Investigation on adaptations of broiler chickens to high dietary free amino acid levels in nitrogen utilisation and plasma amino acid concentrations

ABSTRACT 1. A reduction in crude protein (CP) in feed for broiler chickens necessitates elevated free amino acid (AA) levels to meet the requirement of each AA. This study investigated adaptations following a change to diets with increasing free AA concentrations and possible reasons for the limitation caused by the inclusion of more free AA. 2. Male Ross 308 broiler hatchlings received a starter diet (164 g CP/kg containing 80 g/kg soy protein isolate (SPI)) until d 7. From d 7–22, birds received a diet almost identical to the starter diet or two other diets, where 50% or 100% of digestible AA in SPI were substituted with a free AA mixture. Birds were allocated to metabolism units located in the same barn to determine performance (n = 7 units) and blood traits (n = 14 birds). Total excreta collection was performed on d 7–8, 8–9, 9–10, 11–12, 14–15 and 21–22. Blood samples were collected on d 7, 8, 9, 11, 14 and 21. 3. Average daily weight gain (ADG) and average daily feed intake (ADFI) was unaffected at 50% AA substitution but decreased at 100% AA substitution on d 7–22 (p ≤ 0.001). The 100% substitution led to a decline in ADG and ADFI consistently on all days (p ≤ 0.037) except on d 11–12. A 50% AA substitution resulted in lower ADFI on d 7–8 and 14–15 (p ≤ 0.032). Nitrogen utilisation efficiency (NUE) was on a level of ~ 0.74 and was only affected by treatment up to d 11–12 (p ≤ 0.008). Concentrations of 10, 9, 8, 10 and 4 plasma free AA were affected on d 8, 9, 11, 14 and 21, respectively (p ≤ 0.037). 4. Following a change to diets containing high levels of free AA, NUE and free AA concentrations in the circulation became more balanced within 3 to 7 d. The results suggested that peptide-bound and free AA did not cause different NUE, particularly 3 and 7 d after the diet change.


Introduction
The use of free amino acids (AA) enables a reduction in crude protein (CP) in feed for non-ruminants, including poultry, while achieving target AA concentrations.Using free AA means less dependence on protein-rich materials, like soybean meal, which often is criticised due to environmental challenges for production, long-distance transportation and higher nutrient import to areas of farm animal husbandry.Furthermore, reducing CP in diets (considering that provided by free AA) while fulfilling AA requirements leads to less nitrogen (N) excretion from animals and increases the efficiency of converting feed protein in animalbased food.
Free and peptide-bound AA have different absorption rates, which results in free AA reaching the systemic circulation more rapidly (Denbow 2015;Krehbiel and Mattews 2003).A different appearance of free and peptide-bound AA may result in an pattern present in the systemic circulation that does not allow for maximum protein accretion, even though the dietary AA pattern has been formulated to be balanced.An imbalanced AA pattern in the systemic circulation was presumed to result in increased AA oxidation or intensified degradation of body protein to deliver AA required for synthesis of other proteins (Wu 2013).These mechanisms would contribute to increased protein turnover and, hence, lead to lower N utilisation efficiency (NUE).
The free AA inclusion needed for adjusting concentrations increases with decreasing CP in the feed of broiler chickens.A lower growth associated with increasing free AA inclusion in the diet has often been explained by an upper limit of dietary free AA levels in consequence of different absorption rates of free and peptide-bound AA.However, many of these studies have primarily focused on the utilisation of peptide-bound and free AA as a result of reduced dietary CP (Attia et al. 2022;Bregendahl et al. 2002;Corzo et al. 2005;Namroud et al. 2008).In consequence, interpretations regarding the upper limit of dietary free AA in such studies may also be due to an unrecognised deficiency of AA, which was not considered relevant.
The authors are aware of two studies in which peptide-bound AA were consistently substituted with free AA in poultry.In one study, a noticeable decrease in average daily weight gain (ADG) and gain:feed ratio (G:F) was found when peptide-bound AA were substituted with a mixture of 18 free proteinogenic AA, but NUE was not affected (Siegert et al. 2016).It was inferred that decreased ADFI was the primary reason for reduced growth in birds fed high concentrations of free AA.In a recent study, an upper limit of free AA was determined between 54 and 71 g free AA/kg of feed when digestible peptide-bound AA in soy protein isolate (SPI) were incrementally substituted with a mixture of all 20 proteinogenic AA (Ibrahim et al. 2023).Above this limit, growth and N accretion decreased, with reduced ADFI being considered by the authors as the most likely cause.
The impact of diets high in free AA after change to such diets may have caused reduced growth and N accretion in the study reported by Ibrahim et al. (2023).The authors found similar treatment effects on N accretion and other traits related to N metabolism at 4-7 and 11-14 d after change to the experimental diets.This consistency suggested there was no evidence for relevant adaptions onto high proportions of free AA between these periods.However, impacts of high proportions of free AA within a few days after a diet change are unknown to date.This time period may be of particular relevance because studies on peptide and AA transporters showed that the transporters can adapt to changes in diet composition within one week (Gilbert et al. 2010;Morales et al. 2015).
Therefore, the objective of this study was to determine metabolic adaptations after change to diets with increasing free AA concentrations.The performance of broiler chicken was determined along with excreta and blood sampling at short time intervals after change to diets with high free AA levels to determine N accretion and NUE.Free AA in blood plasma were analysed to determine changes in AA patterns.It was hypothesised that adaptations occurred within the first 7 d after change to diets with high free AA levels.

