Location, animals and experimental design

The study was conducted at the National Center for Disciplinary Research in Physiology and Animal Improvement of the National Institute of Forestry, Agriculture and Livestock Research (Instituto Nacional de Investigaciones Forestales, Agrícolas y Pecuarias,) INIFAP). The protocol was reviewed and approved by the Animal Use Ethics Committee and complied with the Mexican Official Standard (^{NOM-062-ZOO, 1999}). A total of 421 Ross 308 chicks aged 21 to 45 days, housed on the floor (17-18 chicks/cage), with eight replicates per treatment, were fed a diet formulated with corn and soybean paste to meet the nutrient recommendations of the strain, offered free access. Bacitracin methylene disalicylate (BMD) at 11% was included as a growth-promoting antibiotic at a dose of 500 ppm equivalent to 55 ppm of active ingredient; and salinomycin at 12% at a dose of 500 ppm equivalent to 60 ppm of active ingredient for the prevention of coccidiosis. Treatments were as follows: 1) broilers received drinking water directly from a storage tank connected to bell drinkers, 2) 80 % tank water was mixed with 20 % raw WL (RWL) and 3) 80 % tank water was mixed with 20 % pasteurized WL (PWL). The WL was obtained from vermicompost prepared from sheep manure. In treatment 2, before mixing the WL with the drinking water, it was filtered using cotton cloth (diaper cloth). In treatment 3, the WL was filtered and heated at 65 ºC for 1 h before mixing with water. In treatments 2 and 3, the water-WL mixture was put into 20-L plastic jugs and connected to bell troughs. The contents of HA, FA and total humic acids (THA) were 0.30, 0.33 and 0.66 %, respectively.

Recording of production variables

The body weight of the chicks was recorded at the beginning and end of the study and the daily weight gain (DWG, g/d) was estimated as the difference between the final weight (42 d) and the initial weight, divided by the experiment day number. Feed offered and rejected was also monitored to estimate the daily feed intake (DFI, g/d). Feed conversion (FC) was calculated by dividing the DFI by DWG. Daily water consumption (DWCg, ml/d) was quantified and water consumption/feed consumption was calculated. Mortality was recorded daily. At the end of the experiment and after a 12 h fast, five broilers were slaughtered from each pen and carcass and breast weights were recorded. Carcass and breast weights were expressed in grams (g) and the yield was calculated as a percentage (%), dividing by body weight.

Sampling and laboratory analysis

At the beginning of the experiment, before putting the chickens in pens, the floor was cleaned by placing 11 kg of new sawdust and a sample was taken; at the end, the complete litter was weighed and a sample was taken from each pen. At the beginning of the experiment, three groups of chickens from the same flock of chickens used in the experiment were slaughtered; at the end, three chickens were slaughtered per pen with an average weight similar to the average weight per pen. Feed samples were also taken every week.

Litter samples were freeze-dried and ground using a 2 mm sieve. Slaughtered chickens were plucked and feather and body weights were recorded and analyzed separately. The whole body was ground in a meat grinder and a representative sample was taken. The body and feathers were freeze-dried separately and ground. Feed samples were ground using a 2 mm sieve. In litter, broiler carcass, feather and feed samples, dry matter (DM), ash (A) and nitrogen (N) were determined. All laboratory determinations were carried out following recommendations ^{AOAC (2019)}.

Calculations of the chemical composition of chickens and litter

The following formulas were used to estimate total DM, A and N of litter and broilers at the beginning and end of the experiment in each pen:

DM, A and N in litters ,kg =% DM, A and N x weigth of litters ;;

DM, A and N in chickens, kg*=(% de DM, A and N of chickens x weight of chikens)+(%de DM, A and N of feathers x weight of feathers,kg);

*The sum of the N of the body and feathers was made in amounts equivalent to the weight of the body and feathers at the beginning and at the end of the experiment.

