Effects of semi-open water and non-water enrichment on welfare, production, behavior, and microbial exposure of grow-out Pekin ducks

JM Schober; N Wilson; MM Seyoum; JM Lyte; MM Bergman; EM Oluwagbenga; GS Fraley

doi:10.1016/j.psj.2025.105477

. 2025 Jun 25;104(9):105477. doi: 10.1016/j.psj.2025.105477

Effects of semi-open water and non-water enrichment on welfare, production, behavior, and microbial exposure of grow-out Pekin ducks

JM Schober ^a, N Wilson ^a, MM Seyoum ^c, JM Lyte ^b, MM Bergman ^a, EM Oluwagbenga ^a, GS Fraley ^a,^⁎

PMCID: PMC12272575 PMID: 40618487

Abstract

Environmental enrichment plays an important role in the welfare, behavior, and health of commercially raised Pekin ducks. We assessed the effects of three enrichment treatments and a control: Nipple line (CON), whiffle ball (EED), preening cup (PC), and Pekino (PEK) on duck welfare, behavior and bacterial exposure. 525 grow-out Pekin ducks were housed in 4 pens/4 rooms with one of the four enrichment types (N = 4 pens/treatment). Body weights and body condition scores of 10 ducks/pen and FCR, ammonia level, and litter moisture % were recorded weekly. On 16 and 44 days of age, 3 ducks/pen were euthanized and their organs were weighed and whole brains collected. Behavior data were collected using scan sampling with video being recorded for 72 continuous hours for 3 weeks after enrichment placement. Weekly samples were also collected for viable bacterial counts, and 16S rRNA gene sequencing at baseline and 6 h after baseliner. Body morphometrics, FCR, ammonia levels, litter moisture % and viable bacteria were analyzed by 2-way ANOVA with repeated measures. Body condition scores were analyzed with PROC LOGISTIC (SAS 9.4). GLIMMIX procedure (SAS 9.4) was used to analyze behavior. Bacterial communities were characterized using 16S rRNA gene sequencing, and functional potential was predicted using PICRUSt2. PEK and PC ducks were largest in weight (p < 0.0001) and better nostril scores (p = 0.0005) but had dirtier feathers (p < 0.0001), worse litter conditions (p < 0.0001) and more viable bacteria in their water sources (p < 0.0001), while the PC and EED ducks had worse feather quality (p = 0.0021). Alpha and beta diversity metrics revealed that microbiota composition was significantly (p < 0.05) dependent on environmental enrichment type. Likewise, functional pathway analyses revealed distinct (p < 0.05) metabolic capacities, including aerobic respiration and amino acid biosynthesis, between microbiotas of each respective environmental enrichment niche. Our study suggests that semi-open water sources and EEDs may lead to an increase in feather pecking and a decrease in feather quality with a low number of ducks per enrichment as well as an increase in bacteria load and litter moisture.

Keywords: Pekin ducks, Microbiome, Behavior, Enrichment, Open-water

Introduction

Open water sources for Pekin ducks are a current topic of interest for welfare scientists and producers. Pekin ducks are waterfowl, so they have a natural instinct to interact with open water. Certification organizations are beginning to require implementation of open water into commercial duck barns in order for that stakeholder to be certified (e.g. American Humane Society; A Greener World, Global Animal Partnership). The standard housing systems for a Pekin duck commercial barn consists of either solid floors with litter or with raised plastic flooring over a pit or a combination of the two. Pin-metered (nipple) water lines are located over the pit (Chen et al., 2021). Water lines are arranged so that a maximum of 8 adult ducks per pin is available for adequate hydration and to allow for social behaviors and preening to occur (Rice et al., 2014; Schenk et al., 2016; Chen et al., 2021). The use of nipple lines has been criticized for not allowing ducks to perform their natural behaviors such as dabbling, head-dipping, bathing, or swimming (for review see Rodenburg et al., 2005). Thus, numerous researchers have investigated the effects of implementing open-water sources for Pekin ducks.

Open water sources are not only thought of as just providing the ducks with access to water for hydration, but also as a form of environmental enrichment. Critical elements of effective environmental enrichment are (1) assessing the animal’s natural history, individual history, and exhibit restraints and (2) providing species-appropriate opportunities, including ducks specifically dunking their head and wet preening (Mellen and MacPhee, 2001). Jones and Dawkins (2010) suggested that implementing forms of open water enrichment (bell drinker and trough) benefited eye, feather cleanliness, and food pad condition, but ducks housed with the trough had the highest incidence of broken feathers on their breast. O’Driscoll and Broom (2011) suggested that ducks housed with bell drinkers were dirtier than those housed with a bath, and those housed with a trough were intermediate. Ducks with bell drinkers had worse gait scores than those with baths or troughs (O’Driscoll & Broom, 2011). Ducks were less likely to have dirty nostrils when provided with more open water sources and were less likely to have blocked nostrils in the trough and bath treatments compared to those housed with bell drinkers (O’Driscoll & Broom, 2011). However, numerous other studies have shown that open water sources may not benefit commercial duck housing.

Lowman et al. (2016) saw that troughs caused litter to have a higher moisture percentage, elevated coliform and general bacterial growth, and subsequently increased incidents of foot pad lesions compared to ducks housed with nipple lines. That study looked at breeder Pekin ducks, and they also saw that the ducks housed with nipple lines produced significantly more eggs and used less water than ducks housed with a trough (Lowman et al., 2016). Similarly, Schenk et al. (2016) found that ducks housed with troughs showed worse scores for eyes, nostrils, feather quality, feather cleanliness, and foot pads. The troughs also had higher iron, nitrites, pH, and bacterial growth, with the bacterial growth having higher E.coli, coliforms, and Staphylococcus. Further, the water samples from the nipple lines showed no bacterial growth in culture-based assays. Ducks with troughs also showed a greater percent mortality at all ages compared to ducks with nipple lines (Schenk et al., 2016). Recently, our lab (Schober et al., 2023), investigated a form of semi-open water, a preening cup, and saw that there were no differences in body condition scores, body weight or uropygial gland size in ducks housed with preening cups compared to ducks housed with just nipple lines. However, we found that ducks housed with preening cups performed wet preening behaviors more often than the ducks housed with nipple lines, but that did not affect their feather quality/cleanliness or the size of their uropygial gland (Schober et al., 2023). Thus, the relationship between open, or semi-open, water sources, expression of natural behaviors, and duck welfare is still elusive.

Environmental enrichment has also shown to decrease abnormal behaviors, such as feather pecking and feather picking in various poultry species. Feather pecking (aggressive pecking) is pecking that is delivered with considerable force to conspecifics and gives the impression it is intended to hurt (Savory, 1995). For chickens, it is thought that feather pecking is linked to foraging behavior or possibly related to dust bathing behaviors. When a substrate is not functional in relation to foraging behavior or dustbathing, birds redirect pecking to other substrates, such as feathers (Blokhuis and Wiepkema, 1998; Jones et al., 2004; van Staaveren and Harlander, 2020). Feather picking is the act of automutilation, and small amounts of feather picking are viewed as beneficial, in the form of preening or removal of juvenile down but can become harmful under certain conditions (Savory, 1995; Blokhuis and Wiepkema, 1998; Colton and Fraley, 2014).

Colton and Fraley (2014) looked at the effects of a physical, non-water source environmental enrichment device (EED), whiffle-style balls threaded with 4 zip-ties in commercial duck barns. They found that the ducks housed with these EEDs showed decreased feather pecking (on conspecifics) and feather picking (auto-mutilation). Ducks housed in a commercial setting with EEDs also showed better feather quality and cleanliness scores compared to the ducks without EEDs. Thus, non-water-based enrichment may also improve housing systems similar to what some have reported for open water systems.

A more comprehensive understanding of how open water, semi-open water, or physical enrichment without water can affect Pekin duck health, production, and behavior is necessary. In this study, we set out to compare water and non-water-based enrichments in grow-out Pekin ducks. Specifically, we compared two types of semi-open water, preening cups and Pekinos, to non-water-based enrichment for ducks and ducks housed with nipple lines alone. We hypothesized that ducks housed with the Pekinos, or preening cups would have no significant differences in body morphometrics, preening gland size, or behaviors when compared to ducks with just a nipple line. We further hypothesized that litter moisture and ammonia from the pens with the semi-open water sources would increase when compared to the pens with the EED and control. Lastly, we set out to determine if the semi-open water sources would increase bacterial exposure to the ducks compared to non-water-based enrichment or nipple lines alone. Our study suggests that semi-open water sources and EEDs may lead to an increase in feather pecking and a decrease in feather quality when the number of ducks per enrichment was low. Our study also suggests that water enrichments increase bacteria load and litter moisture but may also increase body weight. Further research is needed to address whether implementation of different forms of enrichment reduces abnormal behaviors.

Materials and methods

Animals

All procedures were approved by Purdue's Institutional Animal Care and Use Committee (PACUC # 0923002434). A total of 525 grow-out Pekin ducks were obtained from Culver Duck, Inc. on the day-of-hatch and brought to Purdue University Animal Sciences Research and Education Center (ASREC) farm. They were housed in 5 m² floor pens with 0.17m² per duck, approximately 30 ducks per pen. They were brooded per industry standards and fed industry standard starter feed for the first 10 d of life and grower feed for the remainder of life per industry recommendations (Chen et al., 2021). All feed and water were provided ad libitum. Ducks were placed in single rooms under an 18:6 light cycle, with lights turning on at 0300 h and off at 2100 h, at a temperature maintained at 20-22°C. Brooder temperatures were maintained as recommended, 26°C on the first day of life and drops 1°C a day for 2 weeks (Chen et al., 2021). Water nipple lines were placed over a pit covered with raised plastic flooring, and the remaining area of the pens were covered with pine shavings and added to or replaced as necessary at the same time for all pens. The ducks were grown-out until 43 days of age. We utilized 4 rooms with 4 pens per room with the treatments randomly placed throughout the 4 rooms (N = 4 pens/treatment).

Enrichment placement

The preening cups (PC) and Pekinos (PEK), placed on day 16, were also over the pit as per industry practices (Fraley, conversations with stakeholders) and were placed at the appropriate height to prevent the ducks from climbing in but allowing them to drink from and dunk their heads. Each PC and PEK were connected to the main water supply and were self-filling.

The whiffle balls with zip-ties, also called environmental enrichment devices (EED; Colton and Fraley, 2014) were first placed on day 16 at the same time as the preening cups and the Pekinos. The EED was placed 2 per pen, for 6 days straight, and then removed for 3 days to limit habituation (Colton and Fraley, 2014). This cycle was repeated until the end of the study. Four pens had a nipple line along with a PEK (placed on d 16: “Pekino;” Big Dutchman, Holland, MI USA), 4 pens had a nipple line along with a PC (“IP bell” from IMPEX Watering Solutions, Gainesville, GA USA), 4 pens had a nipple line and balls with zip-ties attached (EED; Colton and Fraley, 2014), and 4 control pens had only a nipple line (CON; 5 ducks/nipple).

Morphometrics and body condition scores

Body weights and body condition scores were assessed on a weekly basis on 15 ducks per pen using a rubric previously published (Karcher et al., 2013; Fraley et al., 2013; Schober et al., 2023). Foot pad quality, eye and nostril health, feather cleanliness and quality were all assessed. Feed intake was measured weekly and feed conversion ratio (FCR) was calculated for each week. On day 16, prior to enrichment placement, 3 ducks per pen were euthanized using FatalPlus (pentobarbital, 390 mg/ml/kg) and their spleens, Bursa of Fabricius, livers, and uropygial glands were weighed (final N = 12/treatment). Three ducks per pen were also euthanized on day 44, twenty-eight days post-enrichment placement using FatalPlus and their spleens, Bursa of Fabricius, livers, and uropygial glands were again weighed (final N = 12/treatment). Whole brains were also removed from 3 ducks/pen/treatment/time point in under 5 min and stored on dry ice until stored at −80°C.

Litter and ammonia

Litter (2 equidistant samples per pen) was collected from all pens each week starting 24 h after placement (d 17). The litter was weighed when collected and then placed in an oven at 105°C for 24 h to dry. Once dry, the litter was weighed again, and the moisture percentage was calculated with the formula (wet litter – dry litter)/wet litter *100. Ammonia readings were taken weekly, with one reading being done at duck height above the shavings and one reading being done at duck height above the pit in each pen (N = 16 readings/treatment/week).

Behavior analyses

WYZE® Cam V3 cameras were placed overhead, across each pen, and each camera was able to visualize its entire, respective, pen. Videos were recorded for 72 continuous hours on days 17-19 (week 1), 27-29 (week 2), and 38-40 (week 3) days of age. Note that the enrichment was placed on 16 days of age. The cameras were turned on the day after the PEK, PC, and EED were placed, and recorded on days the EEDs were in the pens. Videos were analyzed by scan sampling and behavior events were recorded using Excel. We analyzed the first 15 min in each pen for AM (0330 h, 0430 h, 0530 h) and PM (1430 h, 1530 h, and 1630 h). These times were chosen based upon preliminary examination of the videos that indicated these to be peak hours of activity and are similar to time frames reported previously for behavioral analyses of ducks (Rice et al., 2014; Schober et al., 2023). We analyzed each minute during the 15-min interval by counting how many ducks were performing each behavior and then divided that number by the number of ducks in the pen, resulting in a proportion for each behavior.