Experimental setup
All birds received the same starter diet with 164 g CP/kg and containing 80 g SPI/kg until the age of 7 d.From d 7 onwards, birds received a diet almost identical to the starter diet (0FAA) or two other diets where 50% (50FAA) or 100% (100FAA) of the precaecally digestible AA sourced from SPI were substituted with a mix of free AA containing all 20 proteinogenic AA.Using almost identical diets, we investigated levels of free AA inclusion without and with impact on growth and N accretion in a previous study (Ibrahim et al. 2023).In the present study, each diet was tested in seven metabolism units containing 10 birds each and seven metabolism units of 15 birds each to determine performance as well as N metabolism traits and to obtain blood samples, respectively.This separation ensured that blood sampling did not interfere with performance and the N metabolism measurements.All metabolism units were located in the same barn.Units were assigned to dietary treatments according to a randomised complete block design which was optimised using the OPTEX procedure of SAS (version 9.4, SAS Institute, Cary, U.S.A.).
of the recommendations by the Gesellschaft für Ernährungsphysiologie (1999).Non-essential AA concentrations were formulated based on Hofmann et al. (2019Hofmann et al. ( , 2020)).Free AA were added to adjust the AA concentrations in the diets.All diets were calculated to contain the same precaecally digestible AA concentrations based on data from a preliminary digestibility experiment where diets based on the identical basal mixture were investigated (Ibrahim et al. 2023).Digestible asparagine (Asn) + aspartic acid (Asp) and glutamine (Gln) + glutamic acid (Glu) provided by the SPI were substituted with 50/50 mixes of Asp/Asn and Glu/Gln.The proportion of free AA in the diets was 19, 53 and 86 g/kg in 0FAA, 50FAA, and 100FAA, respectively.The SPI (Euroduna Feed Ingredients GmbH, Barmstedt, Germany) was used as a source of peptidebound AA because it has a low hydrolysation degree of up to 3% according to the supplier and almost completely consisted of AA (994.4 g/kg dry matter (DM)).Therefore, the substitution of SPI with free AA had little influence on concentrations of other nutrients.Mass differences up to ~12 g/kg among the diets were balanced using maize starch.
The starter diet and the other diets were pelleted through a 2 mm and 3 mm die, respectively, without using steam.Technical challenges of producing the 3 mm pellets were overcome by adding soybean oil on top of the readily prepared mixtures, resulting in slightly lower proportions of other ingredients, so that AA concentrations were 103% of the recommendations.Concentrations of all nutrients reported herein are based on a standardised DM of 88%, unless otherwise stated.Results of AA analysis of the diets (Table 2) confirmed the calculated values.

