(929.921 – amount of intact pellets g/kg)); (2) PDI = -0.4998 + 0.7980 x Ln ((PDI % – 47.6941)/(95.8331 – PDI %)); (3) protein solubility in KOH = -0.3310 + 2.6324 x Asinh ((protein solubility in KOH g/kg – 64.790)/15.074). All of the statistical analyses and discussions were performed upon the transformed data.

The Hartley’s test, performed for moisture content, PDI, the pellets percentage and protein solubility in KOH, showed a similar variance inside each categorical parameter (F-max was smaller than F-critical at a 0.05 significance level).

Initially data were analyzed as the original statistical model. Therefore, the factor’s coefficients and significances as well as the entire model’s R² values were defined in a full factorial design: two particle sizes x two thermal processings x four moisture addition levels x four fat inclusion levels (Table 2). Backward elimination method was employed to remove the non-significant factors from the prediction equations (Table 3). The resultant models presented R² values close to that of the entire model for the PDI (0.863 and 0.853, respectively, for the entire and backward models) and the amount of intact pellets (0.944 and 0.942, respectively, for the entire and backward models). The R² values of protein solubility in KOH differed after and before backward elimination but it still kept its low value (0.341 and 0.265, respectively, for the entire and backward models).

The feed submitted to conditioner-expander-pelleting treatment resulted in a higher pellet percentage and PDI (862 g/kg of feed and 87%) as compared to conditioner-pelleting treatment (657 g/kg of feed and 69%). These results are in accordance with a survey of North American feed mills which indicated that diets processed on expander before pelleting as compared to conventional conditioning-pelleting improved pellet quality by 15%-25% [13]. Van der Poel [20] reported that conditioner-expander-pelleting treatment increased pellet durability and pellet hardness in a pig diet containing tapioca, pea and soybean meal as compared to steam-pelleting treatment. In the present study, as the conditioner-expander-pelleting treatment was submitted to an additional heat treatment (feed was submitted for 5 s to a mean temperature of 110 °C) and pressure (energy consumption in the expander was 11-13 kW/t of feed/h) over the steam-pelleting process, these processing parameters may have led to higher starch gelatinization and compression on the expander treatment.

Thermal processing interacted with the effect of moisture on the amount of intact pellets. In the conditioner-expander-pelleting treatment, addition of moisture from 0 g/kg to 21 g/kg of feed improved the amount of intact pellets from 817 g/kg to 901 g/kg of feed. The gain observed in conditioner-pelleting treatments in the same moisture addition range, was from 611 g/kg to 714 g/kg of feed. In a similar manner, Lundblad et al. [21] observed that as water was added at a proportion of 0 g/kg to 30 g/kg of feed, there were improvements in the PDIs from 79% to 87% and from 92% to 94% for the conditioner-pelleting and conditioner-expander-pelleting treatments, respectively. A quadratic effect (concave curve) of moisture addition on PDI was confirmed in the present study (p < 0.001).

The diet’s particle sizes achieved in the tests diets were respectively 1,041 microns for the coarse grinding and 743 microns for the medium grinding. Coarser particle size impaired the amount of intact pellets compared to medium particle size. As average particle size of feed changed from medium to coarse size, the amount of intact pellets was reduced from 773 g/kg to 746 g/kg of feed (Table 2). This finding is also in accordance with Dozier [5] who verified that in a corn-soybean meal diets the ideal particle size is approximately 650-700 microns. In addition, other researchers [3, 4] reported that increasing particle size negatively affected pellet quality. Because binding strength between solids can be calculated by the Laplace’s equation (binding strength = 2 x (liquid surface tension/particle diameter/2)), increasing particle size weakens the bonds inside the pellets [6].

 

Table 2 Effect of thermal processing (TP), particle size (PS), moisture addition (MA) and fat inclusion (FI) on pellet quality and protein solubility in KOH of the diets.

 

 

There was an interaction between particle size and thermal processing; the negative effect of coarser particle size on the amount of intact pellets was more substantial for conditioner-pelleting treatments than for conditioner-expander-pelleting treatments. 

Table 3 Regression equation for the amount of intact pellets adjusted by backward elimination method.

Increasing particle size decreased the amount of intact pellets from 677 g/kg to 636 g/kg and from 868 g/kg to 856 g/kg for the conditioner-pelleting and the conditioner-expander-pelleting treatments, respectively (Fig. 1). In conditioner-expander-pelleting treatment, the different particle size did not affect the amount of intact pellets; changes in the feed texture by temperature, pressure and moisture of the annular gap expander probably overwhelmed the breaking forces of coarse particles present in coarser grinding.