The analyzed behaviors included dry preening, wet preening, preening conspecifics (also known as allopreening), drinking from nipple line, drinking from PC/PEK, dunking head in PC/PEK, interacting with EEDs, eating, feather picking, feather pecking, standing, and laying down. The ethogram is provided in Table 1 and was adapted from Liste et al. (2013) and from Schober et al. (2023). Wet preening is defined as using the bill to apply water directly to the feathers while nibbling at or stroking the feathers while applying water directly to the feathers or after tossing water over the body (Jones and Dawkins, 2010; Waitt et al., 2009). We adapted a quadrant around water sources used to describe wet preening from Schober et al. (2023). We are not able to observe water in the bills of the ducks on the videos to confirm wet preening, so the quadrant was used to define the ducks’ behavior as wet preening if they were preening inside of the quadrant immediately following an interaction with water, and dry preening if they were preening outside the quadrant. The quadrant was a one-duck length around the nipple lines and PC, and for this study also around the PEK (Schober et al., 2023).

Table 1.

Ethogram used for the behavior data. Adapted from Schober et al. (2023).

Behavior	Description
Feather Pecking	Directed at another duck, involves removal or attempted removal of another’s feather. Differentiable from preening in that the receiving party will be distressed and make efforts to move away from the attacker.
Drinking from preening cup/Pekino	Dipping beak in water and immediately pulling head back line to attain water from the preening cup/Pekino
Eating	Involves the active consumption of food from a designated food source.
Feather Picking	Involves rapid pulling and removal of own feathers. Feathers from any area can be targeted, but its most often the back and wing feathers that are pulled.
Dunking head in preening cup/Pekino	Rapidly ducking head in water with no clear purpose. Dunking may precede wet preening. Differs from preening/drinking in that the duck does not pull their head back.
Laying down	The duck will be in ventral recumbency, either awake or asleep, but not engaged in another activity. Wings will typically be more relaxed than those of standing ducks and may be stretched out on the side of the duck.
Drinking from nipple line	Standing or sitting directly under the nipple line with their neck extended, reaching one nipple and actively attaining water from it.
Allopreening	Grooming of another duck which may or may not be reciprocated by the other party. The duck performing the service may groom any area of the receiving duck, however the aforementioned areas are the most typical.
Wet Preening	Manipulation of the feathers with the aid of water. The duck will target many areas of its body, with the most common being the chest, back, and wings.
Dry Preening	Manipulation of feathers without the use of water. The most common areas selected for preening are the chest, back and wings.
Standing	Characterized by the duck passively resting in one place on both feet without being engaged in another activity. The wings are folded to the body and the head is most often held in a relaxed posture. Occasionally, the duck will tuck its’ head under its wing while standing.
Interacting with EEDs	Characterized by the duck using their bills to peck at the EEDs in their pens.

Open in a new tab

Brain microdissections

The brains were removed from the −80°C and hemisected with the right half being used for mass spectrometry. These same halves were then further microdissected into 4 target brain areas: caudal mesencephalon (CM), rostral mesencephalon (RM), diencephalon, (DI), and telencephalon (T). These brain areas were selected based on previous studies and a published stereotaxic atlas of the avian brain (Kang et al., 2009, 2020; Kuenzel and Masson, 1988). Details for these procedures for ducks can be found in our lab’s previous studies (Bergman et al., 2024a, 2024b, 2024c).

Mass spectrometry

Brain areas were prepped for mass spectrometry following a previously published protocol (Kim et al., 2014; Bergman et al., 2024a, Bergman et al., 2024b, Bergman et al., 2024c). The brain areas were placed into their own tubes along with 1.0 mm glass beads. 10ul of internal standard, 10ul of acetonitrile (ACN) per mg of brain tissue, and .01 % ascorbic acid were added to tubes as well (Kim et al., 2014). The internal standards contain heavy hydrogen isotopes at a concentration of 10ng/µl for 3,4-dihydroxyphenylacetic acid (DOPAC), homovanillic acid (HVA), 3-methoxytyramine (3-MT), dopamine (DA), 5-hydroxyindoleacetic acid (5HIAA), and serotonin (5-HT). Samples were vortexed for 10 min and then centrifuged for 10 min at 12,000 rpm. Supernatant was then removed and put into a microcentrifuge tube inside of a speedvac and left to evaporate for 24 h. Samples were then stored at −80C until resuspension in 75ul 9:1 HLIC A:B (Kim et al., 2014). Samples were vortexed to homogenize dried components for 30 min and centrifuged at 12,000 rpm for 10 min. Supernatant was collected and run in Agilent Triple Quadrupole Mass Spec (QQQ). Results were collected in neurotransmitter concentrations of ng/mg of tissue. These concentrations were plugged into the neurotransmitter turnover equations for DA and 5HT: (DOPAC+HVA+3MT) / DA and 5HIAA / 5HT (Ahmed and Azmnat, 2017).

Viable bacteria collection and analysis

1.5 mL of water samples were collected from the Pekinos, preening cups, and nipple lines once a week starting on day 18. The collection times were baseline (immediately after dumping out and cleaning the PC and PEK) and 6 h after baseline. The EED were swabbed on the day of placement and when they were removed 6 days later. The swabs were placed in deionized water (DI) in a 1.5 mL centrifuge tube. A commercially available kit for viable bacteria counts was utilized (QUANTOM Viable Cell Staining Kit used with the QUANTOM M50 Cell Counting Slides [Logos Biosystems, Annadale, VA]). The QUANTOM Tx Microbial Cell Counter (Logos Biosystems, Annadale, VA) was used to count the number of viable bacteria in each sample (Logos Biosystems, Annadale, VA). The protocol used is described in Table 2.

Table 2.

Protocol used for viable bacteria on the QUANTOM Tx Micobial Cell Counter (Logos Biosystems, Annadale, VA).

QUANTOM Tx Microbial Cell Counter Protocol
Dilution Factor	2
Min. Fluorescent object size	0.3µm
Max. Fluorescent object size	20µm
Size gating	0.3 ∼ 20 µm
Roundness	50 %
Declustering level	7
Detection sensitivity	4
Focusing method	Autofocus
LED level	5

Open in a new tab

Sample collection for 16S rRNA gene sequencing

We used sterile 50 mL tubes to collect 50 mL of water from the PEK, PC, and nipple lines once a week starting 24 h after placement (d 17) with a baseline and a 6 h after baseline collection. The PEK and PC were dumped out right before the baseline collection and were not dumped at 6 h. The PEK and PC samples were taken from just below the surface of the water. The tubes were gently pressed against a nipple in the nipple line to expel water into the tube. The EEDs were swabbed with #4 qualitative filter paper, placed in 50 mL of sterile PBS solution and vortexed before the EEDs were placed in the pens and 6 days later, when they were taken out of the pens. All the samples were spun down at ∼2000 G for 20 min in a room temperature centrifuge. The supernatant was decanted, and the remaining pellet was resuspended in 1000 µl of sterile PBS. The samples were then vortexed and 1000 µl of the mixture was placed in 1.5 ml sterile centrifuge tubes and stored at −80°C until DNA isolation and downstream 16S rRNA gene sequencing at the Purdue Genomics Center (Purdue University).

DNA extraction and sequencing

DNA was extracted using the ‘DNeasy PowerSoil Pro Kit’, as per the manufacturer’s instructions (cat #: 47014, QIAGEN, Germantown, MD) with a bead beating treatment of 2.5 min at 5.5 m s-1 (FastPreop-24; MP Biomedicals, Solon, OH). DNA extracts were quantified using the Quant-iT™ PicoGreen™ dsDNA Assay Kit (Thermo Fisher Scientific, Inc., Waltham, MA) and measured with a FilterMax F5 micro-plate reader (Molecular Devices, San Jose, CA). Bacterial community composition was determined by sequencing the V4 region of the 16S rRNA gene amplified from DNA extracts by polymerase chain reaction (PCR) using Q5 Hotstart Taq (New England Biolabs; Ipswich, MA) and dual-indexed barcoded 515f/806r primers as described by (Kozich et al., 2013). PCR was performed with 2 ng of DNA template and pooled, purified and normalized to a standard concentration using the SequalPrep normalization kit (Invitrogen, CA, USA). Amplicon libraries were multiplexed and sequenced using the AVITI sequencer (2 × 150 paired-end) at the Purdue Genomics Core (Purdue University, West Lafayette, IN, USA).

Bioinformatic analyses of 16S rRNA gene sequencing data

Raw 16S rRNA gene sequencing data were processed using QIIME 2 (version 2023.7). Paired-end sequences were imported into QIIME 2, and quality control was performed using the DADA2 plugin, which filtered low-quality reads, removed chimeras, and generated amplicon sequence variants (ASVs). The outputs included a feature table and representative sequences for downstream analyses.

Alpha diversity metrics, including Shannon diversity, Faith’s phylogenetic diversity, Observed Features, and Evenness, were calculated using the diversity core-metrics-phylogenetic workflow in QIIME 2. Rarefaction was performed at a sequencing depth of 45,000 sequences per sample to normalize the data across samples.

Beta diversity was assessed using Bray-Curtis dissimilarity, Jaccard distance, unweighted UniFrac, and weighted UniFrac metrics. Principal coordinates analysis (PCoA) plots were generated to visualize community-level differences between groups.

Taxonomic classification of ASVs was performed using a Naive Bayes classifier trained on the SILVA 138 database, specific to the amplified V4 region. The taxonomic composition was summarized at different taxonomic levels (phylum, class, order, and family) using bar plots to illustrate relative abundance patterns across groups.

Functional profiling of microbial communities using PICRUSt2 analysis

Functional profiling of microbial communities was performed using the Phylogenetic Investigation of Communities by Reconstruction of Unobserved States 2 (PICRUSt2) pipeline, which predicts the functional potential of microbiomes from 16S rRNA amplicon data. Briefly, a representative sequence table and associated ASV table were exported from QIIME 2 and used as inputs for PICRUSt2. Representative sequences were aligned to a reference phylogeny using HMMER, and a phylogenetic placement of the ASVs was performed using EPA-ng. Functional predictions were inferred by mapping ASVs to known genomes in the reference database, with outputs including predicted gene family abundances (KEGG Orthologs) and metabolic pathways (MetaCyc).

To ensure high-quality predictions, ASVs with poor alignment to the reference database (alignment proportion <80 %) or exceeding a nearest sequenced taxon index (NSTI) threshold of 2.0 were excluded. The predicted abundances of KEGG Orthologs and pathways were used for downstream statistical analysis and visualization to identify potential functional differences across microbial communities.

Statistical analyses

Morphometrics, environment and FCR

Data were analyzed by two-way ANOVA with repeated measures (SAS 9.4) and the pen was the statistical unit for all data unless otherwise stated, and pairwise comparisons were examined using Tukey’s test. A statistical significance of p ≤ 0.05 considered significant.

Body condition scores

The PROC LOGISTIC (SAS 9.4) with the Firth bias correction for quasi-complete separation and odds ratios were calculated with a statistical significance of p ≤ 0.05 considered significant. Results are reported with corresponding confidence intervals (CI).

Behavior analyses

The GLIMMIX procedure (SAS 9.4) was used for all analyses to examine treatment differences, week differences, and time differences in the proportion of ducks performing dry preening, wet preening, preening others, eating, drinking from nipple line, standing, and laying down. Treatment differences were analyzed for interacting with enrichment (EED interacting with whiffle ball and PC/PEK dunking head in semi-open water). Week and time differences were analyzed for EED interacting with enrichment. Dunking head in semi-open water was analyzed for just the PC and PEK group, as the CON and EED group did not have them. Pen and time nested in week was included as a random effect and time was included as a repeated measure. Pairwise comparisons were examined using Tukey’s test. Data were examined for normality using studentized residual plots, with a statistical significance of p ≤ 0.05 considered significant. Results are described with differences between least square means ± SEM.

16S rRNA analyses

All statistical analyses were conducted in R (R Core Team). Alpha diversity metrics, including Shannon diversity, Faith’s PD, observed features, and evenness, were compared across treatments using the Kruskal-Wallis test, with p-values adjusted for multiple comparisons using the Benjamini-Hochberg (BH) correction. Beta diversity was assessed using Bray-Curtis dissimilarity, and PCoA was performed to visualize microbial community clustering across enrichment treatments and time points. Differences in beta diversity were evaluated using permutational multivariate analysis of variance (PERMANOVA) with 999 permutations. Taxonomic composition and functional pathway abundances were analyzed using the Kruskal-Wallis test, followed by pairwise comparisons where applicable, with statistical significance set at p < 0.05.

Results

Weekly body weights

Weekly body weights were significantly influenced by treatment (F_3,69=10.42, p < 0.0001), days of age (F_5,69=10638, p < 0.0001), and their interaction (F_15,69=5.88, p < 0.0001). For the treatment effect, ducks in the PC group weighed significantly more than ducks in the CON (p < 0.001) and EED (p = 0.0272) groups. Ducks in the PEK group weighed significantly more than ducks in the CON group (p = 0.0003). For the interaction effect, on 35 days of age ducks in the PC group weighed significantly more than ducks in the CON group (P < 0.001). Ducks in the PEK group weighed significantly more than ducks in the CON group (p < 0.0001). On 42 days of age, ducks in the PC group weighed significantly more than ducks in the CON (p < 0.0001) and EED (p = 0.0112) groups. Ducks in the PEK group weighed significantly more than ducks in the CON (p < 0.0001) and EED (p = 0.0116) groups (Fig. 1).