Birds and housing
A total of 525 male Ross 308 broiler hatchlings were obtained from a commercial hatchery (Brüterei Süd ZN der BWE-Brüterei Weser-Ems GmbH & Co. KG) and placed in two floor pens (2 m × 6 m) on dedusted wood shavings.On d 6, birds were transferred to metabolism units in the same barn, ensuring an equal mean bird weight in every unit.Allocation to the metabolism units was carried out a day prior to the introduction of the experimental diets to reduce reallocation stress at the time of diet change.The metabolism units sized 1 m × 1 m × 1 m and 2 m × 1 m × 1 m to house the 10 and 15 birds for performance and blood measurements, respectively.Water and diets were provided ad libitum until the end of the experiment on d 22. Lighting was continuous during the first 3 d after placement, followed by 18 h light and 6 h dark cycle until the end of the experiment.The temperature was set at 34°C for the first 3 d and was gradually decreased to 20°C by d 22.

Experimental procedures
The animals and feed in the metabolic units used to study performance were weighed on days 7, 8, 9, 10, 11, 12, 14, 15, 21 and 22 on a unit basis to record the ADG, ADFI and G:F.Feed DM was determined on these days.The birds were inspected at least twice daily.Dead birds were removed and weighed, and the feed intake of the remaining birds in the respective unit was recorded.Spilled feed pellets were collected daily from trays located underneath each unit, dried and weighed to correct ADFI on a DM basis.Total collection of were substituted by 50% (50FAA), or 100% (100FAA) with free amino acids using a mixture of all 20 proteinogenic amino acids.Digestible asparagine+aspartic acid and glutamine+glutamic acid in SPI of the diet with 0% amino acid substitution were substituted with 50/50 mixes of asparagine/aspartic acid and glutamine/glutamic acid.
2 All diets were calculated to contain the same precaecally digestible AA concentration based on data determined a preliminary digestibility experiment where diets based on the identical basal mixture were investigated (Ibrahim et al. 2023).
excreta was carried out twice daily at 12 h intervals on d 7-8, 8-9, 9-10, 11-12, 14-15 and 21-22.Excreta were immediately frozen at − 20°C after each collection.The excreta samples were thawed at 3°C, then homogenised and DM, ammonia and N concentration were determined.Parts of the homogenised sample was freeze-dried for uric acid (UA) analysis.Trunk blood samples were taken from two birds from each of the other metabolic units on d 8, 9, 11, 14 and 21 by throat cut, after anaesthesia by blunt blow on the head, according to directive 2010/63/EU.Blood samples were collected in lithium heparin tubes and centrifuged for 10 min at 1500 × g and 4°C.Birds selected for blood sampling were marked individually on d 6 and were chosen based on a preplanned schedule to prevent biased selection.On d 7, prior to the diet change, two birds per unit were sampled.Two birds per unit were sampled on day 8, 9, 11, 14 and 21, resulting in 14 sampled birds per treatment per day.

Chemical analyses
Diets were ground through a 0.5 mm sieve in a centrifugal mill (ZM 200;Retsch GmbH,Germany) for analyses of crude ash, crude fat, crude fibre, and starch.A vibrating disc mill (Pulverisette 9, Fritsch GmbH, Germany) was used for the analyses of sodium, potassium, chloride and AA concentrations.The official methods for nutrient analysis in Germany (Verband Deutscher Landwirtschaftlicher Untersuchungs-und Forschungsanstalten 2007) were used for the determination of DM, N, crude fat, crude fibre, crude ash, sodium, potassium and chloride.The AA concentrations in the diets were analysed after oxidation and acid hydrolysis (Siegert et al. 2015).Tryptophan was analysed separately using an HPLC (Fatufe et al. 2005).Analysis of free AA in the blood plasma was carried out without preceding sample hydrolysis.Plasma insulin was analysed using a commercial chicken insulin ELISA Kit (MyBioSource.com, San Diego, U.S.A.).Excreta N using Kjeldahl digestion and excreta DM were determined in duplicate and triplicate, respectively.The analysis of excreta ammonia (NH 3 ; herein, ammonia includes both NH 3 and NH 4 + ) and UA concentrations followed the method described by Hofmann et al. (2019).

