Preference and behavioral responses to synbiotic supplementation via drinking water in laying hens under social stress

Abstract

Synbiotics (SYN) have the potential to enhance animal health, but their efficacy depends on both biological impact and voluntary intake, particularly under stress, when hydration and gut support are critical. This study presents the first investigation of SYN supplementation delivered exclusively through drinking water in laying hens, evaluating both preference and behavioral responses, particularly feather pecking, during a period of social stress induced by repeated mixing of unfamiliar birds. A total of 226 White Leghorn hens (37 weeks old) were housed in enriched floor pens with simultaneous access to color-coded SYN-enriched and plain water containers. Following a 2-week associative learning phase and 1-week washout, hens entered a 6-week preference testing period encompassing pre-stress, stress, and post-stress phases. Water consumption was measured at the group level, while individual jug visits were tracked using RFID technology. Feather pecking was measured (10 min/day) and feather damage assessed according to severity. Hens showed a clear preference for SYN-enriched water, consuming significantly more than plain water (p < 0.0001). While overall intake remained stable, SYN consumption fluctuated across stress phases, with the strongest preference pre-stress and a modest decline during stress. Although hens with higher rates of severe feather pecking (SFP) visited SYN-enriched water more frequently than hens with lower rates (p = 0.0288), suggesting a potential coping mechanism, overall, SFP rates remained stable across all phases. Notably, the level of SFP observed during the pre-stress phase was already sufficient to cause progressive plumage deterioration, which continued throughout the study. The proportion of birds with severe feather damage rose from 39.4% to 53.5%, while those with intact plumage dropped from 37.6% to 19.9% over time. Thus, SYN supplementation via drinking water is feasible and preferred by laying hens, even under stress. These findings highlight the potential of nutraceuticals delivered through drinking water to influence hen behavior; however, the progressive plumage deterioration despite stable SFP rates suggests that SYN supplementation alone may be insufficient, underscoring the need for integrated management strategies.

Introduction

Chronic stress reliably alters feeding motivation across species and exerts complex effects on behavior and brain function, often increasing the consumption of palatable, high-calorie solid or liquid foods, such as sucrose, as a short-term coping strategy, which can temporarily suppress stress responses and promote emotional relief (Ulrich-Lai et al., 2015). However, prolonged or severe stress can blunt reward sensitivity and may lead to anhedonia, a reduced ability to experience pleasure, often reflected in decreased sucrose intake in animal models and considered a hallmark of depressive-like states (Papp et al., 1991).

Critically, stress‑linked changes in feeding are not driven by taste alone but are mediated by the microbiota–gut–brain axis. The gut microbiota, a dynamic ecosystem within the gastrointestinal tract, modulates brain function through microbially derived metabolites and neuroendocrine, immune, and vagal pathways (Forsythe et al., 2014). The gut–brain axis also contributes to sugar preference, underscoring a bidirectional link between feeding behavior and microbial signaling (Tan et al., 2020). Disruptions to the gut microbiota, due to stress, diet, medication, or age, have been linked to cognitive, emotional, and behavioral impairments. While microbiome-targeted interventions show promise in preclinical models, definitive causal links to neurological disease trajectories remain under investigation (Bravo et al., 2011; Forsythe et al., 2016).

Notably, modest sucrose intake can reduce stress‑related behaviors without affecting body weight in rodents, indicating that even small amounts of palatable intake can modulate behavioral and physiological stress (Ulrich-Lai et al., 2015). Together, these findings suggest that nutritionally minimal “comfort feeding” may influence stress-related behavior through gut-brain signaling (Tan et al., 2020). Although this has not yet been studied in birds, these findings point to a feedback loop in which stress alters eating behavior and food preferences (in particular towards sweet items) and disrupts the gut microbiota, both of which, in turn, influence the stress response via the gut-brain axis.

In laying hens, severe feather pecking (SFP) is often triggered by adverse stressors (Brunberg et al., 2016), such as the loss of social bonds and poor environmental conditions (Van Staaveren And Harlander, 2020). Gentle feather pecking (GFP) can also occur and involves repeated, mild pecks at the feathers that may constitute a form of social exploration or allopreening (Riedstra and Groothuis, 2002; Rodenburg et al., 2004). From an ethological perspective, birds performing SFP redirect foraging behavior toward conspecifics’ plumage (Harlander-Matauschek and Bessei, 2005; McKeegan and Savory, 1999; ) as a stress-induced coping mechanism. Although feathers are nutritionally indigestible, their ingestion as part of SFP behavior may act as a form of comfort feeding, altering gut motility and increasing feed passage time (Harlander-Matauschek et al., 2006a, Harlander-Matauschek et al., 2006b). Furthermore, peckers and non-peckers have distinct gut microbiota compositions. Differences include a lower abundance of Lactobacillus in peckers (Birkl et al., 2018; Huang et al., 2023; Van Der Eijk et al., 2019), as well as shifts in bacterial metabolites, such as short-chain fatty acids (Meyer et al., 2013), immune markers (Mindus et al., 2022; Van Der eijk et al., 2020), neurotransmitters and their precursors (Birkl et al., 2017; De Haas and van der Eijk, 2018; Mindus et al., 2021c), all of which modulate stress responses and emotional regulation. Collectively, these findings suggest that the gut-brain axis plays an important role in the development and expression of severe feather pecking (Chen et al., 2022; Mindus et al., 2021b; Van Staaveren et al., 2021).