Total feed consuption,kg=DFI x number of chikens per pen x24 days;

Total consuption of DM, kg=% DM of feed x total feed consumption;

N total intake,kg =DM total intake x % N in feed;

Calculations of gain of chemical components of chickens and litter

The following formulas were used to calculate the gain of chemical components in broilers per pen:

DM gain in chikens, kg = DM total of chikens at the end - DM total of chikens at the beginning of the experiment;

C again in chikens, kg = Total N of chicks at the end- Total N of chiks at the beginning of the experiment-;

N gain in chikens, kg=Total N of chicks at the end-Total N of chicks at the beginning of the experiment-;

The following formulas were used to calculate the gain of chemical components in the litter per pen:

DM gain in the litter ,kg =total DM in the litter at the end - total DM in the litter at the beginning of the experiment;

C gain in litter,kg= Total C in litter at the end - Total C in litter at the beginning of the experiment;

N gain in litter , kg =Total N in litter at the end - Total N in litter at the beginning of the experiment;

Calculations of nitrogen retention and losses

The following formulas were used to obtain the N retention in kg and percentage in broilers and litter per pen:

N retention in chikens ,kg = total N in chicks at 45 days - total N in chiks at 21 days;;

N retention, kg = total N in litter at 45 days - total in litter at 21 dyas;

N retentioen in broilers and litter, kg =total N in broilers an litter at 45 days - total N broilers and litter at 21 days;

The amount of N lost in total, per chicken and per kg of chicken produced was calculated as follows:

Total N lost,kg =Total N consumption, kg - N retenction in broilers and litter , kg;

Statistical analysis

The results were subjected to analysis of variance under a completely randomized design, using the General Linear Models procedures of the SAS statistical package. Prior to the analysis of variance, assumption verification was carried out. The GDP and CDA variables were transformed using the multiplicative inverse and all variables expressed as percentages were transformed to arcsine to comply with the assumption of normality. Statistical differences between means were analyzed using the least significant difference methodology at P < 0.05.

Production and carcass variables

Table 1 shows the productive variables, carcass and breast weight and yield. Initial weight, final weight, feed consumption, weight gain, feed conversion, water consumption, and water consumption/feed consumption and mortality were similar among the three treatments evaluated. These findings do not agree with those of a study in chickens added with RWL in the drinking water (^{Gomez-Rosales and Angeles, 2015}), obtaining benefits in final weight, weight gain and feed conversion; but using diets devoid of antibiotic growth promoter and anticoccidial products. One of present work purposes was to clarify whether the beneficial effects of WL on broiler growth could be maintained despite the presence of BMD and salinomycin; however, productivity results show that the benefits of WL previously observed using antibiotic-free diets were lost. WL pasteurization was carried out with the intention of eliminating the naturally present microorganisms and nullifying any possible growth-promoting effect that the beneficial flora might exert; as has been suggested in tests of plant growth added with WL, as a source of HS (^{Canellas et al., 2015}; ^{Olivares et al., 2015}). Another important difference was that in the previous work a WL from a vermicompost, made with pig and sheep manure containing 0.47, 0.14 and 0.61 % of HA, FA and AHT, respectively, was used (^{Gomez-Rosales and Angeles, 2015}); and in the present work a WL obtained from a vermicompost, made with sheep manure containing 0.30, 0.33 and 0.66 % of HA, FA and THA, respectively, was used.

Breast yield was higher (P < 0.05) in broilers that drank RWL with respect to those that drank only water and RWL (Table 1). Carcass and breast weight and yield were similar between treatments. In a previous study, higher carcass yield was obtained from chickens supplemented with HS extracted from worm composting (^{Domínguez-Negrete et al., 2019}). Higher carcass weight and carcass yield have also been reported in chickens supplemented with increasing amounts of HS in feed and drinking water compared to the control group not supplemented with HS (^{Ozturk et al., 2010}; ^{Ozturk et al., 2012}). In a previous report it was found that HS subjected to heating for more than 40 min retain their detergent properties, but present a lower electron transfer capacity of labile chemical groups that are lost during heating (^{Visser, 1985}); probably by reducing the reactions associated with electron exchange between HS and different acceptors, and highlighting only the surfactant effect, increasing the permeability of the membranes; causing the higher breast yield with PWL, compared to RWL.

Composition and gain of chemical components of broiler chickens

The chemical composition of the slaughtered chickens at the beginning and end of the experiment and the gain of chemical components are presented in Table 2. The initial chemical composition was not analyzed statistically because only a representative sample of the chickens was taken. There were no statistical differences in the chemical composition or chemical component gain of the chickens at the end. The results are not consistent with improvements in protein use efficiency and N and A retention in chickens supplemented with HS (^{Gomez-Rosales and Angeles, 2015}; ^{Disetlhe et al., 2017}), through the excreta collection method.