The average moisture content of mash feed samples prior to conditioning was 102 g/kg of feed. Feed samples collected prior to pellet press confirmed that moisture addition at the conditioner (0, 7, 14 and 21 g/kg of feed) increased moisture content of feed prior to pelleting to 130, 138, 145 and 151 g/kg of feed in average. The moisture content of samples collected at the discharge of pellet press and prior to cooler was 7 g/kg higher for conditioner-expander treatment than for single conditioner treatment (146 g and 139 g of moisture/kg of feed, respectively). Moisture content of heat processed feed in this study was higher than the average values of feed industry once samples were collected prior to cooler.

The amount of intact pellets increased linearly (p < 0.001) with increasing moisture addition in the conditioner (Table 2). The data obtained in the present study shows an increase of 101 g of pellets/kg of feed as water addition in conditioner increased from 0 g/kg to 21 g/kg of feed. Moisture addition resulted in a linear and positive effect (p ≤ 0.05) on the amount of intact pellets and a quadratic response on the PDI. Similar results were reported by Moritz et al. [8] and Buchanan et al. [22] who tested respectively the addition of 25 g to 50 g and 20 g to 40 g of moisture/kg of feed and observed a positive effect on pellet quality. In the same way, Lundblad et al. [21] concluded that addition of 0 g to 30 g of water/kg of mash feed prior to conditioning improved the PDI from 84% to 89% in a maize based pig diets. However no significant differences in pellet quality of broiler diets were reported when a mixture of moisture-mold inhibitor was added to feed at a rate of 0 g/kg to 20 g/kg of feed [22].

Interactions between moisture addition and different processing factors were observed (Tables 2-4 and Figs. 1 and 2). When moisture was added in the conditioner (0 g/kg to 21 g/kg of feed) coarsely ground feed responded by increasing the amount of pellet from 693 g/kg to 803 g/kg of feed. However, in the feed with medium particle size, the amount of pellets improved from 735 g/kg to 812 g/kg of feed. Water addition enhanced pellet quality due to its capillary property which helps keeping particles together [6]. Interaction between particle size and moisture addition may be explained by the fact that once particle to particle interactions weakens proportionately to particle size, the capillary property of moisture on the pellet quality becomes more significant for coarser than medium particle sizes. 

Fig. 1 Interaction of thermal processing, particle size, moisture addition and fat inclusion on the amount of intact pellets (g/kg of feed).

  

Table 4 Regression equation for PDI adjusted by backward elimination method.

  

Table 5 Regression equation for protein solubility in KOH adjusted by backward elimination method.

  

Fig. 2 Interaction of thermal processing, particle size and fat inclusion on PDI (%).

Fat inclusion had a quadratic effect (p ≤ 0.05) on both the amount of intact pellets (convex curve) and the PDI (concave curve). The amount of intact pellets declined quickly as fat inclusion surpassed 35 g/kg of feed, whereas the PDI showed a strong decline as fat inclusion levels exceeded 15 g/kg of feed. Mackinney et al. [23] verified that as fat inclusion increased from 0 g/kg to 50 g/kg of feed, the amount of intact pellets in the feed decreased from 900 g/kg to 490 g/kg of feed. In the present study the amount of intact pellets in the conditioner-expander-pelleting treatment decreased 8% as the fat inclusion varied from 15 g/kg to 45 g/kg of feed, where as in the conditioner-pelleting treatment the amount of intact pellets reduced 26% with increasing levels of fat inclusion. Similar negative effects of fat inclusion on pellet quality were reported previously by Moritz et al. [8] that as average fat addition varied from 36 g/kg to 50 g/kg of feed, the PDI decreased from 75% to 54%. These authors reported that an increase in oil inclusion may have led to reduced frictional heat and an increased feed flow rate on the press which decreased the gelatinized starch from 196 g/kg to 100 g/kg of feed and thus compromising pellet resistance.