Given this connection between gut function and behavioral regulation (Mindus et al., 2021a), dietary interventions that modulate the microbiota-gut-brain-axis, such as the use of synbiotics, offer a potential strategy to mitigate SFP. Synbiotics combine prebiotics (e.g., fructooligosaccharides [FOS] and galactooligosaccharides [GOS]) with probiotics (e.g., Lactobacillus and Bifidobacterium species) to promote beneficial microbial populations (Johnson et al., 2024; Mohammed et al., 2019; Richards et al., 2024). Prebiotics like FOS are mildly sweet (approximately 40% as sweet as sucrose; Franck, 2002) and are used in food products to improve palatability, while delivering prebiotic benefits (Na et al., 2023). In both human and animal nutrition, FOS may increase palatability by acting as comfort agents, influencing gut-brain communication through neural, immune, and endocrine pathways. Although chickens lack the canonical TIR2 sweet receptor, they are capable of detecting and responding to sweet compounds via T1R2-independent pathways (Higashida et al., 2022), suggesting that sweetness can contribute to voluntary intake and behavioral modulation.

In commercial settings, farmers are, at times, required to use alternative strategies, such as synbiotics, to manage stress-related outcomes, including disease or social instability. There is also increased interest in the use of non-invasive functional feed items to treat animal welfare issues. For such interventions to be effective in practice, they must be voluntarily self‑administered, particularly under stress when preferences and motivation to drink/eat can shift. It is, therefore, essential that ingestible interventions are both effective and palatable, providing positive post-ingestive feedback to support ease of administration and the ultimate welfare of the birds.

Accordingly, this study tests the relative preference of laying hens for synbiotic‑enriched (SYN) water versus plain water during chronic, unpredictable social stress induced by mixing unfamiliar birds. The SYN combines mildly sweet prebiotics with probiotics, including but not limited to Lactobacillus, which is deficient in the gut of feather peckers. We formulated two hypotheses: 1. If SYN water provides birds with positive post-ingestive feedback, birds will develop a relative preference for SYN over plain water 2. Exposure to social stress will modulate SYN intake, either increasing it, if birds seek its palatability or comforting properties, or decreasing it, consistent with stress-induced anhedonia reported in other species. Using a choice-based approach, we investigated whether individual variation in SYN preferences was associated with feather pecking behavior observed via video recordings and plumage damage scores during pre-stress, stress, and post-stress periods. This is the first study to evaluate voluntary SYN consumption under social stress in laying hens, addressing a critical gap in welfare-focused nutritional science.

Animals, materials and methods

Ethical statement

This experiment was approved by the Animal Care Committee at the University of Guelph under Animal Utilization Protocol #4113 and carried out 2021.

Animals and housing

A pedigree line of White Leghorn laying hens divergently selected for high (HFP) and low (LFP) severe feather pecking (SFP) activity were used for this study (Kjaer et al., 2001). Eggs from HFP and LFP birds were placed in separate compartments of the incubator to hatch separately. At hatch, 226 non-beak trimmed chicks were individually wing-tagged (5 × 17 mm; Ketchum, Brockville, Canada) and allocated to 10 identical rearing pens of 22 ± 1 birds of the same line. In accordance with standard rearing practices for these lines at the research station (Birkl et al., 2019a; Mindus et al., 2021b), each rearing pen had a brooder (98.5 × 73 × 39 cm) with a heating pad, which was turned on during the first week of age. The entrance to the brooder was covered by white curtains to allow free movement while providing complete darkness. Pen floors (4.7m²) were littered with wood shavings and contained one round metal feeder (43 Ø cm) and a drinker line (7 nipples). Each pen was fitted with 5 perches: a small round perch (87 × 1 cm (L x Ø); 5 cm off the ground), two small round perches extending past the brooder entrance (74 × 2.5 cm Ø; 15 cm off the ground) and two perches, 15 cm perch length/bird at 55 and 120 cm above the ground. Visual contact between pens was prevented by opaque PVC boards, but auditory contact was not blocked. Birds had ad libitum access to water and feed. Commercial rearing lighting conditions, ventilation and vaccination schedules were followed. Birds were fed a 21% protein pullet starter diet from 0 to 6 weeks of age (WOA) and an 18% protein pullet grower diet from 7 to 16 WOA.

At 17 WOA, birds were moved into the layer barn. The LFP and HFP birds were balanced evenly between 12 identical enriched floor pens (i.e., 18-19 birds per pen; 8-9 HFP and 10-11 LFP birds per pen). Each pen (186 cm length (L) x 252 cm width (W) x 284 cm height (H); floor space: 4.7 m²) contained a steel nest box (74 cm L x 45 cm W x 59 cm H) with three compartments elevated by a wooden base, two perches (high narrow perch: 159 cm L x 33 cm W at 123 cm H; low wide perch: 159 cm L x 64 cm W at 64 cm H), a round metal feeder (inner diameter: 23 cm; outer diameter: 36 cm) hanging on a digital scale (WeiHeng Manufacturer, Guangdong, China) and a waterline with 7 nipples (54 cm L at 55 cm height). The floors were covered with 5 cm of wood shavings (Fig. 1). One camera (SNO-5084RN, IR, Samsung, Techwin Co., Ltd., Gyeonggi-do South Korea) was mounted to record aerial footage of each pen. Pens were separated by metal partitions with solid panels up to a height of 75 cm to prevent visual contact between hens at ground level. The set-up allowed for the transmission of sounds and odors between pens. Birds were fed an 18% protein poultry layer-breeder diet starting at 17 WOA (Floradale Feed Mill Ltd., Floradale, Canada). Birds had ad libitum access to water and feed.

Fig. 1.

The experimental timeline showing the pre-supplementation (36 WOA), 3-week associative learning (37-39 WOA), and 6-week preference testing (plain versus synbiotic-enriched (SYN) water, 40-45 WOA) phases.

