INTRODUCTION
Physical treatments, such as expansion and pelleting, are used in the processing of diets with the aim of enhancing feed efficiency. This occurs because during these processes, in addition to starch granule gelatinization, there are also partial protein denaturation, which leads to increased food digestibility and consequently improved broiler performance (Oliveira et al., 2011). Besides technological benefits, feed processing methods are associated with the reduction of feed wastage and feed selection by the animals, less time and energy spent feeding, a decreased segregation of feed ingredients and reduction of pathogenic organisms as well as anti-nutritional factors (Amerah et al., 2007; Abdollahi et al., 2013a; Sellers et al., 2020).
However, physical pellet quality is a critical factor to optimize feed efficiency and growth response of broilers (Abdollahi et al., 2013b; Abdollahi et al., 2018a). Pelleting feed has demonstrated an array of benefits to broiler performance that have varied in magnitude based on pellet quality or the percent of intact pellets provided (Glover et al., 2016). Past research has described increased live weight gain, and decreased feed conversion ratio for broilers fed high quality pellets (Corzo et al., 2011; Lilly et al., 2011).
Factors like size, shape and structure of particles influence digestibility and dispersity of nutrients, apart from altering feed density, pellet quality, fluidity of ingredients within the mixing system, transportation and supply of diet to feeders, along with raising the energy consumed in milling (Ribeiro et al., 2002). Mass moisture is another important factor in pellet quality during processing. Both water and steam added to the mixer during conditioning enhances particle adhesion, thus improving pellet quality, as documented in studies by Buchanan and Moritz (2009).
According to Muramatsu et al. (2013) some strategies can be implemented to enhance pellet quality, and through such the productivity and revenue of the broiler industry may be improved by correct application of these strategies. Therefore, the standardization of moisture, particle size and the processing method are sought, with the objective of decreasing the cost associated with the manufacturing process of diets, as well as increasing nutrient digestibility and broiler performance.
Given that, the aim of this study was to assess the effect of pelleted and pelleted-expanded corn-soybean meal diets prepared with different moisture addition levels (0.8 and 1.6% of water addition to the mixer) and two different particle sizes of ingredients (650 and 850µ) on the performance, carcass yield, metabolizable energy of feeds and ileal digestibility of amino acids of broilers.
MATERIAL AND METHODS
All experimental procedures complied with Ethics Committee on the Use of Animals (protocol number 015/2014) of the Instituto Federal Goiano, Campus of Rio Verde – GO. The experiment was carried out using 720 one-day-old male broiler chicks of the Cobb strain, with a mean body weight of 42 ± 2.1g.
The experimental design was completely randomized in a factorial scheme 2 × 2 × 2 (pelleted or expanded-pelleted feed processing; 0.8% and 1.6% moisture addition in the mixer; particle size of 650 and 850µ) with six replications of 15 birds each.
Birds were housed in cages (90 × 60 × 40 cm) equipped with trough feeders and drinkers, a 100 W lamp for heating, and metal trays for excreta collection. The birds remained under constant lighting (natural and artificial) with temperature check twice a day and curtain handling. The average temperature registered during the experiment was 33 °C ± 3.9°C, with a minimum of 17 °C and a maximum 38 °C.
The broilers received water and feed ad libitum during the initial (1 to 21 days), growth (22 to 35 days) and final (36 to 40 days of age) rearing phases. Feeds (based on corn and soybean meal) were formulated according to the recommendations of the Rostagno et al. (2011) with adaptations in some nutritional levels to meet the demands of the local integrator (Table 1). Due to the use of an industrial mixer with high production capacity, for each experimental diet, batches of 6,000 kg were produced. The surplus feed not used in the research was distributed to a local poultry farmer.
Initially, the ingredients were ground using a hammer mill of joint grinding, setting 5-mm sieves for particles of 650µ and 6.5-mm sieves for 850µ. Mixing time was divided in three phases: dry mixing, liquid addition, and wet mixing. For each particle size (PS) of diets (650 and 850µ) water inclusion levels (0.8 or 1.6% moisture level (ML)) were added during the liquid addition phase mixing using injection nozzles.