Calculations and statistical analysis
The ADG, ADFI, and G:F were calculated from d 7 to 22 and on d 7-8, 8-9, 9-10, 11-12, 14-15 and 21-22 of the experiment.Dead birds were considered in the calculation of ADG and ADFI.The ADFI on a standardised DM of 88%, ADG, G:F, N accretion, and NUE were determined on a metabolism unit basis.N accretion and NUE were calculated using following equations: All traits were statistically analysed by one-way analysis of variance (ANOVA) using the MIXED procedure of SAS.The metabolism unit was considered the experimental unit for all traits except for blood data, where individual birds were the experimental unit.The statistical model was: Where y ij was the dependent trait, α the overall mean, trt i the fixed effect of treatment i, block j the random effect of block j, and e ij the residual error.A random block effect was included if model accuracy, as indicated by the Akaike information, was improved.Homogeneity of variance and normal distribution was confirmed for every trait.Differences between treatments were compared using Tukey tests.Statistical significance was set at p ≤ 0.050.

Performance of broiler chickens
The bird weight on d 7 ranged between 139 and 142 g/ bird and was not significantly different among treatments (p = 0.155).The survival rate during the experimental phase was 99% and was not related to any treatment (five birds died belonging to three different treatments).The ADG, ADFI and G:F from d 7 to 22 was unaffected at 50% AA substitution but was lower at 100% AA substitution compared to 0% AA substitution (Table 3; p ≤ 0.001).Similarly, ADG and ADFI on d 7-8, 8-9, 9-10, 14-15 and 21-22 decreased at 100% AA substitution (Figure 1; p ≤ 0.037) while there was no significant influence at d 11-12.The 50% AA substitution did not impact ADG and ADFI, except for lower ADFI on d 7-8 and 14-15 (p ≤ 0.032).The G:F ratio was reduced at 50% AA substitution on d 14-15 and the 100% AA substitution led to a further reduction in G:F on d 8-9 and 14-15 (p ≤ 0.003).

Nitrogen accretion and nitrogen utilisation efficiency
The N accretion on d 7-8 was reduced at 50% and 100% AA substitution compared to 0% AA substitution (Figure 2; p ≤ 0.001), while N accretion of 50% AA substitution was lower compared to the other treatments on day 8-9 (p ≤ 0.006).The N accretion on d 9-10 decreased with increasing AA substitution (p ≤ 0.001).There was no influence of 50% AA substitution on N accretion on days 11-12, 14-15 and 21-22; however, N accretion on these days was lower for 100% AA substitution (p ≤ 0.031).The NUE was on a level of 0.75 for 0% AA substitution on d 7-8 and followed a similar response pattern among the treatments as N accretion on d 7-8, 8-9 and 9-10.There was no treatment effect on NUE on d 11-12, 14-15 and 21-22, with an average NUE of ~ 0.74.N concentration in gained body weight, which was influenced by treatment on d 7-8, 8-9, 9-10 and 14-15 (Figure 2; p ≤ 0.017), but remained unaffected by treatment on d 11-12 and 21-22 (p ≥ 0.066).