Fig 1 — Weekly body weights. Ducks were weighed on a weekly basis. Note that enrichment devices were placed on 16 days of age. Ducks in the PEK and PC groups were overall weighed more than the ducks in the EED and CON groups (P < 0.0001). On 35 days of age ducks in the PC group weighed more than ducks in the CON group, and ducks in the PEK group weighed more than ducks in the CON group. On 42 days of age, ducks in the PC and PEK groups weighed more than ducks in the CON and EED groups. * =p < 0.05, *** =p < 0.0001.

FCR and feed intake

No significant differences were found among treatments for FCR or feed intake (Fig. 2).

Dissection

The body weight of the ducks was significantly influenced by treatment (F_3,85=4.04, p = 0.0098), days of age (F_1,85=4559.22, p < 0.0001), and their interaction (F_3,85=7.16, p = 0.0002). For the treatment effect, ducks in the PC group weighed significantly more than ducks in the EED group (p = 0.0068). For the interaction effect, on 44 days of age ducks in the PC group weighed significantly more than the ducks in the CON group (p = 0.0297) and ducks in the EED group (p < 0.0001), and the ducks in the PEK group weighed significantly more than the ducks in the EED group (p = 0.0094; Fig. 3A). No significant differences were observed for percentage of body weight for spleen or bursa among treatments (Fig. 3C and D). The liver size was significantly influenced by treatment (F_3,85=4.97, p = 0.0031), where ducks in the EED group had significantly larger relative liver size than ducks in the PC group (p = 0.0017; Fig. 3B). The uropygial gland size was significantly influenced by treatment (F_3,85=4.66, p = 0.0046) where ducks in the PEK group had significantly larger uropygial glands compared to ducks in the CON group (p = 0.045), ducks in the EED group (p = 0.0051), and ducks in the PC group (p = 0.0253; Fig. 3E).

Fig 3 — Dissection data. Ducks were euthanized on 16 days of age, prior to enrichment placement and on 44 days of age. Their body weights were recorded (A), as well as their liver (B), spleen (C), Bursa of Fabricius (D), and uropygial glands (E) were weighed. We observed significant differences for the liver and the uropygial gland. Letters indicate statistical significance, p < 0.05.

Body condition scores

Feather quality

Feather quality was significantly influenced by age (Wald X²=29.3220, p < 0.0001), treatment (Wald X²=8.5157, p = 0.0365) and their interaction (Wald X²=30.7684, p = 0.0021). Data are shown in Table 3.

Table 3.

Feather quality odds ratio.

Feather Quality Probabilities
Days of Age	Contrast	Estimate	Confidence Limits		P Value
	CON vs EED	1.8575	0.9486	3.6372	0.0709
	CON vs PC	1.7396	0.8880	3.4078	0.1066
	CON vs PEK	0.7997	0.3997	1.5998	0.5275
	EED vs PC	0.9366	0.4844	1.8106	0.8455
	EED vs PEK	0.4305	0.2177	0.8516	0.0155
	PC vs PEK	0.4597	0.2323	0.9096	0.0256
27	CON vs EED	4.2093	2.0609	8.5974	<.0001
	CON vs PC	3.5937	1.7682	7.3039	0.0004
	CON vs PEK	1.0000	0.4807	2.0803	1.0000
	EED vs PC	0.8538	0.4320	1.6874	0.6492
	EED vs PEK	0.2376	0.1163	0.4852	<.0001
	PC vs PEK	0.2783	0.1369	0.5655	0.0004
35	CON vs EED	0.5655	0.2910	1.0989	0.0926
	CON vs PC	1.1058	0.5715	2.1398	0.7652
	CON vs PEK	0.1338	0.0630	0.2841	<.0001
	EED vs PC	1.9555	1.0046	3.8064	0.0484
	EED vs PEK	0.2366	0.1123	0.4985	0.0002
	PC vs PEK	0.1210	0.0568	0.2576	<.0001
42	CON vs EED	1.9805	1.0086	3.8888	0.0472
	CON vs PC	1.8406	0.9377	3.6131	0.0762
	CON vs PEK	0.7861	0.3937	1.5696	0.4952
	EED vs PC	0.9294	0.4798	1.8003	0.8281
	EED vs PEK	0.3969	0.2000	0.7878	0.0082
	PC vs PEK	0.4271	0.2153	0.8472	0.0149

Open in a new tab

Age. Ducks at 14 days were 36.9541 times more likely to receive a score of 0 for feather quality when compared to ducks at 27 days (CI: 4.7913-285.0, p = 0.0005), 20.2594 times more likely than ducks at 35 days (CI: 2.5800-159.1, p = 0.0042) and 41.8296 times more likely than ducks at 42 days (CI: 5.4356-321.9, p = 0.0003). Ducks at 20 days were 12.0379 times more likely to receive a score of 0 for feather quality when compared to ducks at 27 days (CI: 3.3693-43.0096, p = 0.0001), 6.5996 times more likely than ducks at 35 days (CI: 1.7950-24.2642, p = 0.0045) and 13.6261 times more likely than ducks at 42 days (CI: 3.8275-48.5094, p < 0.0001).

Treatment. Ducks in the EED group were 0.4305 times less likely to receive a score of 0 for feather quality when compared to ducks in the PEK group (CI: 0.2177-0.8516, p = 0.0155). Ducks in the PC group were 0.4597 times less likely to receive a score of 0 for feather quality when compared to ducks in the PEK group (CI: 0.2323-0.9096, p = 0.0256).

Days of age x Treatment Interaction. At 27 days (Wald X²=27.8104, p < 0.0001), ducks in the CON group were 4.2093 times more likely to receive a score of 0 for feather quality when compared to the ducks in the EED group (CI: 2.0609-8.5974, p < 0.0001), and 3.597 times more likely than the ducks in the PC group (CI: 1.7682-7.3039, p = 0.0004). Ducks in the EED group were 0.2376 times less likely to receive a score of 0 for feather quality when compared to the ducks in the PEK group (CI: 0.1163-0.4852, p < 0.0001). Ducks in the PC group were 0.2783 times less likely to receive a score of 0 for feather quality when compared to the ducks in the PEK group (CI: 0.1369-0.5655, p = 0.0004).

At 35 days (Wald X²=35.6303, p < 0.0001), ducks in the CON group were 0.1338 times less likely to receive a score of 0 for feather quality than ducks in the PEK group (CI: 0.0630-0.2841, p < 0.0001). Ducks in the EED group were 1.9555 times more likely to receive a score of 0 for feather quality when compared to ducks in the PC group (CI: 1.0046-3.8064, p = 0.0484), but 0.2366 times less likely when compared to ducks in the PEK group (CI: 0.1123-0.4985, p = 0.0002). Ducks in the PC group were 0.1210 times less likely to receive a score of 0 for feather quality when compared to ducks in the PEK group (CI: 0.0568-0.2576, p < 0.0001).

At 42 days (Wald X²=10.1056, p = 0.0177), ducks in the CON group were 1.9805 times more likely to receive a score of 0 for feather quality when compared to ducks in the EED group (CI: 1.0086-3.8888, p = 0.0472). Ducks in the EED group were 0.3969 times less likely to receive a score of 0 for feather quality when compared to ducks in the PEK group (CI: 0.2000-0.7878, p = 0.0082). Ducks in the PC group were 0.4271 times less likely to receive a score of 0 for feather quality when compared to ducks in the PEK group (CI: 0.2153-0.8472, p = 0.0149).

Feather cleanliness

Feather cleanliness was significantly influenced by age (Wald X²=165.4089, p < 0.0001), treatment (Wald X²=141.3352, p < 0.0001) and their interaction (Wald X²=92.3067, p < 0.0001). Data are shown in Table 4.

Table 4.

Feather cleanliness odds ratio.

Feather Cleanliness Probabilities
Days of Age	Contrast	Estimate	Confidence Limits		P Value
	CON vs EED	1.2020	0.5974	2.4187	0.6060
	CON vs PC	28.0700	12.9526	60.8317	<.0001
	CON vs PEK	72.3324	30.8351	169.7000	<.0001
	EED vs PC	23.3521	10.8209	50.3950	<.0001
	EED vs PEK	60.1749	25.7538	140.6000	<.0001
	PC vs PEK	2.5769	1.1589	5.7297	0.0202
20	CON vs EED	1.3830	0.5536	3.4548	0.4876
	CON vs PC	1.3830	0.5536	3.4548	0.4876
	CON vs PEK	2.5000	1.0521	5.9405	0.0380
	EED vs PC	1.0000	0.4195	2.3836	1.0000
	EED vs PEK	1.8077	0.7996	4.0867	0.1548
	PC vs PEK	1.8077	0.7996	4.0867	0.1548
27	CON vs EED	1.6625	0.7342	3.7648	0.2229
	CON vs PC	1.4227	0.6209	3.2597	0.4046
	CON vs PEK	3.0653	1.3881	6.7691	0.0056
	EED vs PC	0.8557	0.3927	1.8648	0.6950
	EED vs PEK	1.8438	0.8804	3.8611	0.1047
	PC vs PEK	2.1546	1.0154	4.5719	0.0455
35	CON vs EED	1.5239	0.7157	3.2446	0.2746
	CON vs PC	2.4952	1.1864	5.2481	0.0159
	CON vs PEK	9.4385	4.2686	20.8699	<.0001
	EED vs PC	1.6374	0.8011	3.3471	0.1764
	EED vs PEK	6.1938	2.8853	13.2961	<.0001
	PC vs PEK	3.7826	1.7975	7.9602	0.0005
42	CON vs EED	1.1820	0.5926	2.3575	0.6351
	CON vs PC	17.1353	7.6062	38.6021	<.0001
	CON vs PEK	42.9601	17.4899	105.5000	<.0001
	EED vs PC	14.4974	6.5005	32.3319	<.0001
	EED vs PEK	36.3467	14.9406	88.4221	<.0001
	PC vs PEK	2.5071	1.1263	5.5809	0.0244

Open in a new tab

Age. Ducks at 14 days were 28.5387 times more likely to receive a score of 0 for feather cleanliness than ducks at 20 days (CI: 3.6777-221.50, p = 0.0013), 49.5628 times more likely than ducks at 27 days (CI: 6.4458-381.10, p = 0.0002), 229.30 times more likely than ducks at 35 days (CI: 29.5521-1779.2, p < 0.0001), and 4556.20 times more likely than ducks at 42 days (CI: 557.50-37236.50, p < 0.0001). Ducks at 20 days were 8.0348 times more likely to receive a score of 0 for feather quality than ducks at 35 days (CI: 3.7678-17.1342, p < 0.0001) and 159.60 times more likely than ducks at 42 days (CI: 65.8433-387.10, p < 0.0001). Ducks at 27 days were 4.6265 times more likely to receive a score of 0 for feather cleanliness than ducks at 35 days (CI: 2.2305-9.5964, p < 0.0001) and 91.9276 times more likely than ducks at 42 days (CI: 38.9256-217.10, p < 0.0001). Ducks at 35 days were 19.8698 times more likely to receive a score of 0 for feather cleanliness than ducks at 42 days (CI: 8.6952-45.4051, p < 0.0001).

Treatment. Ducks in the CON group were 28.0700 times more likely to receive a score of 0 for feather cleanliness than ducks in the PC group (CI: 12.9526-60.8317, p < 0.0001) and 72.3324 times more likely than ducks in the PEK group (CI: 30.8351-169.70, p < 0.0001). Ducks in the EED group were 23.3521 times more likely to receive a 0 for feather cleanliness than ducks in the PC group (CI: 10.8209-50.3950, p < 0.0001) and 60.1749 times more likely than ducks in the PEK group (CI: 25.7538-140.60, p < 0.0001). Ducks in the PC group were 2.5769 times more likely to receive a 0 for feather cleanliness than ducks in the PEK group (CI: 1.1589-5.7297, p = 0.0202).

Days of age x Treatment Interaction. At 35 days (Wald X²=34.9788, p < 0.0001), ducks in the CON group were 2.4952 times more likely to receive a score of 0 for feather cleanliness than ducks in the PC group (CI: 1.1864-5.2481, p = 0.0159) and 9.4385 times more likely than ducks in the PEK group (CI: 4.2686-20.8699, p < 0.0001). Ducks in the EED group were 6.1938 times more likely to receive a score of 0 for feather cleanliness compared to ducks in the PEK group (CI: 2.8853-13.2961, p < 0.0001). Ducks in the PC group were 3.7826 times more likely to receive a score of 0 for feather cleanliness than ducks in the PEK group (CI: 1.7975-7.9602, p = 0.0005).

At 42 days (Wald X²=90.1632, p < 0.0001), ducks in the CON group were 17.1353 times more likely to receive a score of 0 for feather cleanliness than ducks in the PC group (CI: 7.6062-38.6021, p < 0.0001) and 42.9601 times more likely than ducks in the PEK group (CI: 17.4899-105.50, p < 0.0001). Ducks in the EED group were 14.4974 times more likely to receive a score of 0 for feather cleanliness compared to ducks in the PC group (CI: 6.5005-32.3319, p < 0.0001) and 36.3467 times more likely than ducks in the PEK group (CI: 14.9406-88.4221, p < 0.0001). Ducks in the PC group were 2.5071 times more likely to receive a score of 0 for feather cleanliness than ducks in the PEK group (CI: 1.1263-5.5809, p = 0.0244).