These experimental diets were produced at an integrated feed mill, specifically for broilers, under different processing types, expansion and pelleting, using a Kahlexpander and a Buhler pelletizer, respectively.
For the pelleted feed, diets were steam-conditioned for 34 seconds at 82°C under a steam pressure of 1.5- 2.0 bar. For the expanded-pelleted feed processing the mash was then transported to an expander, with retention time of 11 seconds, average temperature of 130°C and matrix opening of 4.00 mm. Later, the mash was transported to the pellet press using the same equipment parameters that were used for the simple pelleting treatment.
The percentage of pellet (%) was determined by feed sieving, separating the thin particles from the thick ones by means of a perforated container under agitation (Advantech Manufacturing, INC., 2001). Pellet durability index (PDI, %) was calculated from a 500-g sample of intact pellets placed in a rotating box at 50 rpm for 10 minutes. Subsequently, the feed portion was sieved through a 3.0-mm mesh sieve, while considering in the PDI calculation only the weight of the material that remained on the sieve (Lowe, 2005; Froetschner, 2006). Both pellet percentage and PDI values of feeds are shown on Table 2.
Nutrient digestibility and AMEn evaluation
In order to determine the energy utilization (apparent metabolizable energy – AME and apparent metabolizable energy corrected for nitrogen balance – AMEn) of the experimental feeds, samples of excreta were collected from 10 to 13 and from 27 to 30 days.
During this period, feed intake was measured and underneath each cage, we installed an aluminum tray covered with plastic film for collection of excreta. The sampling procedure was performed at 8:00 am and at 3:00 pm each day, where the samples were packed in plastic bags and frozen for further analyses of dry matter, gross energy and nitrogen content according to Silva & Queiroz (2002). The values of AME and AMEn were calculated through the equations described by Matterson et al. (1965).
The coefficients of ileal digestibility of amino acids and crude protein were determined by collections of ileal digesta samples at 21 days of age. Kaolin, a source of silica, was added to the feed during the initial rearing phase at a level of 1.00%, in order to increase levels of acid-insoluble ash, which was used as an indicator of indigestibility.
At 21 days of age, five birds per replication were slaughtered by cervical dislocation. Shortly after, the ileum was exposed by opening the abdominal cavity, removing a 30-cm segment ending at 4 cm before the ileocecal junction, which was subsequently stored in a plastic container (Adedokun et al., 2007).
Samples of the experimental feed, excreta and digesta were sent to the Animal Nutrition Laboratory of the IF Goiano, campus Rio Verde - GO, Brazil, for dry matter, crude protein and gross energy measurements, according to Silva and Queiroz (2002). The samples of digesta were dried in a lyophilizer to avoid amino acid losses. The total amino acids in feeds and digesta were determined using high-performance liquid chromatography (HPLC), which was carried out by Ajinomoto Animal Nutrition®.
The content of acid-insoluble ash, which is an indigestible fraction in diets and digesta, was measured according to the methodology described by Carvalho et al. (2013). From the laboratory results, we calculated the coefficients of ileal digestibility of amino acids and crude protein according to equations described by Sakomura & Rostagno (2016).
Performance and yields of carcass and cuts
The poultry performance was assessed through the following variables: feed intake, feed conversion and weight gain at 21 and 40 days of age. In this regard, birds and feeds were weighed at the beginning and at the end of each phase.
Feed conversion was calculated based on the average feed intake and average weight of the broilers at the end of the raising period, discounting initial weight. The number of dead broilers was registered daily for feed conversion correction.
Carcass, chest, thigh, drumstick and abdominal fat yields were evaluated at 40 days of age by slaughtering two birds per replication. Before slaughtering by incising the jugular veins and arteries, the broilers had undergone fasting for 12 hours.
Carcass yield percentage was calculated by dividing the carcass weight by the live weight of birds; and the yields of cuts as thigh, drumstick, chest and abdominal fat were calculated in relation to the eviscerated carcass weight.
Data analysis
Data were assessed by the statistical software SISVAR version 5.6 (Ferreira, 2014) and underwent a variance analysis with means compared by use of the Tukey test at 5% significance.