It has been reported that HS are able to regulate N availability for plants in the soil, due to their adsorptive properties, binding directly with ammonia or stimulating the activity of disintegrating bacteria that facilitate N uptake through the roots (^{Canellas et al., 2015}; ^{Olivares et al., 2015}). It was expected that this same effect could be carried out in the chicken intestine, reducing ammonia levels released in the digestive tract, improving health, production and nutrient retention (^{Qaisrani et al., 2015}; ^{Lemme et al., 2019}). Probably the presence of BMD and salinomycin counteracted the effects of HS observed in previous work.

Table 3 shows the chemical composition of the litter at the beginning and end of the experiment and the gain of chemical components. The chemical components at the beginning were not statistically analyzed because only a representative sample of the litter was taken. There were no statistical differences in the chemical composition or chemical component gain of the litter at the end. Previous studies have reported that HS can inhibit urease activity present in soil bacteria (^{Zhang et al., 2013}) and ammonia emissions from excreta of pigs and chickens supplemented with HS (^{Ji et al., 2006}; ^{Maguey-Gonzalez et al., 2018}).

In the present study, it was expected that HS would bind to ammonia inside the intestine and in the litter, reducing N losses by volatilization, and consequently, higher N retention in the litter at the end. Probably the presence of the antibiotics counteracted the effects of HS in the litter.

Retention and losses of nitrogen from broilers and litter

Table 4 shows the results of nitrogen balance and loss during the experiment. None of the variables analyzed showed differences between treatments. These findings do not coincide with the higher breast yield, which could suggest that there was greater efficiency in the use of N for muscle protein synthesis, nor with the higher N retention reported in a previous work in chickens supplemented with RWL (^{Gomez-Rosales and Angeles, 2015}). The results are not consistent with improvements in protein use efficiency and N and A retention in chickens supplemented with HS (^{Gomez-Rosales and Angeles, 2015}; ^{Disetlhe et al., 2017}).

The lack of differences in N balance coincides with the lack of differences in broiler productivity. N retention efficiency in broilers, litter and broilers + litter was 41, 28 and 69%, respectively. N retention in broilers using the culling method was lower than reported nitrogen retention (of 61-80 %) in broilers used in nitrogen balance studies subjected to restricted feeding (^{Gómez et al., 2012}; ^{Gomez-Rosales and Angeles, 2015}); but is more in line with nitrogen retention (of 29-43 %) in free-fed broilers (^{Gomez and Angeles, 2011}).

Theoretically, the rest of the N not recovered in the broilers was excreted to the litter, i.e., 59% of the N excreted by the broilers was expected to be recovered; however, only 28% of N was recovered in the litter, so it is assumed that the missing N was lost as ammonia to the environment, i.e., 31% of the N consumed and excreted was released from the litter by volatilization inside the house during the 24-day period that the broilers remained in the experiment. In litter, N losses in the form of ammonia are due to microbial mineralization of urea and uric acid accounting for up to 80 % of the total nitrogen excreted (^{Zhang et al., 2015}).

N losses per chicken and per kg of chicken indicate that an average of 1.42 and 0.65 g of N per day were volatilized, respectively; considering that 0.216 g of H is required for the formation of 1 g of ammonia from 1 g of N; estimating that per chicken and per kg of chicken produced, 1.72 and 0.78 g of ammonia were generated. The rate of ammonia formation from litter depends mainly on ambient temperature and humidity, and the rate of accumulation inside the house depends on house size, number of chickens and degree of ventilation (^{Feng-Xiang et al., 2012}; ^{Wei et al., 2015}). However, it should be taken into account that when ammonia is released from the litter, it is first inspired by the chick before it is distributed in the house environment. If the house is adequately ventilated, the harmful effects of ammonia on the chicken may be slight; but if the ventilation is not adequate, the chicken will remain exposed to the toxicity of the gas for a longer period of time, because it will be constantly breathing it in. Therefore, it is necessary to continue searching for feed and environmental management alternatives to mitigate the emissions of polluting gases such as ammonia inside the houses, in order to reduce their detrimental impacts on the productivity and health of the birds.