Foot pad scores

Foot pad scores were significantly influenced by age (Wald X²=27.0975, p < 0.0001) and the interaction between days of age and treatment (Wald X²=29.1835, p = 0.0037). Data are shown in Table 5.

Table 5.

Foot pad score odds ratio.

Foot Pad Probabilities
Days of age	Contrast	Estimate	Confidence Limits		P Value
	CON vs EED	0.7826	0.3515	1.7426	0.5484
	CON vs PC	1.6502	0.7776	3.5020	0.1920
	CON vs PEK	0.8518	0.3858	1.8809	0.6915
	EED vs PC	2.1086	0.9712	4.5779	0.0593
	EED vs PEK	1.0884	0.4824	2.4560	0.8383
	PC vs PEK	0.5162	0.2398	1.1110	0.0909
20	CON vs EED	0.4949	0.1702	1.4390	0.1965
	CON vs PC	0.1536	0.0325	0.7265	0.0181
	CON vs PEK	0.0755	0.0094	0.6055	0.0150
	EED vs PC	0.3103	0.0600	1.6041	0.1627
	EED vs PEK	0.1526	0.0178	1.3082	0.0864
	PC vs PEK	0.4917	0.0434	5.5705	0.5665
35	CON vs EED	1.5143	0.5351	4.2856	0.4344
	CON vs PC	2.3043	0.8566	6.1986	0.0982
	CON vs PEK	0.1283	0.0153	1.0776	0.0586
	EED vs PC	1.5217	0.6157	3.7613	0.3632
	EED vs PEK	0.0847	0.0105	0.6851	0.0206
	PC vs PEK	0.0557	0.0071	0.4392	0.0061

Open in a new tab

Age. Ducks at 14 days were 0.0537 times less likely to receive a score of 0 for foot pad when compared to ducks at 20 days (CI: 0.00959-0.3001, p = 0.0009), 0.0176 times less likely when compared to ducks at 27 days (CI: 0.00101-0.3065, p = 0.0056), and 0.0537 times less likely when compared to ducks at 35 days (CI: 0.00959-0.3001, p = 0.0009). Ducks at 20 days were 14.7079 times more likely to receive a score of 0 for foot pad when compared to ducks at 42 days (CI: 2.6068-82.9830, p = 0.0023). Ducks at 27 days were 44.8195 times more likely to receive a score of 0 for foot pad when compared to ducks at 42 days (CI: 2.5612-784.3, p = 0.0092). Ducks at 35 days were 14.7079 times more likely to receive a score of 0 for foot pad when compared to ducks at 42 days (CI: 2.6068-82.9830, p = 0.0023).

Days of age x Treatment Interaction. At 20 days (Wald X²=10.2854, p = 0.0163), ducks in the CON group were 0.1536 times less likely to receive a score of 0 for foot pad when compared to ducks in the PC (CI: 0.0325-0.7265, p = 0.0181) and 0.0755 times less likely than ducks in the PEK group (CI: 0.00942-0.6055, p = 0.0150).

At 35 days (Wald X²=8.9377, p = 0.0301), ducks in the EED group were 0.0847 times less likely to receive a score of 0 for foot pad when compared to ducks in the PEK group (CI: 0.0105-0.6851, p = 0.0061). Ducks in the PC group were 0.0557 times less likely to receive a score of 0 for foot pad when compared to ducks in the PEK group (CI: 0.00706-0.4392, p = 0.0061).

Eye health

Eye health was significantly influenced by age (Wald X²=18.9863, p = 0.0008), treatment (Wald X²=8.1837, p = 0.0424) and their interaction (Wald X²⁼32.9130, p = 0.0010). Data shown in Table 6.

Table 6.

Eye health odds ratio.

Eye Probabilities
Days of age	Contrast	Estimate	Confidence Limits		P Value
	CON vs EED	1.7316	0.7362	4.0728	0.2083
	CON vs PC	0.3428	0.1070	1.0983	0.0715
	CON vs PEK	1.1094	0.4507	2.7308	0.8213
	EED vs PC	0.1980	0.0647	0.6056	0.0045
	EED vs PEK	0.6406	0.2765	1.4843	0.2989
	PC vs PEK	3.2360	1.0213	10.2530	0.0460
20	CON vs EED	0.3220	0.0325	3.1873	0.3326
	CON vs PC	1.7273	0.3938	7.5766	0.4688
	CON vs PEK	10.2308	2.8547	36.6657	0.0004
	EED vs PC	5.3636	0.6074	47.3643	0.1307
	EED vs PEK	31.7691	4.1043	245.9000	0.0009
	PC vs PEK	5.9231	2.0560	17.0636	0.0010
27	CON vs EED	2.0345	0.1795	23.0554	0.5664
	CON vs PC	3.3E-05	0.0000	1.20E+143	0.9525
	CON vs PEK	29.5000	3.8049	228.7000	0.0012
	EED vs PC	1.6E-05	0.0000	5.90E+142	0.9492
	EED vs PEK	14.5000	3.2085	65.5283	0.0005
	PC vs PEK	901594	0.0000	3.30E+153	0.9370

Open in a new tab

Age. Ducks at 14 days were 12.7367 times more likely to receive a score of 0 for eye health than ducks at 20 days (CI: 3.1984-50.7203, p = 0.0003), 11.8444 times more likely than ducks at 27 days (CI: 2.9674-47.2767, p = 0.0005), 4.3165 times more likely than ducks at 35 days (CI: 1.0079-18.4868, p = 0.0488), and 6.0309 times more likely than ducks at 42 days (CI: 1.4549-24.9992, p = 0.0133). Ducks at 20 days were 0.3389 times less likely to receive a score of 0 for eye health than ducks at 35 days (CI: 0.1411-0.8138, p = 0.0155). Ducks at 27 days were 0.3644 times less likely to receive a score of 0 for eye health than ducks at 35 days (CI: 0.1512-0.8784, p = 0.0245).

Treatment. Ducks in the EED group were 0.1980 times less likely to receive a score of 0 for eye health than ducks in the PC group (CI: 0.0647-0.6056, p = 0.0045). Ducks in the PC group were 3.2360 times more likely to receive a 0 for eye health than ducks in the PEK group (CI: 1.0213-10.2530, p = 0.0460).

Days of age x Treatment Interaction. At 20 days (Wald X²=26.6963, p < 0.0001), ducks in the CON group were 10.2308 times more likely to receive a score of 0 for eye health than ducks in the PEK group (CI: 2.8547-36.6657, p = 0.0004). Ducks in the EED group were 31.7691 times more likely to receive a score of 0 for eye health than ducks in the PEK group (CI: 4.1043-245.90, p = 0.0009). Ducks in the PC group were 5.9231 times more likely to receive a score of 0 for eye health than ducks in the PEK group (CI: 2.0560-17.0636, p = 0.0010).

At 27 days (Wald X²=20.6503, p = 0.0001), ducks in the CON group were 29.5000 times more likely to receive a score of 0 for eye health than ducks in the PEK group (CI: 3.8049-228.70, p = 0.0012). Ducks in the EED group were 14.5000 times more likely to receive a score of 0 for eye health than ducks in the PEK group (CI: 3.2085-65.5283, p = 0.0005).

At 42 days (Wald X²=8.5222, p = 0.0364), ducks in the EED group were 0.1807 times less likely to receive a score of 0 for eye health than ducks in the PC group (CI: 0.0567-0.5760, p = 0.0038). Ducks in the PC group were 3.500 times more likely to receive a score of 0 for eye health than ducks in the PEK group (CI: 1.0590-11.5677, p = 0.0400).

Nostril scores

Nostril scores were significantly influenced by age (Wald X²=27.3477, p < 0.0001), treatment (Wald X²=18.1251, p = 0.0004) and their interaction (Wald X²=34.6719, p = 0.0005). Data shown in Table 7.

Table 7.

Nostril health odds ratio.

Nostril Probabilities
Days of age	Contrast	Estimate	Confidence Limits		P Value
	CON vs EED	0.7590	0.3656	1.5761	0.4596
	CON vs PC	0.2912	0.1376	0.6164	0.0013
	CON vs PEK	0.2700	0.1269	0.5743	0.0007
	EED vs PC	0.3837	0.1830	0.8044	0.0112
	EED vs PEK	0.3557	0.1688	0.7496	0.0066
	PC vs PEK	0.9269	0.4317	1.9904	0.8457
20	CON vs EED	3.8642	1.8104	8.2476	0.0005
	CON vs PC	1.3138	0.6349	2.7187	0.4620
	CON vs PEK	0.7464	0.3516	1.5846	0.4464
	EED vs PC	0.3400	0.1607	0.7195	0.0048
	EED vs PEK	0.1932	0.0890	0.4192	<.0001
	PC vs PEK	0.5682	0.2699	1.1961	0.1366
27	CON vs EED	1.6154	0.7337	3.5566	0.2337
	CON vs PC	0.2143	0.0665	0.6909	0.0099
	CON vs PEK	0.0509	0.0065	0.3994	0.0046
	EED vs PC	0.1327	0.0422	0.4167	0.0005
	EED vs PEK	0.0315	0.0041	0.2436	0.0009
	PC vs PEK	0.2373	0.0257	2.1883	0.2044
35	CON vs EED	1.9023	0.8582	4.2166	0.1133
	CON vs PC	0.1729	0.0468	0.6384	0.0085
	CON vs PEK	0.1133	0.0245	0.5240	0.0053
	EED vs PC	0.0909	0.0254	0.3251	0.0002
	EED vs PEK	0.0596	0.0132	0.2681	0.0002
	PC vs PEK	0.6553	0.1055	4.0691	0.6501
42	CON vs EED	0.7571	0.3640	1.5748	0.4565
	CON vs PC	0.2895	0.1366	0.6134	0.0012
	CON vs PEK	0.2683	0.1260	0.5715	0.0006
	EED vs PC	0.3824	0.1823	0.8021	0.0110
	EED vs PEK	0.3544	0.1681	0.7474	0.0064
	PC vs PEK	0.9269	0.4317	1.9903	0.8457

Open in a new tab

Age. Ducks at 14 days were 4.1671 times more likely to receive a score of 0 for nostril than ducks at 20 days (CI: 1.5275-11.3692, p = 0.0053), and 4.1671 more likely than ducks at 42 days (CI: 1.5274-11.3692, p = 0.0053). Ducks at 20 days were 0.0366 times less likely to receive a score of 0 for nostril than ducks at 27 days (CI: 0.00472-0.2844, p = 0.0016) and 0.0745 times less likely than ducks at 35 days (CI: 0.0164-0.3375, p = 0.0008). Ducks at 27 days were 27.2944 times more likely to receive a score of 0 for nostril than ducks at 42 days (CI: 3.5159-211.90, p = 0.0016). Ducks at 35 days were 13.4256 times more likely to receive a score of 0 for nostril than ducks at 42 days (CI: 2.9630-60.8314, p = 0.0008).

Treatment. Ducks in the CON group were 0.2912 times less likely to receive a score of 0 for nostril than ducks in the PC group (CI: 0.1376-0.6164, p = 0.0013) and 0.2700 times less likely than ducks in the PEK group (CI: 0.1269-0.5743, p = 0.0007). Ducks in the EED group were 0.3837 times less likely to receive a score of 0 for nostril than ducks in the PC group (CI: 0.1830-0.8044, p = 0.0112) and 0.3557 times less likely than ducks in the PEK group (CI: 0.1688-0.7496, p = 0.0066).

Days of age x Treatment Interaction. At 20 days (Wald X²=19.7934, p = 0.0002), ducks in the CON group were 3.8642 times more likely to receive a score of 0 for nostril than ducks in the EED group (CI: 1.8104-8.2476, p = 0.0005). Ducks in the EED group were 0.3400 times less likely to receive a score of 0 for nostril than ducks in the PC group (CI: 0.1607-0.7195, p = 0.0048) and 0.1932 times less likely than ducks in the PEK group (CI: 0.0890-0.4192, p < 0.0001).

At 27 days (Wald X²=20.5709, p = 0.0001), ducks in the CON group were 0.2143 times less likely to receive a score of 0 for nostril than ducks in the PC group (CI: 0.0665-0.6909, p = 0.0099) and 0.0509 times less likely than ducks in the PEK group (CI: 0.00647-0.3994, p = 0.0046). Ducks in the EED group were 0.1327 times less likely to receive a score of 0 for nostril than ducks in the PC group (CI: 0.0422-0.4167, p = 0.0005) and 0.0315 times less likely than ducks in the PEK group (CI: 0.00407-0.2436, p = 0.0009).

At 35 days (Wald X²=23.7046, p < 0.0001), ducks in the CON group were 0.1729 times less likely to receive a score of 0 for nostril than ducks in the PC group (CI: 0.0468-0.6384, p = 0.0085) and 0.1133 times less likely than ducks in the PEK group (CI: 0.0245-0.5240, p = 0.0053). Ducks in the EED group were 0.0909 times less likely to receive a score of 0 for nostril than ducks in the PC group (CI: 0.0245-0.3251, p = 0.0002) and 0.0596 times less likely than ducks in the PEK group (CI: 0.0132-0.2681, p = 0.0002).