RESULTS AND DISCUSSION
Nutrient digestibility and AMEn
Both for the starter (10 to 13 days of age) and growth (27 to 30 days of age) phases, processing conditions had no effect on AMEn, as shown on Table 3. Nevertheless, there was significant interaction when considering the type of processing of ration, moisture levels and particle size of ingredients for AMEn for the two rearing periods.
During the collection period from 10 to 13 days of age, in line with the statistical breakdown, we emphasize the highest AMEn values for expanded feeds with1.6% moisture addition and particle sizes 850µ. Conversely, when adding 0.8% moisture level in pelleted feeds with 850µ or 1,6% of moisture/650µ particle size, higher values for AMEn were encountered (Table 4). By breaking down the same interaction, over the period from 27 to 30 days of age, we found no significant differences between the individual factors (particle size and moisture), according to the F test, and hence the interaction was not considered.
When assessing the main effects, higher coefficients of ileal digestibility of amino acids were found in pelleted feeds compared to expanded feed (except for methionine). The addition of 1.6% moisture into the mixer provided higher coefficients of ileal digestibility crude protein (Table 5).
There was a significant interaction between feed processing × particle sizes of milling for threonine, methionine+cystine, tyrosine and arginine (Table 5). For these amino acids, in the statistical breakdown of the interaction (processing × particle size), an increase in the coefficients of ileal digestibility was noted by birds fed expanded feeds with 650µ of particle size; while for pelleted diets, the best results were obtained with a particle size of 850µ.
Differently, for lysine, cystine, glycine, alanine, isoleucine, leucine, serine and valine (Table 6), there occurred a triple interaction (type of processing x particle size of ingredients x moisture levels). In the breakdown of the triple interaction, a higher amino acid digestibility was noted for expanded or pelleted diets with 1.6% of moisture (independent of particle size, 650 or 850µ). However, for diets with 0.8% of moisture, better results of digestibility of amino acids were obtained with granulometry of ingredients of 650µ for expanded and 850µ for pelleted diets.
It is understood that the moisture added during the manufacture of pelleted and expanded feed may interact with other components in the feed, aiding thermomechanical reactions that occur in the pellet. As the temperature increases, the water molecules reach a higher energy state contributing to destabilize proteins and improve their digestibility (Buchanan & Moritz, 2009; Svihus & Zimonja, 2011). Protein denaturation and inhibition of antinutritional factors by the heat treatment of the diet improve protein digestibility and amino acid bioavailability as verified by Mauron (1990).
On the other hand, reduced ingredient particle sizes enlarges feed surface area regarding its volume, thus increasing the contact points between particles. As a result, there is an increase in the interatomic adhesion forces, which increases the capillary force between the solid and liquid phases of pellets, as well as enhancing heat and moisture penetration into feed particles within shorter thermal treatments (California Pellet Mill Co., 2012). These effects may result in greater starch gelatinization and protein denaturation improvement of pellet quality and, consequently, the increased nutrient digestibility (Kilburn & Edwards, 2001; Chewning et al., 2012).
The reason for lower ileal digestibility coefficient of some amino acids obtained when associating0.8%/850µ in expanded diets or 0.8%/650µ in pellet diets is still not clear. However, a general form, pelleting resulted in higher amino acid digestibility compared to expanded feed. Similar to these results, lower digestibility coefficient of total protein and/or amino acids in expanded feeds versus pelleted feeds was observed for Boorojeni et al. (2014) Kim et al. (2018) and Lierman et al. (2019); respectively. Protein and amino acids digestibility can be reduced depending on the intensity of processing, with high temperature and conditioning time (Massuqueto et al., 2020).
It is acknowledged that high temperatures during expander treatment can result in denaturation of proteins or destruction of feed additives, such as in those enzymes that promote digestibility. Furthermore, interactions of high temperatures, moisture and sugar could result in Maillard reaction products, reducing protein and starch solubility and digestibility (Thomas et al.,1998; Svihus & Zimonja et al., 2011). Therefore, in this study the temperature of 130°C during11 seconds to which the expanded feeds were submitted may have been a limiting factor of maximum amino acid digestibility.