An interaction between fat inclusion and particle size was observed for the PDI. An increase in fat inclusion from 25 g/kg to 35 g/kg in medium particle size feed increased the PDI, such behavior was not observed in coarse particle size feed. There was also a combined effect of thermal processing and fat inclusion. The PDI in the conditioner-pelleting treatment decreased from 83% to 60% as the fat inclusion rose from 15 g/kg to 45 g/kg of feed, this negative effect of fat inclusion in conditioner-expander-pelleting treatment was from 91% to 82%. Likewise, the amount of intact pellets in the conditioner-expander-pelleting treatment decreased from 874 g/kg to 807 g/kg of feed as fat inclusion increased from 15 g/kg to 45 g/kg of feed, whereas in the conditioner-pelleting treatment the amount of pellets reduced from 694 g/kg to 550 g/kg of feed for increasing levels of fat inclusion.

Protein solubility (KOH) of mash feed samples collected before thermal processing treatment and soybean meal used in this assay had an average value of 705 g/kg and 819 g/kg of protein, respectively.

Protein solubility of diets decreased in conditioner-expander-pelleting treatment (645 g/kg of protein) as compared to conditioner-pelleting treatment (686 g/kg of protein) (Table 5). A mixture of broken peas, lupins and faba beans was submitted to different thermal processing, and protein dispersibility and nitrogen solubility in water were significantly decreased after expander treatment (14% units decrease relative to non-heat treated feed) when compared to pelleting (10% and 11% units decrease relative to non-heat treated feed) [24]. Malumba et al. [25] submitted corn kernels to different air drying temperatures (54 °C and 130 °C) and verified that high drying temperatures induces the insolubilisation of salt-soluble proteins and zeins. In accordance, when soybean meal was expander treated at 130-136 °C the protein solubility in KOH decreased from 81% to 64% [15].

Coarsely ground diets showed lower protein solubility (663 g/kg of protein) as compared to medium ground diets. In contrast, Benhke [26] stated that moisture and heat penetration is more effective in smaller particles with high surface area per unit weight. Accordingly to this statement, protein present in smaller particles should be subjected to higher thermal gradients with consequent protein denaturation. Because protein solubility in KOH is a method applied for soybean meal but not for composite diets, it is important to consider the findings of these studies as exploratory data.

This study indicates that conditioner-expander-pelleting treatment, moisture addition up to 21 g/kg of feed and fat inclusion limited to 25g/kg of feed are effective strategies to enhance pellet quality. Coarser particle size can impair pellet quality but its effect is more pronounced in conditioner-pelleting treatment. Protein solubility in KOH was decreased as feed was submitted to expander-conditioner in comparison to single conditioning process. Regression equation is a promising tool to model the effects of processing factors and feed formulation on pellet quality. 

The authors thank Ivo Pasche, Lanes Tramontina, Valter Vivan, Valter Sbaraini and João Carlos Barato for their contribution and input to this study. This research was made possible by the support of Fabrício da Silva Delgado, Luiz Stabile Benício and Alessandro Serafin Lopes.

[1] W.B. Cavalcanti, K.C. Behnke, Effect of composition of feed model systems on pellet quality: A mixture experimental approach, II. Cereal Chemistry 82 (2005) 462-467.

[2] R.J. Mina-Boac, G. Maghirang, M.E. Casada, Durability and breakage of feed pellets during repeated elevator handling, Material Written for ASABE Annual International Meeting, ASABE, Portland, Oregon, 2006.

[3] M. Franke, A. Rey, Improving pellet quality and efficiency, Feed Tech. 10 (3) (2006).

[4] J.R.I.E. Mendez, G. Santomá, Feed Manufacturing, the Nutrition of the Rabbit, Cab International, 2008.

[5] W.A. Dozier, III. Cost effective pellet quality for meat birds, Feed Management 52 (2) (2001) 1-3.

[6] M. Thomas, A.F.B. van der Poel, Physical quality of pelleted animal feed. 1. Criteria for pellet quality, Anim. Feed Sci. Technol. 61 (2000) 89-112.

[7] J.L. Briggs, D.E. Maier, B.A. Wakins, K.C. Behnke. Effect of ingredients and processing parameters on pellet quality, Poult. Sci. 78 (1999) 1464-1471.

[8] J.S. Moritz, K.R. Cramer, K.J. Wilson, R.S. Beyer, Feed manufacture and feeding of rations with graded levels of added moisture formulated to different energy densities, J. Appl. Poult. Res. 12 (2003) 371-381.

[9] R. Kulig, J. Laskowski, Effects of conditioning parameters on pellet temperature and energy consumption in the process of plant material pressing, TEKA Kom. Mot. Energ. Roln.—OL PAN 8a (2008) 105-111.