At 42 days (Wald X²=18.2376, p = 0.0004), ducks in the CON group were 0.2895 times less likely to receive a score of 0 for nostril than ducks in the PC group (CI: 0.1366-0.6134, p = 0.0012) and 0.2683 times less likely than ducks in the PEK group (CI: 0.1260-0.5715, p = 0.0006). Ducks in the EED group were 0.3824 times less likely to receive a score of 0 for nostril than ducks in the PC group (CI: 0.1823-0.8021, p = 0.0110) and 0.3544 times less likely than ducks in the PEK group (CI: 0.1681-0.7474, p = 0.0064).

Litter moisture and ammonia

No significant differences were found among treatments for ammonia levels. Litter moisture percentage was significantly influenced by treatment (F_3,45=33.21, p < 0.0001). Litter from the PC group contained a higher moisture content than litter in the CON (p < 0.0001) and EED groups (p < 0.0001). Litter from the PEK group contained a higher moisture content than litter in the CON (p < 0.0001) and EED groups (p < 0.0001). Data shown in Fig. 4.

Fig 4 — Data for (A) ammonia level and (B) litter moisture. No significant differences were found among treatments for ammonia level. PEK litter had a higher moisture percentage than the EED, PC, and CON litter (P < 0.0001).

Behavior

Wet preening

The average proportion of ducks wet preening was significantly influenced by week (F_2,841=11.25, p < 0.0001) and treatment (F_3,841=3.10, p = 0.0262). Ducks wet preened less at week 1 compared to week 2 (−0.2827±0.08138, p = 0.0016) and 3 (−0.3690±0.08139, p < 0.0001). More ducks in the CON group wet preened than ducks in the EED group (0.6465±0.2446, p = 0.0084). Data shown in Fig. 5A.

Fig 5 — Behavior data. Significant differences were found among treatments for wet preening (A), dry preening (B), preening others (C), drinking from nipple line (D), interacting with enrichment source (E; EED interacting with EED and PC/PEK ducks dunking their heads in PC and PEK), and drinking from nipple line/water source (F; EED and CON drinking from nipple line, PEK and PC drinking from PEK and PC). We also observed a week effect for interacting with EED (G; EED ducks interacting with EED). Letters indicate statistical significance, p < 0.05.

Dry preening

The average proportion of ducks dry preening was significantly influenced by week (F_2,841=79.04, p < 0.0001), time (F_5,841=11.80, p < 0.0001), and treatment (F_3,841=6.93, p = 0.0001). Less ducks dry preened at week 1 compared to week 2 (−2.0402±0.2755, p < 0.0001) and week 3 (−3.4443±0.2755, p < 0.0001). Less ducks dry preened at week 2 compared to week 3 (−1.4041±0.2755, p < 0.0001). More ducks dry preened at 0330 h compared to 0430 h (2.1911±0.3895, p < 0.0001) and 0530 h (2.7163±0.3896, p < 0.0001), 1430 h (2.1020±0.3896, p < 0.0001), 1530 h (2.1757 ± 0.3896, p < 0.0001), and 1630 h (1.9965±0.3896, p < 0.0001). More ducks in the PEK group dry preened than ducks in the CON group (2.9865±0.7592, p = 0.0005), and ducks in the EED group (3.0028±0.7592, p = 0.0005). Data shown in Fig. 5B.

Preening others

The average proportion of ducks preening others was significantly influenced by treatment (F_3,841=5.72, p = 0.0007). More ducks in the CON group preened others than ducks in the EED (0.6695±0.1761, p = 0.0009) and PC groups (0.5846±0.1761, p = 0.0052). Data shown in Fig. 5C.

Feather pecking

The average proportion of ducks feather pecking was significantly influenced by week (F_2,841=18.45, p < 0.0001) and time (F_5,841=2.93, p = 0.0124). More ducks feather pecked at week 3 compared to week 1 (0.1519±0.02518, p < 0.0001) and 2 (0.06077±0.02518, p = 0.0423). More ducks feather pecked at week 2 compared to week 1 (0.09114±0.02518, p = 0.0009). More ducks feather pecked at 0330 h compared to 1530 h (0.1248±0.03560, p = 0.0064), and 1630 h (0.1063±0.03560, p = 0.0345).

Feather picking

No significant differences were observed for feather picking.

Drinking from the nipple line

The average proportion of ducks drinking from the nipple line was significantly influenced by week (F_2,841=8.18, p = 0.0003), time (F_5,841=28.21, p < 0.0001), and treatment (F_3,841=60.08, p = 0.0001). More ducks drank from the nipple line at week 3 compared to week 1 (1.1672±0.3088, p = 0.0005) and 2 (0.9697±0.3088, p = 0.0050). More ducks drank from the nipple line at 0330 h compared to 0430 h (2.6123±0.4368, p < 0.0001) 0530 h (3.7772±0.4368, p < 0.0001), 1430 h (4.0768±0.4368, p < 0.0001), 1530 h (3.5799±0.4368, p < 0.0001), and 1630 h (3.6561±0.4368, p < 0.0001). More ducks also drank from the nipple line at 0430 h compared to 1430 h (1.4646±0.4368, p = 0.0108). More ducks in the CON group drank from the nipple line than ducks in the EED (2.3442 ± 0.7854, p = 0.0154), PC (8.4766±0.7854, p < 0.0001), and PEK groups (8.4051±0.7854, p < 0.0001). More ducks in the EED group drank from the nipple line than ducks in the PC (6.1324±0.7854, p < 0.0001) and PEK groups (6.0609±0.7854, p < 0.0001). Data shown in Fig. 5D.

Eating

The average proportion of ducks eating was significantly influenced by week (F_2,841=6.94, p = 0.0010) and time (F_5,841=3.06, p = 0.0097). Less ducks ate at week 3 compared to week 1 (−0.3763±0.1049, p = 0.0010) and 2 (−0.2791±0.1049, p = 0.0216). More ducks ate at 0430 h compared to 0530 h (0.4338±0.1483, p = 0.0411), and 1530 h (0.4991±0.1483, p = 0.0103).

Standing

The average proportion of ducks standing was significantly influenced by week (F_2,841=89.19, p < 0.0001) and time (F_5,841=14.72, p < 0.0001). More ducks stood on week 3 compared to week 1 (10.3747±0.7843, p < 0.0001) and 2 (6.4351±0.7843, p < 0.0001). More ducks also stood on week 2 compared to week 1 (3.9395±0.7843, p < 0.0001). More ducks stood at 0330 h compared to 0430 h (6.4392±1.1091, p < 0.0001), 0530 h (6.4454±1.1091, p < 0.0001), 1430 h (7.5165±1.1091, p < 0.0001), 1530 h (8.4114±1.1091, p < 0.0001), and 1630 h (6.8773±1.1091, p < 0.0001).

Laying down

The average proportion of ducks laying down was significantly influenced by week (F_2,841=92.93, p < 0.0001) and time (F_5,841=33.23, p < 0.0001). More ducks laid down at week 1 compared to week 2 (6.2830±1.0877, p < 0.0001) and 3 (14.7744±1.0877, p < 0.0001). More ducks also laid down at week 2 compared to week 3 (8.4914±1.0877, p < 0.0001). Less ducks laid down at 0330 h compared to 0430 h (−12.1399±1.5383, p < 0.0001), 0530 h (−16.0966±1.5383, p < 0.0001), 1430 h (−15.5315±1.5383, p < 0.0001), 1530 h (−16.0847±1.5383, p < 0.0001), and 1630 h (−14.7895±1.5383, p < 0.0001).

Drinking from semi-open water source

The average proportion of ducks drinking from the PEK or PC was significantly influenced by week (F_2,417=22.09, p < 0.0001). Though there was no significant difference between the PC and PEK groups for drinking from the semi-open water source. More ducks drank from the semi-open source at week 1 compared to week 2 (0.7501±0.2185, p = 0.0019) and 3 (1.4523±0.2185, p < 0.0001). More ducks also drank from the semi-open source at week 2 compared to week 3 (0.7022±0.2185, p = 0.0040).

Dunking head in semi-open water source

The average proportion of ducks dunking their head in the semi-open water source was significantly influenced by week (F_3,417=49.33, p < 0.0001), and time (F_5,417=22.94, p < 0.0001). More ducks dunked their heads in the semi-open water source at week 3 compared to week 1 (2.9682±0.3015, p < 0.0001) and 2 (1.8260±0.3015, p < 0.0001). More ducks also dunked their heads in the semi-open water source at week 2 compared to 1 (1.1422±0.3015, p = 0.0005). More ducks dunked their heads into the semi-open water source at 0330 h compared to 0430 h (2.4141±0.4263, p < 0.0001), 0530 h (3.6525±0.4263, p < 0.0001), 1430 h (3.3139±0.4263, p < 0.0001), 1530 h (3.3721±0.4263, p < 0.0001), and 1630 h (3.8599±0.4263, p < 0.0001). More ducks also dunked their heads into the semi-open water source at 0430 h compared to 0530 h (1.2384±0.4263, p = 0.0445), and 1630 h (1.4458±0.4263, p = 0.0099).

Interacting with enrichment/dunking head in semi-open water

The average proportion of ducks interacting with the enrichment (PEK, PC dunking head in preening cup and EED) was significantly influenced by treatment (F_2,636 = 9.23, p = 0.0001). More ducks in the PEK group interacted with their enrichment than the ducks in the EED group (1.9176±0.4464, p < 0.0001). Data shown in Fig. 5E.

Drinking from nipple line/drinking from water enrichment

The average proportion of ducks drinking from the nipple line (EED, CON) and drinking from the water enrichment (PEK, PC) was significantly influenced by treatment (F_3,841=37.65, p < 0.0001). More ducks in the CON group drank from the nipple line than ducks in the PEK (7.1835±0.8480, p < 0.0001) and PC groups (7.4621±0.8480, p < 0.0001) drank from their open water source. More ducks in the EED group drank from the nipple line than the ducks in the PEK (4.8393±0.8480, p < 0.0001) and PC groups (5.1179±0.8480, p < 0.0001) drank from their open water source. Data shown in Fig. 5F.

Interacting with EED

The average proportion of ducks interacting with the EED was significantly influenced by week (F_2,195=6.57, p = 0.0017). More ducks interacted with the EED on week 1 compared to week 2 (0.6926±0.2612, p = 0.0234) and 3 (0.9049±0.2612, p = 0.0019). Data shown in Fig. 5G.

Mass spectrometry

Data from the CD, RM, and DI regions showed no significant differences among DA or 5HT, or their metabolites (data not shown), levels nor in the respective DA and 5HT turnovers. In the PEK group only, a difference (p = 0.006) was observed in the T for 5HT levels but only between 16 and 44 days of age. No differences were observed in the metabolite levels, and thus not in the respective turnover.

Viable bacteria

There was a significant treatment effect (F₃=33.6067, p < 0.0001) and a significant collection time effect (F₂=31.0065, p < 0.0001). The CON nipple line contained less viable bacteria than the PC (p < 0.0001) and the PEK (p < 0.0001). The EED contained less viable bacteria than the PC (p < 0.0001) and the PEK (p < 0.0001). The PC had less viable bacteria than the PEK (p = 0.0139). Results shown in Fig. 6. These results show that the PEK had the most amount of viable bacteria in the water than any of the other treatments, while the EED and CON groups had the least amount of viable bacteria.

Fig 6 — Data for viable bacteria. The CON nipple line contained less viable bacteria than the PC and PEK. The EED contained less viable bacteria than the PC and PEK, and the PC had less viable bacteria than the PEK. Letters indicate statistical significance, p < 0.05.

Microbial diversity and community composition

Shannon diversity differed significantly between treatments (H = 9.07, Kruskal-Wallis test, p = 0.028) (Fig. 7). The EED group exhibited significantly lower Shannon diversity compared to the CON group (H = 7.87, padj=0.032). However, no significant differences were observed between EED and PEK (H = 1.03, padj=0.308) or EED and PC (H = 2.47, padj=0.235).

Faith’s PD showed significant variation among treatments (H = 27.68, Kruskal-Wallis test, p < 0.0001) (Fig. 7). Faith’s PD was significantly higher in the CON group compared to the PEK (H = 24.23, padj <0.0001), PC (H = 6.61, padj=0.015), and EED (H = 8.90, padj=0.006) groups. Additionally, Faith’s PD was significantly lower in the PEK group compared to the EED group (H = 12.52, padj=0.001).

Observed features also varied significantly between treatments (H = 10.41, Kruskal-Wallis test, p = 0.015) (Fig. 7). The EED group exhibited significantly lower observed features compared to the CON group (H = 6.21, padj=0.038). However, no significant differences were noted between EED and PEK (H = 1.70, padj=0.235) or EED and PC (H = 0.26, padj=0.607).

Evenness differed significantly across treatments (H = 11.12, Kruskal-Wallis test, p = 0.011) (Fig. 7). Evenness was significantly lower in the EED group compared to the CON (H = 7.92, padj=0.019), PEK (H = 7.47, padj=0.019), and PC (H = 5.48, padj=0.039) groups. However, no significant differences were observed between PEK and PC (H = 1.70, padj=0.235) or between EED and PC (H = 0.26, padj=0.607).

Beta diversity analysis revealed distinct microbial community structures across treatments and timepoints (Fig. 8). At baseline, EED samples formed a tight, distinct cluster separate from other treatments, while PC and CON samples showed considerable overlap in their microbial communities. PEK treatment formed a discrete cluster with partial overlap with PC samples (Fig. 8A). After 6 h, marked shifts in community structure were observed (Fig. 8B). EED, PC, and PEK treatments formed more compact, partially overlapping clusters. However, CON samples maintained a distinct community structure separate from other treatments (Fig. 8B).