Moreover, high values of AMEn for broilers in the starter phase (10 to 13 days of age) was determined when the particle size of 850µ was used in expanded diets (0.8% of moisture) and pelleted feeds (0.8 or 1.6% of moisture). Similar to these results, Naderinejad et al. (2016) observed increases in EMAn in pelleted diets, as a result of medium and coarse grindings (DGM of maize 651µ and 796µ, respectively). The authors attributed those improvements in starch digestibility and AME occurred with the increase in particle size in pelleted diets, were associated with higher gizzard weights and reduction in gizzard pH. Coarse feed particles are known to stimulate gizzard activity, resulting in larger numbers of small particles in the digesta processed in the small intestine, which besides increasing pancreas mass, also indicate increased activity and enzyme secretory capacity (Williams et al., 2008; Naderinejad et al., 2016; Abadi et al., 2019).
Contrastingly, feeding finely ground diets to poultry reduces the physiological activity and mechanical stimulation of the gizzard (Zentec et al., 2020). An underdeveloped gizzard function as a transit rather than a grinding organ, failing to regulate feed flow and consumption to leads to feed overconsumption, nutrient overload and shorter retention time for the nutrients in the GIT which at the end, can diminish nutrient digestion and absorption (Svihus et al., 2011; Boroojeni et al., 2016).
In mash diets, the effects of coarse particle size of ingredients on the GIT development are evident. Moreover, in hydrothermally processed feeds (pelleting, extrusion, and expansion) the use of coarse particles (larger than 1,000 or 1,500µ) can cause the fracturing of pellets, increasing the fine percentage and decreasing the broilers feed intake (Frank & Ray, 2006). As such, the recommendation is to prepare corn-soybean meal feeds for broilers, with a particle size of approximately 600 to 900µ (Amerah et al., 2007).
The study by Muramatsu et al. (2013) describes that the addition of moisture associated with a coarser particle size (1,041µ) of ingredients may increase the pellet quality. According to the authors, the 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. Thus, the investigation into how the broiler industry can benefit from highly digestible highquality pellets become ever more significant (Abadi et al., 2019; Attar et al., 2019; Massuqueto et al., 2020; Poor et al., 2021).
Therefore, according to the results noted in this study, the use of 850µ of particles in pellet feed with 1.6% of moisture in the mix, can be used as a strategy to increase the consumption of nutrients and energy, while improving productivity and revenue of broiler industry, as the reduction in grain particle size increases mill energy consumption and decreases the feed production rate (Parsons et al., 2006; Muramatsu et al., 2013).
Performance and yields of carcass and cuts
The processing methods (pellet or expanded) did not affect broiler performance. Greater weight gain (WG) at 21 and 40d was noted in birds fed diets with a 1.6% moisture addition. An improvement in feed conversion (FC) for higher moisture addition to the mixer was observed only at 21 days of age (Table 7).
The statistical breakdown of the interaction between feed processing and particle size of ingredients; indicates that broilers at 21d of age fed expanded feeds with a particle size of 650µ had higher weight gain when compared to the particle size of 850µ. However, there was no difference between the particle sizes for the animals fed pelleted feeds (Table 8).
Previous research has demonstrated that increases in pellet quality can directly improve broiler weight gain and feed conversion (Moritz et al., 2003; Hott et al., 2008; Netto et al., 2019). However, these improvements in performance normally are associated with the increase in feed intake, a fact that did not occur in this study.
For Muramatsu et al. (2014) alternative procedures such as the expansion process, increasing moisture level addition, fat inclusion restriction or post pellet liquid fat application, could be incorporated into the mill feed production process to produce physical pellets of a high quality. The authors affirmed that heat processing is the factor that most influenced pellet quality, accounting for 44% PDI variability, while the moisture addition is responsible for 16% PDI variability. However, the percentage of pellets and PDI of all diets that receive1.6% water addition in the mix presented small difference from those receiving 0.8% of moisture (with an increase of 3.8% and 1.6% in percentage of pellets and PDI respectively) (Table 2).
Thus, the probability is low that increased performance of broilers fed diets with 1.6% water addition may be result from improved pellet quality.