Synbiotic-enriched (SYN) versus plain water set-up

From 37 to 45 WOA, the waterline was removed and water was provided in 11.36 liter hanging jugs with four nipples (Poultry Waterer, Four Nipples, Millside Industries Inc., Wallenstein, Canada) adapted for the study (Fig. 2). Jugs were color-coded with blue or yellow duct tape (Duck Heavy Duty Duct Tape, Shurtape Technologies, LLC., Ohio, United States) so that hens could associate un-supplemented plain and synbiotic-supplemented (SYN) water jugs with a specific color in case birds could not detect the flavor or chemical properties of the synbiotic. Blue and yellow were chosen as chickens have been shown to distinguish between these colors (Bona et al., 2018; Jones et al., 2001). Birds used in this study did not have previous interactions with these colors. To eliminate color bias, half of the pens received SYN water in yellow jugs and the other half in blue jugs. The color assignment was consistent throughout the experiment. To avoid contamination, each jug was assigned to only hold SYN or plain water. The location of the two jugs inside each pen (i.e., closer to the wall or the aisle) varied daily in a semi-randomized pattern to allow birds to recognize which jug contained SYN while preventing a side bias.

Fig. 2.

A schematic representation (A) and a photograph (B) of the pen layout. Each pen included a steel nest box with 3 compartments elevated with a wooden base, two perches (high narrow perch and low wide perch), a round metal feeder, a surveillance camera on a wooden mount and two water containers that were installed from 37 to 45 WOA (i.e., during associative learning and preference testing periods).

Preparation of synbiotic-enriched (SYN) and plain water

A commercial multi-species synbiotic powder was used in the present study. The synbiotic contained probiotic microorganisms (Gram-positive Bifidobacterium animalis, and lactic acid bacteria, including Enterococcus faecium, Pediococcus acidilactici, Lactobacillus reuteri and Lactobacillus salivarius), prebiotic fructooligosaccharides, and silicon dioxide. Bacterial counts were performed by Gelda Scientific, 6320 Northwest Dr., Mississauga, Canada. The synbiotic was confirmed to contain a minimum viable bacterial count of 2.0 × 10¹¹ Colony Forming Units/kg. Based on the manufacturer’s suggestion (20 mg of synbiotic/hen/day) and a conservatively estimated average water intake of ∼ 250 ml per day per hen, a dose of 80 mg of synbiotic per litre of water was used.

Prior to use, the synbiotic was ground into smaller particles for 20 seconds using a Cuisinart Food Processor. Then, 400 mg of the resulting powder was weighed into 1.5 ml microcentrifuge tubes using an analytic balance (Mettler Toledo Analytic Balance, Model XSR105DU, Mississauga, Canada) and stored in a cool and dry place until use.

To prepare SYN water jugs, 500 ml lukewarm water (i.e., 36.7 to 40.6°C) was poured into a 700 ml Oster® blending cup and mixed with an immersion blender (Oster® Detachable Hand Blender with Blending Cup 2611-33A, Brampton, Canada) for 10 seconds. Then, ∼400 mg of the synbiotic powder was added and solubilized with the immersion blender for another 10 seconds. In a separate 5 L pitcher, 2,700 ml hot and 1,800 ml cold tap water was mixed using a wire whisk. Then, the 500 ml SYN solution was poured into the 5 L pitcher and mixed for 15 seconds. Once well blended, the SYN solution was weighed on a bench scale accurate to 0.01 kg (Ohaus® Defender 3000 Bench Scale, D31300BX, Parsippany, United States), poured into its corresponding jug, and hung inside the pen. The water for plain water jugs was prepared following the same protocol but without adding SYN to the lukewarm water.

SYN and plain water jugs were filled 5-times a week (Monday to Friday). From Monday to Thursday, jugs were filled with 5 L. On Fridays, 10 L were added to each jug to provide water over the weekend.

RFID system

Radio frequency identification (RFID) technology was used to record the number and duration of visits to plain and SYN water jugs by individual hens. The components of the RFID system used is listed in Table 1.

Table 1.

Components of the RFID system used to record water consumption behavior.

RFID System Component	Number of Units	Specifications	Supplier
Passive integrated transponder (PIT) tags	226	HPT 12 PIT tags, 12.5 mm, 134.2 kHz	Biomark®, Idaho, United States
Antennas	12	30 cm L x 10 cm W x 2 cm H,	Biomark®, Boise, Idaho, United States
Multiplex RFID reader unit	1	-	Biomark®, Boise, Idaho, United States
Power supply	1	3-12 V 2AMP Regulated DC Power Supply	Parts Express, Ohio, United States
Handheld PIT tag reader	1	601 Handheld Reader	Biomark®, Boise, Idaho, United States

Each bird was assigned a unique passive integrated transponder (PIT) tag encased in Gorilla tape (The Gorilla Glue Company, Ohio, United States). The PIT tag corresponded to the hen’s unique wing tag provided at hatch and the backpack. The PIT tags were glued to leg bands (BCP 16 mm Poultry Leg Bands, Ring Size 7, United Kingdom) that were padded with strips of foam for comfort.

The RFID system’s reader unit collected and stored daily recordings from all 12 antennas used in the study on an external hard drive. The reader unit was turned off every morning during daily tasks inside the pens. The raw data collected included the antenna ID that the data was collected from, date, time and the PIT tag ID, which corresponded to the hen’s backpack ID (used to monitor pecking behavior; see Section 2.8) and their unique wing tag ID (Fig. 3). Each timestamp represented one millisecond when a PIT tag was in the range of the antenna’s field. All 12 antennas were plugged into the multiplex RFID reader connected to a power supply unit during daily recordings. The microchip of the reader unit stored data from antennas, which was downloaded daily using the Biomark® software. A Perl script (Garant et al., 2022) was used to translate and summarize Biomark® output into Excel files (CSV format) showing jug visit behaviors for each hen. Visits to the same jug by the same hen were only counted as two separate visits if they occurred 30,000 milliseconds (30 seconds) apart (Garant et al., 2022).

