The present study aimed to assess the benefits and disadvantages of different common feed processing methods in broiler fattening. One focus of these examinations was to test the impact of these methods on feed with regard to starch modification and nutrient digestibility. Another focus lies on the impact of variously processed feeds on chicken performance and the gastrointestinal tract of broilers.
4.1. Impact on Starch Modification and Feed Digestibility
In contrast to the previous studies of Liermann et al. [12
], only marginal effects of HTPM on starch disintegration of feed were detected with fattening pig feed in the present investigations despite similar processing conditions. Furthermore, only slight alteration could be visualized by ESM. A possible explanation could be the high content of plant oil in the present basal diet. As reported by Thomas et al. [13
] the presence of lipids can impede the gelatinization of starch or enhance the gelatinization temperature of this polysaccharide. HTPM loosened the tight connections of starch granules, whereas expander treatment seemed to be most effective.
Starch granules of ileal chyme showed characteristic pores and pin holes, which are clear indications of enzymatic attack of amylase as previously demonstrated by Fuwa et al. [14
] and Kienzle et al. [15
]. Slightly visual differences were observed between chickens fed coarsely ground and pelleted feed. Starch granules found in ileal chyme of chickens fed finely ground and pelleted feed were in a high extent attacked by digestive enzymes. The number and size of pin holes increase with increasing small intestinal starch digestibility [15
]. Small pits can be interpreted as precursors of pin holes [15
]. It has to be emphasized that the slightly lower modification of starch in ileal chyme of broilers fed coarsely ground non-compacted and finely ground and pelleted feed might be less an indication for lower digestibility of the whole dietary starch but rather than an indication for the low digestibility of these visualized starch granules themselves. Already, Fuwa et al. [14
] demonstrated by ESM that the granules of some plants that are resistant to the action of amylases show surfaces similar to intact granules after enzyme attack. Moreover, it is assumed that only less digestible starch granules reach the chosen location of chyme collection. Interestingly, no indications of starch modification by HTPM as demonstrated in Figure 1
were shown in starch granules of ileal chyme of all chicken. This aspect could be an evidence for the rapid digestion of modified starch. The aspect that the starch content in ileum did not differ markedly between feeding groups may indicate that there were no differences in starch digestibility between these groups.
Although the protein digestibility was affected in a HTPM and grinding depended manner coarsely ground non-compacted feed showed no lesser protein digestibility compared to the other feeds indicating no beneficial effects of a higher processing degree on these parameter.
Coarsely ground feed was not lesser digestible than finely ground feed which might be due to the pronounced capability to grind coarse chyme-components by the gizzard in the presence of sufficient grit [17
Pelleting resulted in slightly higher total digestive tract crude protein digestibility compared to expanded and pelleted feed. It is known that high temperatures during expander treatment can result in denaturation of proteins or destruction of feed additives such as enzymes which promote digestibility. Furthermore, interactions of high temperatures, moisture and sugar could result in the development of indigestible Maillard products [13
]. These adverse effects on feed could also be responsible for the lower protein digestibility of expanded and pelleted feed compared to solely pelleted feed. The improvement of total digestive tract crude fibre and ether extract digestibility compared to non-compacted feed caused by HTPM might be a result of higher cell wall disruption and the release of integrated fat of feed components [20
It is known that there are strong relations between nutrient digestibility or chicken performance and chyme viscosity [21
]. Furthermore, previous studies demonstrated an impact of technical feed treatment on chyme viscosity [23
]. In the present study, no effect on chyme viscosity was expected to arise from non-starch polysaccharides (NSP) by technical feed treatment because of the addition of NSP-hydrolysing enzyme (xylanase, β-glucanase) to the feed. As reported in the studies of Salih et al. [21
]; Inborr and Bedford [23
] and Dänicke et al. [25
], chyme viscosity can be markedly reduced by supplementation of the mentioned enzymes. Indeed, chyme viscosity was low in all chickens. Nevertheless, finely ground expanded and pelleted feed resulted in significantly higher viscosity of ileal chyme compared to all other feeds. Interestingly, the highest chyme viscosity of broilers fed finely ground, expanded and pelleted feed was not related with the highest feed extract viscosity. Possibly, the increase in ileal chyme viscosity is related to the damage of supplemented enzymes by the high processing temperatures. In studies by Inborr and Bedford [23
] and Vranjes et al. [19
], markedly decreased enzyme activities were shown due to pelleting or extrusion. However, this effect was not observed when feeding coarsely ground, pelleted and expanded feed although equivalent processing temperatures were used. Because of the small differences of ileal chyme viscosity between feeding groups no obvious effects on nutrient digestibility were detected due to different technically treated feeds.