[10] A.C. Fahrenholz, Evaluating factors affecting pellet durability and energy consumption in a pilot feed mill and comparing methods for evaluating pellet durability, Thesis presented in Kansas Universtiy, Manhattan, Kansas, 2012.

[11] D.A. Fairfield, Pelleting for profit-feed and feeding digest, National Grain and Feed Association Part 1, No. 6, 2003, p. 54.

[12] M.R. Abdollahi, V. Ravindran, T.J. Wester, G. Ravindran, D.V. Thomas, Effect of improved pellet quality from the addition of a pellet binder and/or moisture to a wheat-based diet conditioner at two different temperatures on performance, apparent metabolisable energy and ileal digestibility of starch and nitrogen in broilers, Anim. Feed Sci. Technol. 175 (2012) 150-157.

[13] I. Fancher, D. Rollins, B. Trimbe, Feed processing using the annular gap expander and its impact on poultry performance, J. Appl. Poult. Res. 5 (1996) 386-394.

[14] E. Prestlokken, F. Fôrutvikling, Expander treatment, HFE 305 Feed Manufacturing Technology [Online], http://www.umb.no/statisk/iha/kurs/nova/feed_technolog y/4.pdf (acessed Feb., 2012).

[15] J.A.F. Veloso, S.L.S. Medeiros, C.L.C. Arouca, N.M. Rodriguez, E.O.S. Saliba, S.G. Oliveira, Chemical composition, physico-chemical and nutritional evaluation and the effect of expander treatment on corn and soybean meal or growing pigs, Arq. Bras. M. Vet. Zoot. 57 (5) (2005) 623-633. (in Portuguese)

[16] C.M. Parsons, K. Hashimoto, K.J. Wedekind, D.H. Baker, Soybean protein solubility in potassium hydroxide: An in vitro test of in vivo protein quality, J. Anim. Sci. 69 (1991) 2918-2924.

[17] Association of Official Analytical Chemists International (AOAC), Method 930.15 in Official Methods of Analysis, 16th ed., AOAC, Arlington, VA, 1998.

[18] American Society of Agricultural and Biolgical Engineers (ASABE), Method of Determining and Expressing Fineness of Feed Materials by Sieving. American National Standard Institute S319.3, Feb. 3, 2006.

[19] N.M. Razali, Y.B. Wah, Comparisons of Shapiro-Wilk, Kolgomorov-Smirnov, Lilliefors and Anderson-Darling tests, J. Statistic. Model. and Anal. 1 (2011) 21-33.

[20] A.F.B. van der Poel, H.M.P. Fransen, M.W. Bosch, Effect of expander conditioning and/or pelleting of a diet containing tapioca, pea and soybean meal on the total tract digestibility in growing pigs, Anim. Feed Sci. Technol. 66 (1997) 289-295.

[21] K.K. Lundblad, J.D. Hancock, K.C. Behnke, E. Prestløkken, L.J. McKinney, M. Sørensen, The effect of adding water into the mixer on pelleting efficiency and pellet quality in diets for finishing pigs without and with use of an expander, Anim. Feed Sci. Technol. 150 (2009) 295-302.

[22] N.P. Buchanan, J.S. Moritz, Main effects and interactions of varying formulation protein, fiber, and moisture on feed manufacture and pellet quality, J. Appl. Poult. Res. 18 (2009) 274-283.

[23] J.M. Hott, N.P. Buchanan, S.E. Cutlip, J.S. Moritz, The effect of moisture addition with a mold inhibitor on pellet quality, feed manufacture, and broiler performance, J. Appl. Poult. Res. 17 (2008) 262-271.

[24] L.J. Mackinney, N.D.O. Skinner, R.G. Teeter, Pellet Quality Effects on Broiler Growth and Efficiency, Oklahoma Agriculture Experiment Station Report , 2001.

[25] J.O. Goelema, A. Smits, L.M. Vaessen, A. Wemmers, Effects of pressure toasting, expander treatment and pelleting on in vitro and in situ parameters of protein and starch in a mixture of broken peas, lupins and faba beans, Anim. Feed Sci. Technol. 78 (2006) 109-126.

[26] P. Malumba, S. Janas, T. Masimango, M. Sindic, D. Claude, F. Béra, Influence of drying temperature on the wet-milling performance and the proteins solubility indexes of corn kernels, J. Food Eng. 95 (2009) 393-399.