At baseline (24 h after enrichment device placement), bacterial communities exhibited distinct treatment-dependent patterns (Fig. 9C). Among the genera, Pseudomonas showed the most pronounced differences (Kruskal-Wallis, H = 22.8, p = 0.00035), with higher abundance in EED compared to all other treatments. Similarly, Achromobacter displayed significant variation across treatments (H = 17.8, p = 0.00203), with the EED group exhibiting higher levels compared to the other treatments. Opportunistic pathogens also differed significantly, with Moraxella showing higher abundances in PEK and PC compared to EED (H = 14.9, p = 0.00665; padj=0.001 and 0.014, respectively). Staphylococcus abundance varied significantly between treatments (H = 12.4, p = 0.01491), with the CON group exhibiting higher levels compared to PEK (padj=0.003).

By 6 h after baseline, notable shifts in the bacterial community structure were observed (Fig. 9D). Pseudomonas continued to display significant differences between treatments (H = 26.4, p < 0.0001). However, the pattern shifted, with CON and EED treatments showing significantly higher levels compared to PC and PEK treatments. This indicates temporal changes in Pseudomonas colonization across enrichment systems. Acinetobacter emerged as a prominent component at this time point (H = 17.8, p = 0.002), with the CON treatment exhibiting higher abundance compared to PC and PEK treatments (padj=0.007 and 0.002, respectively).

Moraxella maintained significant variation across treatments at 6 h (H = 24.6, p < 0.0001), with consistently higher levels in PEK and PC treatments compared to other groups. The difference was particularly pronounced between PC and EED treatments, indicating that open water systems consistently enrich Moraxella growth, regardless of exposure duration. Additionally, Corynebacterium showed significant treatment effects at this time point (H = 18.9, p = 0.001). Open water systems (PEK and PC) exhibited higher levels compared to conventional systems, further emphasizing the distinct microbial communities associated with different enrichment devices. These findings highlight the temporal dynamics and treatment-specific microbial signatures, with implications for pathogen exposure and microbial management in duck production systems.

Functional pathways

Functional pathway analysis revealed distinct metabolic profiles across treatments at baseline (24 h after enrichment device placement) and 6 h after baseline (Fig. 10). At baseline, PC and EED treatments exhibited notably higher abundances in several metabolic pathways, including aerobic respiration I, amino acid biosynthesis pathways (e.g., l-isoleucine and l-valine), and various metabolic superpathways. In contrast, CON and PEK treatments displayed comparatively lower pathway abundances (Fig. 10).

Fig 10 — Functional profiling of microbial communities reveals differences in metabolic pathways across enrichment treatments (CON, EED, PC, PEK) and time points (A: baseline, B: 6 h after baseline) in grow-out Pekin ducks. The heatmap displays the top 30 metabolic pathways based on abundance. Treatments include Nipple line (CON, control), Whiffle ball (EED, non-water enrichment), Preening cup (PC, semi-open water), and Pekino (PEK, open water). Pathway abundances are represented as log-transformed values, with darker shades indicating higher abundance. Sequencing and functional pathway analysis were conducted as described in the Methods.

By 6 h after baseline, the metabolic profiles demonstrated a marked shift, with pathway abundances becoming more uniform across all treatments (Fig. 10). Core metabolic pathways, such as the TCA cycle, nucleotide biosynthesis, and amino acid metabolism, showed comparable abundance levels across treatments.

The early differences in metabolic potential between treatments (most notably between PC/EED and CON/PEK) at baseline were not sustained at 6 h after baseline (Fig. 10).

Discussion

The purpose of our study was to evaluate whether preening cups, Pekinos, or EED placed with 30 ducks per pen would alter their production, environment, welfare, and behavior compared to ducks housed with nipple lines alone. To accomplish this goal, we placed grow-out Pekin ducks in pens with either nipple-lines alone, or with a combination of nipple line and preening cup, nipple line and 2 Pekinos, or nipple line with 2 EED. After the enrichment was placed, significant differences were observed for wet preening and dry preening, where more ducks in CON group wet-preened than the EED ducks and more ducks in the PEK group dry-preened than the CON and EED ducks. Previous studies showed an increase in wet preening when ducks were housed with an open water source compared to ducks housed with only nipple lines (Jones et al., 2009; Schober et al., 2023). Significant differences were also observed for drinking from the nipple line, where more CON and EED ducks drank from their nipple line than the PC and PEK ducks drank from their nipple line. This suggests that the semi-open water sources were a more efficient, or preferred, vehicle for the ducks to access water. This finding agrees with our previous observations in Schober et al. (2023), where the control ducks drank more from the nipple line than the PC ducks.

Looking at duck interactions with enrichments holistically, more ducks dunked their heads in the PEK than ducks interacted with their enrichment in the EED group. More ducks also drank from the nipple line in the CON and EED group than ducks drank from the semi-open water source in the PEK and PC groups. One possible explanation is that the ducks who drank from the semi-open water source could have ingested more water during that one interaction compared to CON and EED ducks that are unable to consume as much water at once with the nipple lines. This causes the CON and EED ducks to interact more with their nipple lines for drinking than the ducks in the PC and PEK groups drinking from their semi-open water sources. We showed that ducks prefer the semi-open water source over nipple lines for access to water, but the presence of the semi-open water sources does not increase nor decrease the efficacy of their preening behaviors as shown by behavior, body condition scores, and uropygial gland.

Following dissections, we also observed specifically on day 44, that the PC ducks weighed more than the EED and CON ducks, and that the PEK ducks weighed more than the EED ducks. This aligns with the significant treatment effect of weekly body weights, where on 42 days of age, the PC and PEK ducks weighed more than the CON and EED ducks. Previous studies have shown no differences between ducks housed with open water versus ducks housed with only nipple lines (Jones at al., 2009; Schenk et al., 2016). While a study looking at narrow bell drinkers, baths, and troughs showed that ducks housed with baths or troughs weighed more than ducks housed with narrow bell drinkers (O’Driscoll & Broom, 2011). The significant differences with body weight within the current study could be due to the PC ducks and PEK ducks ingesting more water. It could also be due to the ducks with the semi-open sources having wetter feathers. Future studies should examine the carcass weight of ducks housed with open water compared to ducks housed with just nipple lines.

The PEK ducks also had a larger uropygial gland than the CON, EED, and PC ducks, but only before the PEK was actually placed. The larger uropygial gland in this group could be due to the semi-open water, but that is unlikely, and this difference was likely just due to random effects. There have been differing results on whether or not open water access affects the size of the uropygial gland. One study showed that taking away open-water access decreases the size of the uropygial gland in Sanshui white ducks (Mi et al., 2020). While a different study has shown that the uropygial gland size is not affected with implemented semi-open water sources in Pekin ducks (Schober et al., 2023). These discrepancies suggest that factors such as species differences, environmental conditions, and specific husbandry practices may influence the outcomes.

Litter from the PC and PEK pens contained a higher percentage of moisture than the litter of the CON and EED pens. Previous studies have also shown that open water sources cause an increase in litter moisture percentage (Lowman et al., 2016; Schenk et al., 2016). This could be due to ducks being able to dunk their heads into the semi-open water and splashing water onto their backs and breast. When the ducks then lay down in the litter, this causes the litter to become wet. However, there were no differences among treatments for ammonia levels within the pens. This is likely due to the pit system underneath the water lines implemented in our research barns as well as all 4 treatments being in each room. We also saw an increase in viable bacteria for both the PEK and PC compared to EED and CON, similar to bacteria loads shown in previous studies with open water sources (Schenk et al., 2016). This observation could be due to the ducks’ ability to place their whole bills/heads into the water. Ducks are known to use water while they eat, 5x as much as other poultry species (Chen et al., 2021), in fact, that can cause high protein food remnants to contaminate the PEK and PC. This could lead to an increase in live bacteria due to deposition by the duck. Also feed contamination adds nutrients to the water producing a viable culture for bacteria to thrive. Schenk et al. (2016) also saw an increase in water contamination in water troughs. Both open and semi-open water sources have shown to increase litter moisture % and viable bacteria.

PEK and PC ducks had better nostril scores but had dirtier feathers than the EED and CON ducks. EED and PC ducks had worse feather quality, where EED and PC ducks had worse quality. Schober et al. (2023) saw an increase, though not significant from behavioral analyses, in feather pecking within the pens with the preening cups. However, body condition scores from those ducks suggested that considerable feather pecking did occur, but not observed on videos. Bergman et al. (2024) saw an increase in dopamine within the brains of those same ducks with the preening cups. Dopamine is a hormone associated with aggression, which supports the idea that ducks with the PC performed more aggressive behaviors (feather pecking) than ducks with just the nipple line. Previous studies have linked aggression and increased feather pecking to increased DA activity (Bergman et al., 2024; Biscarini at al., 2010; Dennis et al., 2011; de Haas and Eijk, 2018; Narvaes and Almeida, 2014). Bergman et al. (2024) concluded that providing ducks with a semi-open water source may cause an increase in reward as they interact with it, resulting in an increase activity of DA in their brains. Thus, predisposing them to aggressive and resource guarding behaviors. However, in this current study we did not see significant differences among groups for feather pecking. The poor body condition scores for feather quality in the EED and PC ducks suggest that feather pecking was occurring, so this discrepancy could be due to times of the day analyzed where they are not exhibiting those behaviors.

The impact of environmental enrichment strategies (e.g. whiffle ball) and open water availability on duck welfare extends beyond behavior to include potential duck exposure to diverse microbial communities. Indeed, drinking water in poultry production contains diverse microbial genera, of which some (e.g. Escherichia) are typical members of the poultry gut and skin microbiotas (Rychlik et al., 2023; Musteanagic et al., 2023). As ducks are waterfowl, the use of semi- or open-water sources provide unique aquatic niches in the production environment that may cultivate distinct microbial communities to create unique opportunities for microbial exposure that could be strategized by the producer to improve duck welfare. The findings of this study demonstrate that enrichment system type affects microbial exposure and pathogen risk in grow-out Pekin ducks. Semi-open water systems (PEK and PC) were associated with higher microbial diversity and proliferation of opportunistic pathogens like Moraxella and Corynebacterium (Schenk et al., 2016; Liste et al., 2013). These taxa were consistently more abundant in open-water systems compared to non-water-based (EED) or closed water (CON) systems. Such patterns suggest that open water promotes microbial colonization by providing a nutrient-rich environment which, although beneficial for some natural behaviors (O’Driscoll & Broom, 2011), increases potential pathogen exposure. Nevertheless, it is important to highlight that the present study characterized the microbial communities in the production environment itself, and not in the gut or other physiological system, of ducks. Hence, future studies are warranted to examine whether the microbial communities unique to each enrichment modality (e.g. whiffle ball or preening cup) overlap with the microbiota of the ducks’ gut, skin, or other system.

Temporal changes in microbial communities highlight the dynamic interaction between enrichment systems and microbial community development. At baseline, open-water systems exhibited higher abundances of Moraxella, while non-water systems like EED showed reduced levels of opportunistic pathogens (Colton and Fraley, 2014). By 6 h, microbial shifts were apparent, with Pseudomonas dominating in CON and EED treatments, indicating a change in colonization patterns likely influenced by water availability and environmental conditions. These temporal dynamics emphasize the need for frequent monitoring and appropriate management of enrichment systems to mitigate microbial risks (Rodenburg et al., 2005).

Functional profiling revealed that open- and semi-open-water systems initially supported higher metabolic activity, as evidenced by enriched pathways such as amino acid biosynthesis and aerobic respiration (Campbell et al., 2014). However, these differences were not sustained, with pathways becoming more uniform across treatments by 6 h. This suggests that while open-water systems initially enrich distinct microbial activity, environmental factors such as moisture levels and nutrient availability may drive convergence over time, supporting previous findings on environmental impacts in duck production systems (Fraley et al., 2013).

Moreover, several microbial genera identified in each environmental niche have been reported to synthesize neurochemicals that are involved in avian behavior. For example, Corynebacterium spp. and Streptococcus spp. were previously reported to synthesize serotonin de novo (Roshchina et al., 2010). Considering the role of the gut microbiota in affecting behavior in poultry, environmental enrichment or water sources could be leveraged to shape the neuroactive potential of the microbiota to positively affect duck behavior and welfare (Van der Eijik et al., 2020). Although previous studies from our group noted increased feather pecking in ducks provided preening cups (Schober et al., 2023) despite no change in brain serotonergic activity (Bergman et al., 2024), it is warranted to understand how acquisition of microbial species from environmental water-based microbial communities may drive duck peripheral serotonergic signaling to affect other aspects of the central nervous system that drive feather pecking behavior linked to preening cups, or other enrichment modalities. However, in mammals it is known that serotonin cannot cross the blood-brain-barrier and therefore affect behaviors (Chen et al., 2021; Bektas et al., 2020); however, this needs further investigation in birds.

Thus, while open- and semi-open-water enrichment systems promote natural behaviors, they provide exposure to unique microbial communities and pathogen risks compared to non-water-based systems. Factors such as moisture levels, nutrient availability, and environmental management contribute to these outcomes (Schenk et al., 2016). It is still unclear if open water implementation is the best for duck health, production, or welfare. It is possible that implementation of environmental enrichment at a high density could increase resource guarding/aggression. Future research should study the effects of environmental enrichment at different numbers of ducks to enrichment to see if there is a critical number to see improvements in production and welfare but not enough to cause them to resource guard the enrichment. Future studies should also explore strategies to optimize enrichment, such as improved water management (filtration system) and regular cleaning and disinfecting, to balance welfare and production benefits with microbial safety.