4.2. Impact on Fattening and Slaughtering Performance
Broilers fed non-compacted feed were markedly inferior compared to broilers fed compacted feeds considering BWG which in turn reduced the BW on slaughtering day. The lower BW of the broilers fed non-compacted feed resulted on the one hand in lighter left breast meat and on the other hand in a lower weight of abdominal fat. Decreased BWG was clearly caused by the lower DFI of non-compacted feed fed broilers. This result seemed to be mainly based on feed form which inhibited the ability of broilers to consume similar quantities of feed by the beak as compared to their compacted-feed fed counterparts. In case of the coarsely ground non-compacted feed also a selective feed intake can be considered as total feed intake-decreasing factor. This inhibition of the ability to consume the feed due to feed form resulted in turn in a higher feed wastage and spillage by the birds which explained the increased FGR of birds fed non-compacted feeds although no differences in feed digestibility compared to compacted feed were observed.
Expanded and pelleted feed reduced also the BWG compared to pelleted feed. Although the feed form did not differ from solely pelleted feed, also in this case the reduction of the DFI is the main reason for this decreasing effect. Possibly, this effect is based on differences in pellet texture and quality. Indeed, it is known that expanding influences pellet hardness and durability inter alia by starch modification [13
], as also observed by ESM which plays a key role in the pellet quality and texture [26
]. That these characteristics can affect broiler performances was proven by Parsons et al. [28
]. Also, the differences in broiler performance between coarsely ground and finely ground expanded feeds might be explained by effects on pellet quality and texture. Interestingly, a fine grinding seemed to have adverse effects on the fattening performance when the feed was additionally expanded and pelleted compared to coarsely ground expanded and pelleted feed. In contrast, a fine grinding slightly increased the fattening performance when the feed was fed in meal form or pelleted compared to their coarsely ground counterparts. Possibly, these effects are also based on differences in feed texture and pellet quality which affected the DFI. On the other hand, it could also be a result of the differences in the protein digestibility of expander treated feeds, which was slightly reduced in finely ground expanded and pelleted feed compared to coarsely ground and pelleted feed. Furthermore, the starch content was slightly higher in coarsely ground expanded and pelleted feed compared to finely ground expanded and pelleted feed which might be also an influencing factor on BWG. Also, in the case of birds fed expanded and pelleted feed, the lower fattening performance compared with birds fed pelleted feed resulted in a slightly lower slaughtering performance.
4.3. Impact on Organ Traits
Chicken fed pelleted feed showed increases in proventriculus size but decreases in organ weight compared to other feeding groups. Therefore, it is more likely that the organ was stretched than an increase in muscular tissue occurred. It is suggested that a massive expansion of the proventriculus will limit the available space for other visceral organs in the abdominal cavity. The authors Jones and Taylor [29
] and Taylor and Jones [30
] derived a relationship between proventricular dilatation and the incidence of ascites.
There was a significant positive correlation between the DFI of broilers considering the whole fattening period and the ISC (r
= 0.307; p
= 0.002) as well as the width (r
= 0.401; p
< 0.001) and length (r
= 0.404; p
< 0.001) of the proventriculus. While the width and the length of the proventriculus were positively correlated to the DFI the weight of this organ was negatively correlated to the DFI (r
= −0.544; p
< 0.001). Therefore, a higher DFI caused a lighter but greater proventriculus. These aspects clearly indicate that the high DFI of chicken fed pelleted feed directly increase the risk for the development of proventricular dilatations in broilers. Furthermore, it is the explanation for the lower ISC and proventricular width of broilers fed coarsely ground non-compacted feed which showed lower DFI than broilers of other groups. Already Svihus and Hetland [31
] have reported that some chickens tend to overconsume when feeding on finely ground, pelleted feed. Svihus et al. [32
] recommended the inclusion of whole wheat in the diet to avoid overconsumption. However, a high DFI is not the explanation for the low ISC of broilers fed finely ground, expanded and pelleted feed because these broilers consumed similar amounts of feed compared to broilers of the other feeding groups which had higher ISC. Possibly, there is a relationship to the higher ileal chyme viscosity measured in the intestine of these broilers.