Excretion of nitrogenous compounds
The NH 3 -N excretion rose with increasing AA substitution across all sampling days (Figure 3; p ≤ 0.001), except when comparing 50% and 100% AA substitution on d 21-22.However, NH 3 -N excretion relative body weight increased in line with AA substitution for all sampling days (p ≤ 0.001).The UA-N excretion on d 7-8 increased with increasing AA substitution (Figure 3; p ≤ 0.001).There was no difference in UA-N excretion on d 8-9, 9-10 and 11-12 between 50% and 100% AA substitution, but UA-N excretion on these days was lower for 0% substitution (p ≤ 0.001).The UA-N excretion relative to body weight was higher in the diets with 50% and 100% AA substitution compared to the diet without AA substitution (p ≤ 0.012), except for on d 14-15.The NH 3 -N/(NH 3 -N+UA-N) ratio was highest for 100% AA substitution on all sampling days (p ≤ 0.015), except for d 9-10.For d 11-12 and d 14-15, this ratio increased with higher AA substitution (Figure 3; p ≤ 0.001) while there was no difference in this ratio between 0% and 50% AA substitution on the other sampling days.The NH 3 -N/(NH 3 -N+UA-N) ratio on d 7-8, 8-9 and 21-22 increased only by 100% AA substitution compared to 0% and 50% AA substitution (p ≤ 0.020).

Free amino acid and insulin concentrations in blood plasma
Concentrations of 10, 9, 8, 10 and four free proteinogenic AA in blood plasma were significantly different among treatments on d 8, 9, 11, 14 and 21, respectively (Figure 4; p ≤ 0.037).In general, the order of free AA concentrations Table 3. Effects of incremental substitution of digestible amino acids from 80 g soy protein isolate (SPI)/kg in one diet (0FAA) with 50% (50FAA) or 100% (100FAA) of free amino acids on average daily weight gain, average daily feed intake, and the gain: feed ratio of broiler chickens from d 7 to 22 post-hatch (n = 7 units of 10 birds each) and on insulin in blood plasma on d 7, 8, 9, 11, 14, and 21 after change to the experimental treatments on day 7 (n = 14 individual birds).Values are presented in supplementary Table 1.
was 100% > 50% > 0% AA substitution.Lysine was the only AA in blood plasma that increased with higher AA substitution on all sampling days (p ≤ 0.001).For most AA, the biggest numerical differences in concentrations in the blood plasma among treatments were observed on d 8 and 9.These differences among treatments diminished by d 14 and 21.The treatments did not influence insulin concentrations in blood plasma (Table 3; p ≥ 0.091).