According to Abdollahi et al. (2018a) there are several mechanisms that underpin the advantages of pellet feeding over mash diets, but the main factor is simply the increased feed intake through facilitation of ingestion. The macro-structural characteristics of pellets (percentage of intact pellets, durability, hardness, length and diameter) also have the ability to influence feed consumption and need to be given special attention. In addition, birds fed pelleted feed decrease ration waste, through which the reduction in energy used for maintenance occurs, as observed by Moritz et al. (2002), Moritz et al. (2003), Hott et al. (2008) and Abdollahi et al. (2012).
Considering that all the experimental diets were thermally processed, the structural characteristics of good pellets in expanded and pelleted feeds, may have masked the physical pellets quality effect on feed consumption. Therefore, the increase in WG (21 and 40d) and FC (21d) in birds fed diets with 1.6% of moisture may be due to improvements in amino acid digestibility and AMEn as previously demonstrated. In terms of improvements in digestibility, other nutrients such as starch or fat can also be benefitted through the hydrothermal processes, which justifies in part increases in broilers performance (Abdollahi et al., 2013a; Bergeron et al., 2018; Lierman et al., 2019; Massuqueto et al., 2020).
However, the utilization of nutrients in broiler diets has not been clearly delineated in the literature due to the pronounced effect of pelleting on broiler performance being associated primarily at the feed intake (Abdollahi et al., 2018b). According to Boroojeni et al., (2016), nutrient availability of diets processed by heat is greatly dependent on feed formulation, chemical composition of the feed ingredients or/and the final feed, along with the presence and concentration of anti-nutritional factors, as well as the hygiene status of feed.
In addition, some researchers affirm that the conventional pelleting process in a good pellet quality is usually obtained at the expense of nutritional quality (Abdollahi et al., 2013a; Kim et al., 2018). Therefore, the need exists for research to identify and evaluate possible strategies for the manufacture of high digestible and quality pellets.
No effect was found from the processing methods on particle size and moisture level for the yields of carcass and drumstick (Table 9). However, the poultry fed expanded feeds, as well as diets with 1.6% moisture addition, presented higher breast yield.
Particle size had a significant interaction with processing methods only for abdominal fat (Table 10). Through the statistical breakdown of this interaction, we noted that birds fed expanded feed with particle size of 650µ showed a greater abdominal fat accumulation. However, for those fed pelleted feeds, the fat accumulation was higher with the use of a particle size of 850µ.
Increased breast yield was observed for broilers fed expanded feeds, as well as for diets with a higher water addition level (1.6%). Furthermore, increased abdominal fat was noted in broilers fed with expanded feeds with milling of ingredients of 650µ, as well as those pelleted with 850µ. The results obtained for carcass characteristics corroborate with those determined for the ileal digestibility of amino acids, where higher values were observed with those birds that consumed expanded diets with 650µ and pelleted with 850µ (regardless of moisture levels). The higher availability of amino acids in the feed is positively correlated with improvements in carcass yield and cuts (Lilly et al., 2011) and negatively correlated with abdominal fat (Corzo et al., 2010). However, other nutritional factors, such as energy, protein and mineral consumption may be related to an increased abdominal fat in broilers (Fouad & ElSenousey, 2014).
Researches considering influences of pellet and expanded feeds on the broiler carcass characteristics are scarce. Massuqueto et al. (2020) observed that carcass, breast, thigh, and drumstick yields also were not affected by the physical form of the diet (mash or pelleted), although the pelleted diets resulted in higher amounts of abdominal fat than the mash diets.
In contrast, Cutlip et al. (2008) assessing four steam pressure × temperature combinations (138 kPa at 82.2°C, 138 kPa at 93.3°C, 552 kPa at 82.2°C, or 552 kPa at 93.3°C) in broiler pelleted feeds, noted that variations in conditioning temperature did not affect breast yield or abdominal fat.
CONCLUSION
Compared to pelleted diets, the consumption of expanded corn-based diets does not improve the performance of broilers at 40 days of age. Therefore, to increase nutrient absorption, EMAn and broiler performance, the recommendation is to prepare pelleted diets, with the addition of 1.6% water in the mix and an 850µ particle size.
This article was originally published in Revista Brasileira de Ciência Avícola, 2022 / v.24 / n.4 / 001-010. http://dx.doi.org/10.1590/1806-9061-2020-1391. This is an Open Access article under a Creative Commons Attribution License. 