Fig. 3.

Radio-frequency identification (RFID) technology overview. (A) & (B) Unique PIT tags were glued into birds’ leg bands that were scanned using a handheld PIT tag reader. (C) & (D) Antennas generating a radio frequency field were placed inside containers underneath the water jugs. Antennas picked up PIT tag signals when birds stepped onto containers to visit water jugs (E) & (F).

Water jug apparatus set-up

The water jugs were customized so that only a single bird could access the drinker at a time, ensuring that only the PIT tags of birds inside the jug apparatus was detected. To this end, three out of the four nipples were blocked off by taping clear plastic vials over the nipples. Color-coded partitions (2 boards of 23 cm L x 38 cm H with an opening of 17-19 cm) that match the jug color were taped to the bottom of the hanging jugs to surround the single accessible nipple, creating a compartment that fits a single hen (Fig. 4). Each hanging jug had a clear, plastic step-up container with a lid (1.5 L Storage Box, 37 cm L x 15 cm W x 8 cm H, Dollarama, Montreal, Canada) that was attached to the bottom of the partition using zip-ties and s-hooks (1″ Zinc Plated S-Hooks, Hillman, Pickering, Canada). The container was lined with shelf liner padding and housed an RFID antenna (∼10 cm range). The antenna was elevated within the container using a rectangular prism made of whiteboard to ensure that the PIT tag signal was captured when hens stepped onto the container (Garant et al., 2022). Pilot testing was conducted prior to data collection to confirm that the partitions restricted access to a single hen and that antenna detection performance was reliable under the final layout. The RFID system therefore recorded a hen’s presence at the drinking position (i.e., jug visit), but did not measure whether water was consumed during that visit. To eliminate radio-frequency interference between antennas, the containers were placed ∼60 cm apart (measured between the centers of the containers). Each pen had two jug apparatuses (SYN and plain water) that were hung from the ceiling.

Fig. 4.

Representative images of the back and rump of a hen with (A) a score of 0 for no feather damage, (B) a score of 1 for moderate wear or featherless area smaller than 2.8 cm in diameter, and (C) a score of 2 for severe damage or featherless area larger than 2.8 cm in diameter.

Water consumption recordings

From 37 to 45 WOA, the water consumption per pen was measured each morning from Monday to Friday. To this end, jugs were weighed using a compact scale, accurate to 0.001 kg (Ohaus® Valor® 1000, V11P6T). The jug weight at the end of a 24 h period was subtracted from the jug weight at the beginning of the 24 h period (i.e. immediately after it was filled for the day).

Due to equipment constraints, RFID data collection was limited to six pens at a time. Between 37 and 45 WOA, the first group was recorded from Monday evening to Wednesday morning, and the second group from Wednesday evening to Friday morning. To minimize interference from human activity and ensure consistency, only data recorded during specific light-cycle hours were retained. Specifically, recordings from 17:00 to 19:29 h on Day 1 and from 05:00 to 08:59 h on Day 2 were extracted from the raw Biomark® files. Therefore, each pen was recorded for 6.5 hours per week over a 9-week period (37–45 WOA), resulting in a total of 58.5 hours of RFID data per pen and 702 hours across all 12 pens.

Files were processed using a custom-built Perl script (Perl Version 5.36.0) executed in the macOS Terminal (Terminal Version 2.13 (447)). The following outputs were calculated for each hen: (1) the total time (milliseconds) spent at each jug, (2) the average duration (milliseconds) of visits to each jug, and (3) the number of visits to each jug (Garant et al., 2022). A threshold of 30 seconds was used to differentiate between two separate visits to a jug by the same bird (Garant et al., 2022). Thus, the duration and frequency of visits to each jug, but not the behavior (e.g., drinking, not drinking) were measured via RFID.

Feather pecking behavior observations and feather damage scoring

From 36 to 52 WOA, numbered silicon backpacks were used to identify hens, as described in Birkl et al. (2018, 2019b). The backpack number corresponded to a unique wing tag, leg band and PIT tag ID during the entirety of the trial. The incidence of gentle feather-pecking (GFP) and severe feather-pecking (SFP) behaviors was documented from video recordings collected when hens were 40 to 45 WOA (Fig. 1). During this period, daily 15-minute recordings were taken from Monday to Friday at 09:00 h, 11:30 h, 15:00 h, and 16:00 h. Personnel were not allowed to enter the room during and 15 minutes prior to the recordings. Videos were trimmed to 10-minute clips for behavioral analyses.

Pecking behavior was classified as described in van Staaveren and Harlander (2020). The actors of GFP and SFP were identified for each ten-minute recording per pen. Actors were classified as hens that performed the pecking behavior. GFP was described as repetitive, gentle pecks directly towards the tips of feathers of conspecifics without pulling, in continuous bouts that lasted longer than four seconds. SFP was described as an instantaneous forceful peck directed towards the feathers and involved pulling, removing and/or ingesting feathers (van Staaveren and Harlander, 2020).

From these behavioral observations, hens that exhibited the SFP phenotype were categorized into low peckers, described as hens that pecked at equal to or lower than the median SFP rate per period, and high peckers, described as hens that pecked at a rate greater than the median SFP rate per period. Therefore, the designations as low and high peckers reflect only the phenotype of the birds without considering the genotype.

In addition to the video recordings, each bird was evaluated for feather cover damage by a trained assessor who was blinded to the treatment. Feather cover was scored using a validated 3-point scale (Decina et al., 2019), where: 0 indicated no damage or slight wear; 1 denoted moderate wear, including deformed or worn feathers or at least one featherless area smaller than 2.8 cm in diameter (approximating the size of a Canadian $2 coin); and 2 reflected severe damage, with at least one featherless area exceeding 2.8 cm (Fig. 4) (Fig. 4). Four areas of the body were examined: neck, back and rump, belly, and tail.