Discussions
Most indications for adaptations to the presence of high levels of free AA in diets seemed to occur within 3 to 7 d after diet change, thus overall confirming the hypothesis.This was suggested by altered NUE within the initial 3 d and increasing concentrations of most of the free AA in blood plasma with higher AA substitution levels within the first 7 d after diet change.Subsequently, NUE was unaffected by treatment and plasma free AA concentrations narrowed among treatments.Peptide and transporters in the small intestine adapting to the presence of free AA (Gilbert et al. 2010;Morales et al. 2015) may have contributed to that, but such transporters were not measured herein.The absence of treatment effects on NUE after 3 d after the diet change may imply that AA absorption by the peptide and AA transporters was no limiting factor for protein accretion.
The birds responded to increasing AA substitution with reduced ADFI immediately after the diet change, which impacted on G:F and ADG.On the three days following diet change, substituting peptide-bound with free AA affected performance, N accretion and NUE differently and without any discernible pattern.The high G:F on d 7-8 and 8-9 was mainly due to reduced ADFI on these days, while the ADG, N accretion and NUE partly were a result of ADFI from the preceding day.The reduced growth in the first days after the diet change to 50% AA substitution was compensated for by the birds later on in the experiment because there were no performance differences between 0% and 50% AA substitution over the whole 14 d.Contrary to that, 100% AA substitution resulted in lower performance compared to the other treatments throughout the experimental period.These results allign with a preliminary study using nearly identical diets (Ibrahim et al. 2023).A compensation of the reduced ADFI by higher ADFI of 50% compared to 0% AA substitution in the middle of the experimental period could explain the similar growth of the birds fed those treatments.However, effects of seamless diet change cannot explain the low ADFI of all treatments on the second day after the diet change.
Blood plasma free AA concentrations apparently are unsuitable as indicators for impacts of high free AA levels in the diets on NUE after three days of diet change.While there was no treatment effect on NUE from three days after diet change onwards, concentrations of most free AA in the blood plasma were higher the more free AA were included in the diet up to seven days after diet change.In addition, NUE was not affected after three days of diet change.Hence, the lower N accretion at 100% AA substitution on d 7-22 was probably caused by reduced N intake and not by decreased utilisation of the rapidly absorbed free AA.Absorption rates of peptide-bound and free AA may differ (Denbow 2015;Krehbiel and Mattews 2003), but such an effect had no impact on N accretion.Nevertheless, the treatment effects on plasma free AA concentrations were possibly caused by different absorption rates of peptide-bound and free AA, which may have contributed to the varying NUE within the first three days after change to diets with 50% and 100% AA substitution.Likewise, NUE was not affected, despite the lower growth of birds fed a free AA mixture instead of SPI in another study on broiler chickens (Siegert et al. 2016).This supported the interpretation that the lower growth was mainly due to a lower ADFI and was not caused by a different utilisation of peptide-bound and rapidly absorbed free AA.
A recent study on pigs described increasing plasma insulin concentrations when peptide-bound AA were substituted with free AA (Eugenio et al. 2023).The authors reported that higher absorption rates led to peaks in plasma AA concentration which were accompanied by higher plasma insulin concentrations.Higher insulin was a potential cause for lower ADFI in the study of Eugenio et al. (2023) because  3.
increased insulin concentrations in the circulation were found to reduce ADFI in poultry (Shiraishi et al. 2008) but we are not aware of similar findings for pigs.Possibly, lower ADFI in consequence of higher plasma insulin concentrations may represent a strategy to avoid very high AA concentrations in circulation in both poultry and pigs.However, no treatment effects on plasma insulin concentrations were determined in the present study.One explanation for the diverging results of the study on pigs (Eugenio et al. 2023) and the present study, is that the broiler chickens were fed ad libitum while the pigs were fed specific meals.The more constant nutrient intake when fed ad libitum may mute peaks of plasma AA concentrations caused by the rapid absorption of free AA.This would make higher insulin secretion to avoid very high AA concentrations in the circulation less necessary.Another consideration is that the ad  alanine, 2211; arginine, 1217; asparagine, 1347; aspartic acid, 1255; cysteine, 246; glutamic acid, 720; glutamine, 5029; glycine, 4196; histidine, 599; isoleucine, 1235; leucine, 1609; lysine, 2312; methionine, 724; phenylalanine, 787; proline, 1581; serine, 2246; threonine, 2107, tyrosine, 591; valine, 2279.libitum feed provision prior to bleeding may have contributed to the absence of treatment effects on plasma insulin concentrations in the present study.Insulin concentration is affected by ad libitum feeding or the refeeding method in broiler chickens (Bigot et al. 2003;Richards and Mcmurtry 2008) and pigs (Reynolds et al. 2010;Sanz Fernandez et al. 2015).In addition, ADFI is regulated by several physiological mechanisms in addition to insulin (Bungo et al. 2011).The different taste of the same AA present in free form or as peptides observed in humans (Linde et al. 2009) may have influenced ADFI, but the consequences of such taste differences in broiler chicken are unknown.The ADFI may have been influenced by a change in acidbase balance in the birds fed the diet with 100% AA substitution in the first days after diet change.The daily NH 3 -N excretion raised with increasing AA substitution in all observation periods after diet change.Differences between treatments had the same response pattern when NH 3 -N excretion was related to body weight, which suggested that treatment effects on NH 3 -N excretion were not a consequence of the different animal weight among treatments.Parts of raising NH 3 -N excretion with increasing AA substitution can be attributed to a higher proportion of urinary N excreted in form of NH 3 -N.This was suggested by the higher NH 3 -N/(NH 3 -N+UA-N) ratios with increasing AA substitution, given that urinary N consists mostly of NH 3 -N and UA-N (Goldstein and Skadhauge 2000).Increasing proportions of urinary N excreted as NH 3 -N may represent an adaptive response to excrete acids to maintain the acid-base balance (Hamm and Simon 1987).Consistently, a challenged acidbase balance in the diets with high AA substitution was accompanied by increased urinary NH 3 -N excretion in a preliminary experiment (Ibrahim et al. 2023).The provision of HCl with increasing AA substitution due to higher dietary concentrations of L-lysine•HCl may also have contributed to a shift in the acidbase balance (Patience 1990).However, blood traits indicating whether the acid-base balance of the birds was challenged in the present study were not reported herein.
In conclusion, the present study provided no evidence of different NUE after three days post-diet change and declining differences in free AA concentrations in the blood plasma seven days after diet change.The reduced growth of birds fed 100% AA substitution could be attributed to a decline in ADFI in the first three days after diet change.Most of the adaptations associated with high free AA levels in the diet with 50% and 100% substitution took place within the first three to seven days after diet change.That was indicated by changes in ADFI, N accretion, and NUE in the first three days as well as in plasma free AA concentrations within the first seven days.Thereafter, NUE and plasma free AA concentrations became more balanced among treatments.The increase in NH 3 -N excretion in the diets with 100% compared to 0% AA substitution might represent an adaptative response to excrete acid and could have been caused by a shift in the acid-base balance.