Data availability

The data generated in this study can be accessed through the National Center for Biotechnology Information (NCBI) BioProject repository under the accession number PRJNA1212770.

Disclosures

The authors declare no conflict of interest.

Acknowledgments

The USDA is an equal opportunity provider and employer. Mention of trade names or commercial products in this publication is solely for the purpose of providing specific information and does not imply recommendation or endorsement by the U.S. Department of Agriculture. The authors wish to thank Heidi Parnin and Drew Frey of Culver Duck Inc. (Middlebury, IN) for their support as well as Jason Fields and his staff at Purdue ASREC for their excellent care of our ducks. We are also grateful to Sally Reinink and Steve Walcott of Big Dutchman (Holland, MI) for supplying the Pekinos.

References

Ahmed M., Azmat A. Decreased brain serotonin turnover rate following administration of Sharbat-e-Ahmed Shah produces antidepressant and anxiolytic effect in rats. Metab. Brain Dis. 2017;32:1785–1790. doi: 10.1007/s11011-017-0065-6. [DOI] [PubMed] [Google Scholar]
Bektaş A., Erdal H., Ulusoy M., Uzbay I.T. Does seratonin in the intestines make you happy? Turk. J. Gastroenterol. 2020;31(10):721–723. doi: 10.5152/tjg.2020.19554. Oct. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bergman M.M., Schober J.M., Novak R., Greif A., Plue C., Frey D., Parnin H., Fraley G.S. Differences in central dopamine, but not serotonin activity and welfare associated with age but not with preening cup use in commercial grow-out Pekin duck barns.: preening cups may not harm ducks in barns. Poult. Sci. 2024;104542 doi: 10.1016/j.psj.2024.104542. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bergman M.M., Schober J.M., Novak R., Greif A., Plue C., Fraley G.S. Transportation increases circulating corticosterone levels and decreases central serotonergic activity in a sex dependent manner in Pekin ducks. Poult. Sci. 2024;104494 doi: 10.1016/j.psj.2024.104494. [DOI] [PMC free article] [PubMed] [Google Scholar]
Bergman M.M., Schober J.M., Oluwagbenga E.M., Frey D., Parnin H., Fraley G.S. Preening cups in duck housing are associated with an increase in central dopamine activity that suggests negative affective state. Poult. Sci. 2024;103588 doi: 10.1016/j.psj.2024.103588. [DOI] [PMC free article] [PubMed] [Google Scholar]
Biscarini F., Bovenhuis H., van der Poel J., Rodenburg T.B., Jungerius A.P., van Arendonk J.A.M. Across-line SNP association study for direct and associative effects on feather damage in laying hens. Behav. Genet. 2010;40:715–727. doi: 10.1007/s10519-010-9370-0. [DOI] [PubMed] [Google Scholar]
Blokhuis H.J., Wiepkema P.R. Studies of feather pecking in poultry. Vet. Q. 1998;20:6–9. doi: 10.1080/01652176.1998.9694825. [DOI] [PubMed] [Google Scholar]
Campbell C.L., Colton S.., Haas R., Rice M., Porter A., Schenk A., Meelker A., Fraley S.M., Fraley G.S. Effects of different wavelengths of light on the biology, behavior, and production of grow-out Pekin ducks. Poult. Sci. 2014;94:1751–1757. doi: 10.3382/ps/pev166. [DOI] [PubMed] [Google Scholar]
Chen X., Shafer D., Sifri M., Lilburn M., Karcher D., Cherry P., Wakenell P., Fraley S.M., Turk M., Fraley G.S. Centennial review: history and husbandry recommendations for raising Pekin ducks in research or commercial production. Poult. Sci. 2021;100 doi: 10.1016/j.psj.2021.101241. [DOI] [PMC free article] [PubMed] [Google Scholar]
Chen Y., Xu J., Chen Y. Regulation of neurotransmitters by the gut microbiota and effects on cognition in neurological disorders. Nutrients. 2021;13(6):2099. doi: 10.3390/nu13062099. Jun 19. [DOI] [PMC free article] [PubMed] [Google Scholar]
Colton S., Fraley G.S. The effects of environmental enrichment devices on feather picking in commercially housed Pekin ducks. Poult. Sci. 2014;93:2143–2150. doi: 10.3382/ps.2014-03885. [DOI] [PubMed] [Google Scholar]
de Haas E.N., van der Eijk J.A.J. Where in the serotonergic system does it go wrong? Unravelling the route by which the serotonergic system affects feather pecking in chickens. Neurosci. Biobehav. Rev. 2018;95:170–188. doi: 10.1016/j.neubiorev.2018.07.007. [DOI] [PubMed] [Google Scholar]
Dennis R.L., Cheng H.W. The dopaminergic system and aggression in laying hens. Poult. Sci. 2011;90:2440–2448. doi: 10.3382/ps.2011-01513. [DOI] [PubMed] [Google Scholar]
Fraley S.M., Fraley G.S., Karcher D.M., Makagon M.M., Lilburn M.S. Influence of plastic slatted floors compared with pine shaving litter on Pekin Duck condition during the summer months. Poult. Sci. 2013;92:1706–1711. doi: 10.3382/ps.2012-02992. [DOI] [PubMed] [Google Scholar]
Jones B., Blokhuis H., Jong I., Keeling L., McAdie T.M., Preisinger R. Feather pecking in poultry: the application of science in a search for practical solutions. Anim. Welf. 2004;13(suppl. S 13) [Google Scholar]
Jones T.A., Dawkins M.S. Effect of environment on Pekin duck behaviour and its correlation with body condition on commercial farms in the UK. Br. Poult. Sci. 2010;51:319–325. doi: 10.1080/00071668.2010.499143. [DOI] [PubMed] [Google Scholar]
Jones T.A., Waitt C.D., Dawkins M.S. Water off a duck’s back: showers and troughs match ponds for improving duck welfare. Appl. Anim. Behav. Sci. 2009;116:52–57. [Google Scholar]
Kang S.W., Leclerc B., Mauro L.J., El Halawani M.E. Serotonergic and catecholaminergic interactions with co-localized dopamine-melatonin neurons in the hypothalamus of the female turkey. J. Neuroendocrinol. 2009;21:10–19. doi: 10.1111/j.1365-2826.2008.01804.x. [DOI] [PubMed] [Google Scholar]
Kang S.W., Christensen K., Aldridge D.J., Kuenzel W.J. Effects of light intensity and dual light intensity choice on plasma corticosterone, central serotonergic and dopaminergic activities in birds, Gallus gallus. Gen. Comp. Endocrinol. 2020 doi: 10.1016/j.ygcen.2019.113289. [DOI] [PubMed] [Google Scholar]
Karcher D.M., Makagon M.M., Fraley G.S., Fraley S.M., Lilburn M.S. Influence of raised plastic floors compared with pine shaving litter on environment and Pekin duck condition. Poult. Sci. 2013;92:583–590. doi: 10.3382/ps.2012-02215. [DOI] [PubMed] [Google Scholar]
Kim T.-H., Choi J., Kim H.-G., Kim H.R. Quantification of neurotransmitters in mouse brain tissue by using liquid chromatography coupled electrospray tandem mass spectrometry. J. Autom. Methods Manag. Chem. 2014;506870 doi: 10.1155/2014/506870. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kozich J.J., Westcott S.L., Baxter N.T., Highlander S.K., Schloss P.D. Development of a dual-index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the miseq illumina sequencing platform. Appl. Environ. Microbiol. 2013;79:5112–5120. doi: 10.1128/AEM.01043-13. [DOI] [PMC free article] [PubMed] [Google Scholar]
Kuenzel W.J., Masson M. The Johns Hopkins University Press; Baltimore, MD, USA: 1988. In: a Stereotaxic Atlas of the Brain of the Chick (Gallus domesticus) [Google Scholar]
Liste G., Kirkden R.D., Broom D.M. A commercial trial evaluating three open water sources for farmed ducks: effects on water usage and water quality. Br. Poult. Sci. 2013;54:24–32. doi: 10.1080/00071668.2013.763900. [DOI] [PubMed] [Google Scholar]
Lowman Z.S., Parkhurst C.R., Romano J. Effect of nipple lines vs. water trough on Pekin duck breeder performance and well-being. Int. J. Poult. Sci. 2016;15:52–56. [Google Scholar]
Mellen J., MacPhee M.S. Philosophy of environmental enrichment: past, present, and future. Zoo Biol. 2001;20:211–226. [Google Scholar]
Mi J., Wang H., Chen X., Hartcher K., Wang Y., Wu Y., Liao X. Lack of access to an open water source for bathing inhibited the development of the preen gland and preening behavior in Sanshui White ducks. Poult. Sci. 2020;99:5214–5221. doi: 10.1016/j.psj.2020.08.018. [DOI] [PMC free article] [PubMed] [Google Scholar]
Mustedanagic A., Matt M., Weyermair K., Schrattenecker A., Kubitza I., Firth C.L., Loncaric I., Wagner M., Stessl B. Assessment of microbial quality in poultry drinking water on farms in Austria. Front. Vet. Sci. 2023;10 doi: 10.3389/fvets.2023.1254442. [DOI] [PMC free article] [PubMed] [Google Scholar]
Narvaes R., de Almeida R.M.M. Aggressive behavior and three neurotransmitters: dopamine, GABA, and serotonin—A review of the last 10 years. Psychol. Neurosci. 2014;7:601–607. [Google Scholar]
O’Driscoll K.K.M., Broom D.M. Does access to open water affect the health of Pekin ducks (Anas platyrhynchos) Poult. Sci. 2011;90:299–307. doi: 10.3382/ps.2010-00883. [DOI] [PubMed] [Google Scholar]
Rice M., Meelker A., Fraley S.M., Fraley G.S. Characterization of Pekin duck drinking and preening behaviors and comparison when housed on raised plastic versus pine litter flooring. Appl. Poult. Res. 2014;23:737–741. [Google Scholar]
Rodenburg T.B., Bracke M.B.M., Berk J., Cooper J., Faure J.M., Guémené D., Guy G., Harlander A., Jones T., Knierim U., Kuhnt K., Pingel H., Reiter K., Servière J., Ruis M.A.W. Welfare of ducks in European duck husbandry systems. Worlds. Poult. Sci. J. 2005;61:633–646. [Google Scholar]
Roshchina V.V. In: Evolutionary Considerations of Neurotransmitters in Microbial, Plant, and Animal Cells BT - Microbial Endocrinology: Interkingdom Signaling in Infectious Disease and Health. Lyte M., Freestone P.P.E., editors. Springer New York; New York, NY: 2010. pp. 17–52. Pages. Pages. [Google Scholar]
Rychlik I., Karasova D., Crhanova M. Microbiota of chickens and their environment in commercial production. Avian Dis. 2023;67:1–9. doi: 10.1637/aviandiseases-D-22-00048. [DOI] [PubMed] [Google Scholar]
Savory C.J. Feather pecking and cannibalism. Worlds. Poult. Sci. J. 1995;51:215–219. [Google Scholar]
Schenk A., Porter A.L., Alenciks E., Frazier K., Best A.A., Fraley S.M., Fraley G.S. Increased water contamination and grow-out Pekin duck mortality when raised with water troughs compared to pin-metered water lines using a United States management system. Poult. Sci. 2016;95:736–748. doi: 10.3382/ps/pev381. [DOI] [PMC free article] [PubMed] [Google Scholar]
Schober J.M., Merritt J., Swanson M., Tetel V., Oluwagbenga E., Frey D., Parnin H., Erasmus M., Fraley G.S. Preening cups increase apparent wet preening behaviors, but have no impact on other behaviors, body condition, growth, or body morphometrics of grow-out Pekin ducks. Poult. Sci. 2023;103145 doi: 10.1016/j.psj.2023.103145. 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]
van der Eijk J.A.J., Rodenburg T.B., de Vries H., Kjaer J.B., Smidt H., Naguib M., Kemp B., Lammers A. Early-life microbiota transplantation affects behavioural responses, serotonin and immune characteristics in chicken lines divergently selected on feather pecking. Sci. Rep. 2020;10:2750. doi: 10.1038/s41598-020-59125-w. Available at. [DOI] [PMC free article] [PubMed] [Google Scholar]
van Staaveren N., Harlander A. Burleigh Dodds Science Publishing; 2020. Cause and Prevention of Injurious Pecking in Chickens. Understanding Thebehaviour and Improving the Welfare of Chickens; pp. 509–566. [Google Scholar]
Waitt C., Jones T., Dawkins M.S. Behaviour, synchrony and welfare of Pekin ducks in relation to water use. Appl. Anim. Behav. Sci. 2009;121:184–189. [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Data Availability Statement

The data generated in this study can be accessed through the National Center for Biotechnology Information (NCBI) BioProject repository under the accession number PRJNA1212770.