Chronic social stress treatment

To induce SFP, hens were exposed to repeated, unpredictable social stress/mixing on seven random days (Birkl et al., 2019; Mindus et al., 2021b) during the 3-week stress period (41 to 43 WOA). Across the 3‑week stress period, hens experienced seven unannounced mixing events, each designed to disrupt established social hierarchies. On each mixing day, all birds within a color‑coded treatment group (i.e., all hens assigned to the yellow or blue water‑jug condition) were caught, removed from their home pens, and redistributed into new temporary groups. Hens were intentionally mixed so that no bird remained with its full original pen‑mates and so that each temporary group contained birds that had not previously been housed together. This ensured that every hen encountered multiple unfamiliar individuals at each event. For each mixing event, birds were drawn from all pens belonging to the same treatment color, and new social groups were assembled by randomly combining 3–6 hens from each source pen, depending on availability. This procedure maximized the number of novel social encounters while maintaining treatment integrity (i.e., SYN-blue or plain-yellow comparisons were preserved).

Bodyweight and feed consumption measurements

All hens were weighed (Ohaus® Valor® 1000 scale, V11P6T) at 36 WOA (pre-supplementation), 39 WOA (end of the associative learning period), 43 WOA (end of stress period), 45 WOA (end of post-stress period) and 52 WOA (post-supplementation, long-term).

From 37 to 45 WOA, daily feed consumption per pen was measured using digital crane hook hanging scales (accuracy ±0.05 kg; WeiHeng Manufacturer, Guangdong, China). The amount consumed was calculated by subtracting the pre-fill weight from the previous day’s post-fill weight.

Statistical analysis

The SAS software (SAS® Studio Version 3.81, Enterprise Edition, SAS Institute Inc., Cary, North Carolina) was used for all statistical computations. Unless specified differently, generalized linear mixed models (PROC GLIMMIX) were used to analyze the data. The assumptions of normally distributed residuals and homogeneity of variance were examined graphically with the use of QQ plots. Scatter plots of studentized residuals against predicted values and treatment values, and a Shapiro-Wilk test of normality were used to confirm the assumptions of the variance analysis. To detect possible outliers, studentized residuals outside a ± 3.4 envelope were used. Data was transformed where necessary. Least square (LS) means and standard errors on the data scale were recovered using the ilink option when non-Gaussian distributions were used. Values are presented as LS means ± standard error, unless stated otherwise. Differences between means were compared pairwise using a Tukey-Kramer adjustment. Statistical significance was considered at P < 0.05.

To assess hens’ liquid intake (ml/bird/day) preference during pre-stress, stress, and post-stress periods, PROC GLIMMIX with a Gaussian distribution was used to assess the fixed effects of liquid type (i.e., SYN and plain water), period (i.e., pre-stress, stress, and post-stress), and their interactions. Pen was included as a repeated effect to account for repeated measurements within the same pen across periods.

The number of GFP and SFP performed by actors per 10-minute recordings was calculated per pen per period. To find the relationship between the average number of GFP and SFP bouts, and average daily SYN water consumption at the group-level, PROC GLIMMIX with Gaussian distributions were used to assess the fixed effect of period, SYN water consumption (ml/bird/day), and their interaction on the number of pecks with a random pen effect.

RFID recordings for total time spent at jugs, average duration of visits and the number of visits to SYN water jugs was calculated per period per hen. Any hens that did not have RFID recordings for one or more periods, either due to missing recordings or total durations being less than 1 minute, were removed from the dataset. A subset of 199 hens from the 226 hens were used for the analyses. For SFP, the average weekly severe feather pecks per daily 10-minute video recordings per individual hen was calculated per period. The median SFP was calculated for each period, and each hen was categorized as a low pecker or high pecker within said period. PROC GLIMMIX with lognormal distributions were used to assess the fixed effects of SFP category (low or high pecker), period, and their interaction on the (1) total visit duration (minutes), (2) average visit duration (minutes), and (3) number of visits to SYN water jugs, with a random effect of hen accounting for repeated measurements on the same hens in each period.

The maximum body feather score was determined by identifying the highest score out of the four body region scores per individual bird in each period. PROC GLIMMIX with multinomial distributions were used to assess the fixed effect of period on feather cover scores (i.e., scores for each individual region and highest body scores). The odds of having lower feather damage were modelled and presented as odds ratios (OR) with associated 95% confidence intervals (CI). PROC GLIMMIX with a Gaussian distribution was used to assess the fixed effect of period, daily SYN consumption, and their interaction on daily feed consumption/bodyweight with a random pen effect.

Results

Hens prefer SYN water across all periods, but social stress modulates intake patterns

Hens showed a significant preference for synbiotic-enriched (SYN) water over plain, un-enriched water, consuming 78.2 ± 1.33 ml per bird per day of SYN water compared to 68.4 ± 1.33 ml per bird per day of plain water over the entire duration of the experiment (F₁,₇₀₂ = 38.57, p < 0.0001).

However, there was a trend toward an interaction between liquid type (SYN or plain) and period, as the magnitude of preference for SYN water tended to change over time (F_{2, 702} = 2.89, p = 0.0562). Indeed, hens consistently consumed more SYN than plain water, but the difference varied by period. The highest difference was observed in the pre-stress period, followed by a reduction in SYN water consumption in the stress period and a slight increase in the post-stress period (Table 2). While SYN water intake fluctuated across periods (with a trend between Pre-Stress and Stress: t₇₀₂ = 2.72, p = 0.0723), plain water intake remained stable (all p > 0.9), contributing to the observed interaction trend. In total, hens consumed an average of 149.7 ± 3.7 ml, 143.8 ± 2.5 ml, and 146.4 ± 2.8 ml liquid per bird per day during the pre-stress, stress, and post-stress periods, respectively (P > 0.05).