3Figure 1 .
Figure 1.Effects of incremental substitution of digestible amino acids from 80 g soy protein isolate (SPI)/kg in one diet (0FAA) with 50% (50FAA) or 100% (100FAA) of free amino acids on average daily weight gain (ADG; panel a), average daily feed intake (ADFI; panel b), and the gain:feed ratio (G:F; panel c) of broiler chickens from d 7 to 22 post-hatch (n = 7 units of 10 birds each).Error bars indicate the pooled standard error.Treatments without a common letter within a time period differ significantly (p < 0.050).Values are presented in supplementary Table 1.

Figure 2 .
Figure 2. Effects of incremental substitution of digestible amino acids from 80 g soy protein isolate (SPI)/kg in one diet (0FAA) with 50% (50FAA) or 100% (100FAA) of free amino acids on nitrogen (N) accretion (panel a), N utilisation efficiency (panel b), N accretion relative to average daily weight gain (ADG; panel c) of broiler chickens determined on selected days after change to the experimental treatments on d 7 (n = 7 units of 10 birds each).Error bars indicate the pooled standard error.Treatments without a common letter within a time period differ significantly (p < 0.050).Values are presented in supplementary Table 2.

Figure 3 .
Figure 3. Effects of incremental substitution of digestible amino acids from 80 g soy protein isolate (SPI)/kg in one diet (0FAA) with 50% (50FAA) or 100% (100FAA) of free amino acids on excretion of ammonia-nitrogen (NH 3 -N; panel a) and uric acid-nitrogen (UA-N; panel c) and their ratio to each other (panel e) and their relation to body weight (panels b and d) of broiler chickens determined on selected days after change to the experimental treatments on day d (n = 7 units of 10 birds each).Error bars indicate the pooled standard error.Treatments without a common letter within a time period differ significantly (p < 0.050).Values are presented in supplementary Table3.

Table 2 .
Analysed nutrient composition of the experimental diets and digestible amino acids in all diets (g/kg on 88% dry matter basis unless otherwise stated).
10FAA (no substitution of amino acids from soy protein isolate; SPI): basal mix + SPI; digestible amino acids in SPI contained in 0% amino acid substitution (0FAA) Digestible asparagine+aspartic acid and glutamine+glutamic acid in SPI of the 0FAA diet were substituted with 50/50 mixes of asparagine/aspartic acid and glutamine/glutamic acid.