[bib0001] Ahmed M., Azmat A. Decreased brain serotonin turnover rate following administration of Sharbat-e-Ahmed Shah produces antidepressant and anxiolytic effect in rats. Metab. Brain Dis. 2017;32:1785–1790. doi: 10.1007/s11011-017-0065-6. [DOI] [PubMed] [Google Scholar]

[bib0002] Bektaş A., Erdal H., Ulusoy M., Uzbay I.T. Does seratonin in the intestines make you happy? Turk. J. Gastroenterol. 2020;31(10):721–723. doi: 10.5152/tjg.2020.19554. Oct. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0003] Bergman M.M., Schober J.M., Novak R., Greif A., Plue C., Frey D., Parnin H., Fraley G.S. Differences in central dopamine, but not serotonin activity and welfare associated with age but not with preening cup use in commercial grow-out Pekin duck barns.: preening cups may not harm ducks in barns. Poult. Sci. 2024;104542 doi: 10.1016/j.psj.2024.104542. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0004] Bergman M.M., Schober J.M., Novak R., Greif A., Plue C., Fraley G.S. Transportation increases circulating corticosterone levels and decreases central serotonergic activity in a sex dependent manner in Pekin ducks. Poult. Sci. 2024;104494 doi: 10.1016/j.psj.2024.104494. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0005] Bergman M.M., Schober J.M., Oluwagbenga E.M., Frey D., Parnin H., Fraley G.S. Preening cups in duck housing are associated with an increase in central dopamine activity that suggests negative affective state. Poult. Sci. 2024;103588 doi: 10.1016/j.psj.2024.103588. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0006] Biscarini F., Bovenhuis H., van der Poel J., Rodenburg T.B., Jungerius A.P., van Arendonk J.A.M. Across-line SNP association study for direct and associative effects on feather damage in laying hens. Behav. Genet. 2010;40:715–727. doi: 10.1007/s10519-010-9370-0. [DOI] [PubMed] [Google Scholar]

[bib0007] Blokhuis H.J., Wiepkema P.R. Studies of feather pecking in poultry. Vet. Q. 1998;20:6–9. doi: 10.1080/01652176.1998.9694825. [DOI] [PubMed] [Google Scholar]

[bib0008] Campbell C.L., Colton S.., Haas R., Rice M., Porter A., Schenk A., Meelker A., Fraley S.M., Fraley G.S. Effects of different wavelengths of light on the biology, behavior, and production of grow-out Pekin ducks. Poult. Sci. 2014;94:1751–1757. doi: 10.3382/ps/pev166. [DOI] [PubMed] [Google Scholar]

[bib0009] Chen X., Shafer D., Sifri M., Lilburn M., Karcher D., Cherry P., Wakenell P., Fraley S.M., Turk M., Fraley G.S. Centennial review: history and husbandry recommendations for raising Pekin ducks in research or commercial production. Poult. Sci. 2021;100 doi: 10.1016/j.psj.2021.101241. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0010] Chen Y., Xu J., Chen Y. Regulation of neurotransmitters by the gut microbiota and effects on cognition in neurological disorders. Nutrients. 2021;13(6):2099. doi: 10.3390/nu13062099. Jun 19. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0011] Colton S., Fraley G.S. The effects of environmental enrichment devices on feather picking in commercially housed Pekin ducks. Poult. Sci. 2014;93:2143–2150. doi: 10.3382/ps.2014-03885. [DOI] [PubMed] [Google Scholar]

[bib0012] de Haas E.N., van der Eijk J.A.J. Where in the serotonergic system does it go wrong? Unravelling the route by which the serotonergic system affects feather pecking in chickens. Neurosci. Biobehav. Rev. 2018;95:170–188. doi: 10.1016/j.neubiorev.2018.07.007. [DOI] [PubMed] [Google Scholar]

[bib0013] Dennis R.L., Cheng H.W. The dopaminergic system and aggression in laying hens. Poult. Sci. 2011;90:2440–2448. doi: 10.3382/ps.2011-01513. [DOI] [PubMed] [Google Scholar]

[bib0014] Fraley S.M., Fraley G.S., Karcher D.M., Makagon M.M., Lilburn M.S. Influence of plastic slatted floors compared with pine shaving litter on Pekin Duck condition during the summer months. Poult. Sci. 2013;92:1706–1711. doi: 10.3382/ps.2012-02992. [DOI] [PubMed] [Google Scholar]

[bib0015] Jones B., Blokhuis H., Jong I., Keeling L., McAdie T.M., Preisinger R. Feather pecking in poultry: the application of science in a search for practical solutions. Anim. Welf. 2004;13(suppl. S 13) [Google Scholar]

[bib0016] Jones T.A., Dawkins M.S. Effect of environment on Pekin duck behaviour and its correlation with body condition on commercial farms in the UK. Br. Poult. Sci. 2010;51:319–325. doi: 10.1080/00071668.2010.499143. [DOI] [PubMed] [Google Scholar]

[bib0017] Jones T.A., Waitt C.D., Dawkins M.S. Water off a duck’s back: showers and troughs match ponds for improving duck welfare. Appl. Anim. Behav. Sci. 2009;116:52–57. [Google Scholar]

[bib0018] Kang S.W., Leclerc B., Mauro L.J., El Halawani M.E. Serotonergic and catecholaminergic interactions with co-localized dopamine-melatonin neurons in the hypothalamus of the female turkey. J. Neuroendocrinol. 2009;21:10–19. doi: 10.1111/j.1365-2826.2008.01804.x. [DOI] [PubMed] [Google Scholar]

[bib0019] Kang S.W., Christensen K., Aldridge D.J., Kuenzel W.J. Effects of light intensity and dual light intensity choice on plasma corticosterone, central serotonergic and dopaminergic activities in birds, Gallus gallus. Gen. Comp. Endocrinol. 2020 doi: 10.1016/j.ygcen.2019.113289. [DOI] [PubMed] [Google Scholar]

[bib0020] Karcher D.M., Makagon M.M., Fraley G.S., Fraley S.M., Lilburn M.S. Influence of raised plastic floors compared with pine shaving litter on environment and Pekin duck condition. Poult. Sci. 2013;92:583–590. doi: 10.3382/ps.2012-02215. [DOI] [PubMed] [Google Scholar]

[bib0021] Kim T.-H., Choi J., Kim H.-G., Kim H.R. Quantification of neurotransmitters in mouse brain tissue by using liquid chromatography coupled electrospray tandem mass spectrometry. J. Autom. Methods Manag. Chem. 2014;506870 doi: 10.1155/2014/506870. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0022] Kozich J.J., Westcott S.L., Baxter N.T., Highlander S.K., Schloss P.D. Development of a dual-index sequencing strategy and curation pipeline for analyzing amplicon sequence data on the miseq illumina sequencing platform. Appl. Environ. Microbiol. 2013;79:5112–5120. doi: 10.1128/AEM.01043-13. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0023] Kuenzel W.J., Masson M. The Johns Hopkins University Press; Baltimore, MD, USA: 1988. In: a Stereotaxic Atlas of the Brain of the Chick (Gallus domesticus) [Google Scholar]

[bib0024] Liste G., Kirkden R.D., Broom D.M. A commercial trial evaluating three open water sources for farmed ducks: effects on water usage and water quality. Br. Poult. Sci. 2013;54:24–32. doi: 10.1080/00071668.2013.763900. [DOI] [PubMed] [Google Scholar]

[bib0025] Lowman Z.S., Parkhurst C.R., Romano J. Effect of nipple lines vs. water trough on Pekin duck breeder performance and well-being. Int. J. Poult. Sci. 2016;15:52–56. [Google Scholar]

[bib0026] Mellen J., MacPhee M.S. Philosophy of environmental enrichment: past, present, and future. Zoo Biol. 2001;20:211–226. [Google Scholar]

[bib0027] Mi J., Wang H., Chen X., Hartcher K., Wang Y., Wu Y., Liao X. Lack of access to an open water source for bathing inhibited the development of the preen gland and preening behavior in Sanshui White ducks. Poult. Sci. 2020;99:5214–5221. doi: 10.1016/j.psj.2020.08.018. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0028] Mustedanagic A., Matt M., Weyermair K., Schrattenecker A., Kubitza I., Firth C.L., Loncaric I., Wagner M., Stessl B. Assessment of microbial quality in poultry drinking water on farms in Austria. Front. Vet. Sci. 2023;10 doi: 10.3389/fvets.2023.1254442. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0029] Narvaes R., de Almeida R.M.M. Aggressive behavior and three neurotransmitters: dopamine, GABA, and serotonin—A review of the last 10 years. Psychol. Neurosci. 2014;7:601–607. [Google Scholar]

[bib0031] O’Driscoll K.K.M., Broom D.M. Does access to open water affect the health of Pekin ducks (Anas platyrhynchos) Poult. Sci. 2011;90:299–307. doi: 10.3382/ps.2010-00883. [DOI] [PubMed] [Google Scholar]

[bib0033] Rice M., Meelker A., Fraley S.M., Fraley G.S. Characterization of Pekin duck drinking and preening behaviors and comparison when housed on raised plastic versus pine litter flooring. Appl. Poult. Res. 2014;23:737–741. [Google Scholar]

[bib0034] Rodenburg T.B., Bracke M.B.M., Berk J., Cooper J., Faure J.M., Guémené D., Guy G., Harlander A., Jones T., Knierim U., Kuhnt K., Pingel H., Reiter K., Servière J., Ruis M.A.W. Welfare of ducks in European duck husbandry systems. Worlds. Poult. Sci. J. 2005;61:633–646. [Google Scholar]

[bib0035] Roshchina V.V. In: Evolutionary Considerations of Neurotransmitters in Microbial, Plant, and Animal Cells BT - Microbial Endocrinology: Interkingdom Signaling in Infectious Disease and Health. Lyte M., Freestone P.P.E., editors. Springer New York; New York, NY: 2010. pp. 17–52. Pages. Pages. [Google Scholar]

[bib0036] Rychlik I., Karasova D., Crhanova M. Microbiota of chickens and their environment in commercial production. Avian Dis. 2023;67:1–9. doi: 10.1637/aviandiseases-D-22-00048. [DOI] [PubMed] [Google Scholar]

[bib0037] Savory C.J. Feather pecking and cannibalism. Worlds. Poult. Sci. J. 1995;51:215–219. [Google Scholar]

[bib0038] Schenk A., Porter A.L., Alenciks E., Frazier K., Best A.A., Fraley S.M., Fraley G.S. Increased water contamination and grow-out Pekin duck mortality when raised with water troughs compared to pin-metered water lines using a United States management system. Poult. Sci. 2016;95:736–748. doi: 10.3382/ps/pev381. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0039] Schober J.M., Merritt J., Swanson M., Tetel V., Oluwagbenga E., Frey D., Parnin H., Erasmus M., Fraley G.S. Preening cups increase apparent wet preening behaviors, but have no impact on other behaviors, body condition, growth, or body morphometrics of grow-out Pekin ducks. Poult. Sci. 2023;103145 doi: 10.1016/j.psj.2023.103145. 2023. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0040] van der Eijk J.A.J., Rodenburg T.B., de Vries H., Kjaer J.B., Smidt H., Naguib M., Kemp B., Lammers A. Early-life microbiota transplantation affects behavioural responses, serotonin and immune characteristics in chicken lines divergently selected on feather pecking. Sci. Rep. 2020;10:2750. doi: 10.1038/s41598-020-59125-w. Available at. [DOI] [PMC free article] [PubMed] [Google Scholar]

[bib0041] van Staaveren N., Harlander A. Burleigh Dodds Science Publishing; 2020. Cause and Prevention of Injurious Pecking in Chickens. Understanding Thebehaviour and Improving the Welfare of Chickens; pp. 509–566. [Google Scholar]

[bib0043] Waitt C., Jones T., Dawkins M.S. Behaviour, synchrony and welfare of Pekin ducks in relation to water use. Appl. Anim. Behav. Sci. 2009;121:184–189. [Google Scholar]

PERMALINK

Effects of semi-open water and non-water enrichment on welfare, production, behavior, and microbial exposure of grow-out Pekin ducks

JM Schober

N Wilson

MM Seyoum

JM Lyte

MM Bergman

EM Oluwagbenga

GS Fraley

Abstract

Introduction

Materials and methods

Animals

Enrichment placement

Morphometrics and body condition scores

Litter and ammonia

Behavior analyses

Table 1.

Brain microdissections

Mass spectrometry

Viable bacteria collection and analysis

Table 2.

Sample collection for 16S rRNA gene sequencing

DNA extraction and sequencing

Bioinformatic analyses of 16S rRNA gene sequencing data

Functional profiling of microbial communities using PICRUSt2 analysis

Statistical analyses

Morphometrics, environment and FCR

Body condition scores

Behavior analyses

16S rRNA analyses

Results

Weekly body weights

Fig. 1.

FCR and feed intake

Fig. 2.

Dissection

Fig. 3.

Body condition scores

Feather quality

Table 3.

Feather cleanliness

Table 4.

Foot pad scores

Table 5.

Eye health

Table 6.

Nostril scores

Table 7.

Litter moisture and ammonia

Fig. 4.

Behavior

Wet preening

Fig. 5.

Dry preening

Preening others

Feather pecking

Feather picking

Drinking from the nipple line

Eating

Standing

Laying down

Drinking from semi-open water source

Dunking head in semi-open water source

Interacting with enrichment/dunking head in semi-open water

Drinking from nipple line/drinking from water enrichment

Interacting with EED

Mass spectrometry

Viable bacteria

Fig. 6.

Microbial diversity and community composition

Fig. 7.

Fig. 8.

Fig. 9.

Functional pathways

Fig. 10.

Discussion

Data availability

Disclosures

Acknowledgments