Table 2.

Average (LSM ± SE) synbiotic-enriched (SYN) and plain water consumption (mL per hen per day) in laying hens during a pre-stress (40 WOA), stress (41-43 WOA) and post-stress period (44-45 WOA) in a choice test.

Period	SYN Water Consumption (mL per hen per day ± SE)	Plain water Consumption (mL per hen per day± SE)	SYN vs Plain Difference (mL)	t₇₀₂	P-value
Pre-Stress	82.7 ± 2.58	67.0 ± 2.58	15.7	4.46	0.0001
Stress	74.9 ± 1.60	69.0 ± 1.60	5.9	2.91	0.0433
Post-Stress	77.2 ± 1.89	69.2 ± 1.89	8	3.22	0.0170

Gentle and severe feather pecking behaviors are stable and unaffected by SYN

The number of gentle feather pecks remained stable across all periods, with 0.04 ± 0.01 pecks/bird/10 minutes during the Pre-Stress period and 0.03 ± 0.01 pecks/bird/10 minutes during both the Stress and Post-Stress periods, with no significant effects of period (F _2,256.7 = 0.18, p = 0.8377), liquid type (F_1,256 = 1.00, p = 0.3194), or their interaction (F _2,256.7 = 0.11, p = 0.8940).

Overall, the number of severe feather pecks recorded was higher than gentle pecks, but similarly stable across periods, ranging from 0.79 to 0.89 pecks/bird/10 minutes. No significant effects of period (F _2,257 = 1.32, p = 0.2682), liquid type (F _1,257 = 0.93, p = 0.3369), or their interaction (F _2,257 = 1.28, p = 0.2797) was observed.

High feather pecking hens visit SYN-enriched water more frequently despite no differences in visit duration measured via RFID

The average time spent at a water jug per visit was short, approximately 0.30–0.33 minutes or 18–20 seconds. Total time spent at the SYN water jugs per hen per period was around 5.26–5.94 minutes. Across all periods, hens visited the SYN water jugs approximately 17 times per 6.5-hour period.

There was no significant effect of period, SFP category (low or high peckers), or their interaction on total visit duration to SYN water jugs (all P > 0.05). Similarly, there was no effect of period or an interaction between the period and SFP category on the number of visits to the SYN water jugs (all P > 0.05). However, SFP category had a significant effect on the number of visits to SYN water jugs (F₁,₅₄₇.₃ = 4.80, p = 0.0288), with high peckers visiting more frequently (18.69 ± 1.08 visits/6.5 h) than low peckers (16.37 ± 0.89 visits/6.5 h) across all periods.

Feather cover damage and severe pecking rates

Feather cover scores for the head/neck, back/rump, and tail regions did not differ significantly across periods (Table 3). In contrast, feather cover on the belly progressively deteriorated across periods, with the proportion of birds exhibiting severe feather damage (score 2) increasing from nearly 20% in the Pre-Stress period to approx. 33% in the Post-Stress period. Total feather cover also deteriorated progressively across periods, with the proportion of birds exhibiting severe feather damage (score 2) increasing approx. 40% in the Pre-Stress period to approx. 54% in the Post-Stress period. The increase in hens with severe damage was in both cases logically associated with a decline in the proportion of hens with no visible damage to the feather cover. These descriptive results are confirmed by higher odds of having better feather cover on the belly area in both the Pre-Stress and Stress periods compared to the Post-Stress period. No significant difference was found between the Pre-Stress and Stress period. Similarly, birds were more likely to have better overall feather cover in the Pre-Stress and Stress period compared to the Post-Stress period, with no difference between Pre-Stress and Stress period.

Table 3.

The number (n) and percent of observations (%), including odds ratios (OR) and 95% confidence intervals (CI) for feather cover quality in the head/neck, back/rump, belly, and tail regions of laying hens.

Body Region / Period	Score 0 n (%)	Score 1 n (%)	Score 2 n (%)	reference vs comparison period	OR	95% CI
Head / Neck
Pre-Stress	122 (53.98%)	54 (23.89%)	50 (22.12%)	Pre-Stress / Stress	0.984	0.690-1.404
Stress	126 (56.25%)	42 (18.75%)	56 (25.00%)	Stress / Post-Stress	1.178	0.824-1.685
Post-Stress	119 (52.65%)	42 (18.58%)	65 (28.76%)	Pre-Stress / Post-Stress	1.160	0.815-1.650
Back / Rump
Pre-Stress	199 (88.05%)	4 (1.77%)	23 (10.18%)	Pre-Stress / Stress	0.754	0.412-1.378
Stress	203 (90.63%)	5 (2.23%)	16 (7.14%)	Stress / Post-Stress	1.300	0.711-2.376
Post-Stress	199 (88.05%)	8 (3.54%)	19 (8.41%)	Pre-Stress / Post-Stress	0.980	0.555-1.731
Belly
Pre-Stress	163 (72.12%)	19 (8.41%)	44 (19.47%)	Pre-Stress / Stress	1.005	0.669-1.511
Stress	160 (71.43%)	24 (10.71%)	40 (17.86%)	Stress / Post-Stress	2.495*	1.710-3.640*
Post-Stress	109 (48.23%)	44 (19.47%)	73 (32.30%)	Pre-Stress / Post-Stress	2.509*	1.717-3.666*
Tail
Pre-Stress	194 (85.84%)	12 (5.31%)	20 (8.85%)	Pre-Stress / Stress	1.067	0.633-1.798
Stress	190 (84.82%)	16 (7.14%)	18 (8.04%)	Stress / Post-Stress	0.990	0.591-1.656
Post-Stress	192 (84.96%)	16 (7.08%)	18 (7.96%)	Pre-Stress / Post-Stress	1.056	0.627-1.779
Overall Feather Cover
Pre-Stress	85 (37.61%)	52 (23.01%)	89 (39.38%)	Pre-Stress / Stress	1.114	0.788-1.574
Stress	80 (35.71%)	49 (21.88%)	95 (42.41%)	Stress / Post-Stress	1.767*	1.246-2.505*
Post-Stress	45 (19.91%)	60 (26.55%)	121 (53.54%)	Pre-Stress / Post-Stress	1.968*	1.390-2.786*