In contrast to the studies of Jones and Taylor [29
], which included whole grains in pellets, coarsely grinding before pelleting could not inhibit the dilatation of the proventriculus, although grinding appears to affect the proventricular weight. Possibly, the differences in particle size between coarsely ground and pelleted feed as well as finely ground and pelleted feed were balanced after the compacting process by secondary grinding which was described in previous studies of Wolf et al. [33
Compacted feed significantly decreased the gizzard weight compared to coarsely ground non-compacted feed. The heaviest gizzards had broilers fed coarsely ground non-compacted feed. Similar results were reported by Attia et al. [34
] and Preston et al. [24
]. The latter authors associated an increase of gizzard weight with more developed gizzard muscles. In contrast to this suggestion, no significant differences between the thicknesses of gizzard muscles of the different feeding groups were proven in the present study. Furthermore, no significant correlation existed between gizzard weight and thickness of main muscles (r
= 0.037; p
= 0.720). In studies of Wu et al. [35
] the increase in gizzard weight was also not related to a thicker tunica muscularis
. A general increase of this organ in broilers fed coarsely ground non-compacted feed compared to broilers fed other feeds, as assumed by Betscher et al. [3
], was not confirmed. In the present study only the thick, lateral main muscles were considered. Perhaps, higher gizzard weights are based on alterations in the development of the thin craniodorsal and caudoventral muscles.
Svihus et al. [36
] and Svihus [32
] hypothesized that structural feed supports the development of a heavier gizzard and gizzard functionality. These authors also suggested that this effect improved the ability of the gizzard to regulate the feed intake and in turn avoid overconsumption. Indeed, the weight of the gizzard was negatively correlated to the DFI (r
= −0.678; p
< 0.001) in the present study. Furthermore, a significant negative correlation between gizzard weights and the BWG was detected (r
= −0.575; p
In general, it should be noted that the minimum recommended particle sizes (30% particles > 1000 mm) by Svihus [32
] were fulfilled in both non-compacted feeds. However, it has to be emphasized that a secondary grinding by compacting processes might reduce feed particle size of these feed variants. Therefore, it is assumed that the structural requirements of broilers fed hydro-thermal processed feed were unmet. During slaughtering of broilers from growth experiment often bedding material was detected in the gastrointestinal tract. Apparently, chicken fed compacted feed seemed to consume more litter material than chicken fed non-compacted feed. Svihus [32
] discussed that the intake of litter material is the consequence of the compensation-demand of poultry after feeding poorly structured feed.
In accordance to studies of Taylor and Jones [30
] who reported correlations between proventricular dilatation and duodenal WL ratio significant correlations were detected between duodenal WL ratio and proventricular size (proventricular length: r
= 0.414, p
< 0.001; proventricular width: r
= 0.254, p
= 0.013). Also, the WL ratio of further intestinal segments and of the total small intestine was correlated with proventricular length and width (r
> 0.28; p
< 0.01). These data clearly indicate that alterations in the upstream digestive tract also changes morphometric characteristics in the downstream digestive tract. Possibly, these effects are based on physico-chemical changes in the intestinal content. However, the intestinal WL ratio did not correlate to chyme viscosity in the present study as previously reported by Taylor and Jones [30
]. Further, characteristics of the intestinal chyme which are known to differ after feeding different technically treated feeds are for example the particle size distribution or the pH [3
Significant treatment-associated differences between weights of heart, liver and pancreas are inconsistent and neither explainable by higher protein level of feeds nor by higher digestibility of feed or performance. Betscher et al. [3
] demonstrated in a meta-study that the inclusion of whole wheat results in a significantly increased pancreas weight. They associated the higher organ weight with an increase in pancreatic secretion. In the present study, finely ground non-compacted feed significantly increased the weight of the pancreas compared to all other feeding groups. As reported in studies of Nir et al. [39
] overfeeding resulted in higher weight of pancreas and liver, however not in higher heart weight. Moreover, these organ weights did not correspond with the DFI measured in the present study.