Feather cover score is based on Decina et al. (2019) where 0=no damage, 1=bald area ≤ $2 coin, 2= bald area >$2 coin. Feather cover condition was measured during pre-stress (39 WOA), stress (43 WOA), and post-stress (45 WOA) (pre-stress: n = 226, stress: n = 224, post-stress: n = 226). For overall feather cover score, the highest severity between the four body regions was recorded per individual bird in each period. The odds of having better feather cover is modelled where OR >1 indicates a better a feather cover in the latter period of the comparison compared to the first mentioned reference period. Significant OR and CI are marked with asterisks (*).

Body weight & feed consumption

The average body weight was 1.55±0.02 kg during Pre-Stress, 1.53±0.02 kg at the end of the Stress period, and 1.53±0.02 at the end of the Post-Stress period. There was a significant effect of period on weight (F_2,21.01=3.67, p = 0.0430). However, post hoc comparisons between pre-stress, stress and post-stress periods did not reveal statistically significant pairwise differences, indicating that biologically meaningful changes did not occur despite the marginal period effect in the overall model. However, there was no effect of SYN water consumption (F_1,16.41=0.04, p = 0.8347) on hens’ body weight. There was some evidence of an interaction effect between period and SYN water consumption affecting body weight (F_2,21.22=21.22, p = 0.0528); however, more data is required for further analysis to confirm the interaction.

Feed consumption was 114.8 ± 4.00 g per bird per day during Pre-Stress, 113.8 ± 3.32 g per bird per day during Stress period, and 109.4 ± 3.48 g per bird per day Post-Stress (LSM±SE). There was no effect of period (F_2,343.7 = 0.05, p = 0.9505) or an interaction between period*SYN water consumption (F_2,343.8 = 0.06, p = 0.9376) on daily feed consumption at the group-level.

Discussion

The present study employed a relative preference test to assess whether laying hens would voluntarily consume synbiotic (SYN)-enriched water across the 6-week test period (pre-stress, stress, post-stress). The results show that hens can discriminate between water types and actively prefer the SYN water, which combines mildly sweet prebiotics and beneficial probiotics. This preference suggests that palatability contributes to voluntary intake. Palatability of SYN water is likely mediated by sweet taste perception and/or beneficial bacteria and their metabolites via positive post-ingestive feedback.

Hens exhibited a consistent preference for freshly prepared SYN water over plain water across all experimental periods (pre-stress, during stress and post-stress), suggesting that SYN water acts as a reward. This preference may be influenced by one or more components of the SYN used in this study, which includes prebiotic fructooligosaccharides (FOS), chicory root (a natural source of FOS), and probiotic strains such as Bifidobacterium animalis, Enterococcus faecium, Pediococcus acidilactici, Lactobacillus reuteri, and Lactobacillus salivarius. We propose that the observed preference is, in part, mediated by FOS, which have a fruity aroma and sweet taste in humans (Franck, 2002). Mice lacking the canonical sweet taste receptor subunit T1R2 still develop and maintain preferences for sugar solutions, which is attributed to gut-brain signaling based on nutrient identity rather than caloric content (Nelson et al., 2001; Sclafani et al., 2015; Tan et al., 2020). Similarly, although chickens lack T1R2 (Baldwin et al., 2014), they still exhibit behavioral preferences for sweet substrates (e.g., Gentle 1972; Harlander-Matauschek and Rodenburg, 2011; Higashida et al., 2022; Savory, 2010), suggesting that their preference for FOS-containing solutions may be mediated by post-ingestive nutrient sensing rather than traditional sweet taste perception. Indeed, avian, extra-oral taste receptors and enteroendocrine cells in the gastrointestinal tract may serve as functional analogs to the gut-brain signalling pathway in mammals, enabling nutrient-driven preferences via central feedback mechanisms (Higashida et al., 2022; Tan et al., 2020).

The consistent preference for SYN water observed in this study may not only reflect its palatability, but also a learned behavioral response driven by microbiota-mediated feedback. Prior research has shown that SYN supplementation, like the formulation used here, can increase beneficial gut bacteria (e.g., Lactobacillus, Bifidobacterium), enhance immune function, and reduce behavioral reactivity in poultry (Johnson et al., 2024; Mohammed et al., 2019). These effects are particularly relevant in birds prone to stress-induced behaviors, such as feather pecking. Indeed, hens selected for high feather pecking behavior have repeatedly been shown to exhibit gut dysbiosis, characterized by reduced Lactobacillus abundance (Birkl et al., 2018; Van der Eijk et al., 2019). This microbial imbalance is associated with altered gut function, and neuroimmune dysregulation, including changes in serotonin signaling and pro-inflammatory markers (Huang et al., 2023; Mindus et al., 2022). Such findings support the hypothesis that gut microbiota composition can influence behavior via the bidirectional microbiota–gut–brain axis (van Staaveren et al., 2021). Intervention studies further reinforce this link. For example, it was demonstrated that oral administration of Lactobacillus rhamnosus JB-1 in stressed hens not only increased Lactocaseibacillus levels, but also improved T cell function and prevented escalating SFP (Mindus et al. 2021b, c). In mouse models, modulation of T-cell activity has been linked to stress resilience induced by feeding of this same L. rhamnosus strain (Liu et al., 2020). These findings suggest that SYN water may act as a reward, not only due to its taste and aroma, but also because it promotes a physiological balance, which reinforces preference through positive post-ingestive feedback. Finally, learned associations between the solution’s postingestive effects and its visual cue (color) may have strengthened the preference over time. Together, these mechanisms, palatability, nutrient‑driven reward, microbiota‑mediated feedback, and associative learning, provide a biologically plausible explanation for the hens’ consistent preference for SYN water.

In the present study, birds experienced stress through social mixing to disrupt established social hierarchies, a known stressor in chickens that induces frustration and feather pecking (Birkl et al., 2019a,b; Mindus et al., 2021b). However, we housed HFP and LFP birds together to mimic commercial flocks, in which only a proportion of hens exhibit SFP. Nonetheless, given that the two lines differ in behavioral predispositions and stress sensitivity, their co‑housing may also act as a mild background stressor, potentially influencing social dynamics independently of the imposed mixing events. As prebiotics and probiotics are known to confer benefits under stressful conditions, and based on rodent comfort-feeding studies, we hypothesized that hens would increase SYN intake during the stress challenge. However, SYN water intake decreased during the stress period compared to the pre-stress period (74.9 ml vs. 82.7 ml, respectively), while plain water intake remained stable. The near-significant interaction between water type and period suggests that stress may have selectively reduced the appeal of SYN water. One possible explanation is stress-induced reduction in reward sensitivity, or anhedonia, which is well documented in mammals and commonly assessed using sucrose-preference tests (Papp et al., 1991). In the present study, we propose that FOS exerts post-ingestive rewarding effects, similarly to sucrose. Therefore, the reduced SYN intake during stress may reflect a blunted post-ingestive reward response, aligning with the idea that chickens, like mammals, experience stress-induced changes in hedonic processing.

Feather‑pecking behavior remained stable across all experimental phases, with no significant changes in either GFP or SFP. At first glance, this might suggest that the social‑mixing stressor was too weak to alter affiliative GFP or escalate SFP. However, prior work from our groups shows that social disruption of this type can affect behavior even when chicks are reared with dark brooders, which we use as standard practice at our research facility; despite dark brooding, we still observe the development of feather pecking in non‑beak‑trimmed hens (e.g., Birkl et al., 2019; Mindus et al., 2021). Moreover, while some studies report that dark brooders mitigate feather pecking (Jensen et al., 2006; Gilani et al., 2012) and reduce injurious pecking (Sirovnik and Riber, 2022), the protective effects are not universal across genotypes or environments, and thus dark brooding does not eliminate feather pecking in all contexts. Taken together, these considerations make it unlikely that insufficient stress intensity alone explains the stable SFP observed here. A more plausible interpretation is that the SYN intervention exerted a buffering effect against stress‑related increases in SFP, thereby maintaining stable pecking rates across periods, an interpretation consistent with findings that oral Lactobacillus rhamnosus supplementation can prevent stress‑induced escalation of SFP, whereas non‑supplemented hens show increases (Mindus et al., 2021b,c). At the same time, we cannot exclude that repeated exposure to social mixing produced partial habituation, contributing to stabilized SFP. Interpreting GFP as a stress‑sensitive social outcome is biologically meaningful: in wild birds, affiliative contacts such as allopreening increase following social instability (Radford, 2008), and in domestic hens, GFP has been linked to social exploration, recognition, and group cohesion, and varies with social context (Riedstra and Groothuis, 2002; Rodenburg et al., 2004). Because published timelines for hierarchy stabilization after mixing are limited and primarily concern aggressive interactions rather than SFP, we note this as an alternative explanation and a target for future research.

Despite this potential buffering effect, the constant presence of SFP throughout the experimental period contributed to a progressive deterioration in feather cover, particularly in the belly region. Feather damage scoring revealed a marked decline in plumage condition: the proportion of birds with intact feathers (Score 0) dropped from 37.6% during the pre-stress period to 19.9% in the post-stress period, while those with severe damage (Score 2) increased from 39.4% to 53.5%. These findings highlight a critical welfare concern—namely, that even when SFP frequency remains stable, its chronic nature can lead to cumulative and potentially irreversible physical damage. Because plumage integrity is essential for thermoregulation, mobility (Garant et al., 2022), and social signaling (Morris, 1956), the ongoing deterioration in feather condition has broad implications for bird welfare. The stable pecking rates and worsening feather damage suggests that the success of proposed interventions to curtail feather damage cannot be assessed solely by pecking behavioral frequency. Instead, welfare assessments should prioritize the preservation of feather cover and integument health. This also underscores the importance of early intervention before SFP becomes a chronic behavior and physical damage accumulates (Decina et al., 2019).

Hens exhibiting a higher incidence of SFP behavior frequented the SYN jugs more often than their low SFP counterparts. Although this difference was modest in absolute terms, it may be partly explained by the generally heightened motor activity of high SFP hens (Kjaer, 2017.), which increases the likelihood of encountering the additive. Alternatively, it is also plausible that these visits are intentional, goal-directed to the SYN jugs. This interpretation is supported by findings from Heinsius et al. (2020), showing that high feather pecking (HFP) birds are capable of inhibiting impulsive pecks in a Go/No-Go task, indicating that their behavior is not merely reactive, but involves goal-directed control. Additionally, operant chamber testing demonstrates that HFP hens exhibit elevated pecking drive (Harlander-Matauschek et al., 2006), particularly under delayed reinforcement (Birkl et al., 2019a), consistent with increased motor output, an effort toward valued outcomes. Interestingly, research in mice has shown that even small amounts of comfort feeding can modulate stress-related behavior (Ulrich-Lai et al., 2015) raising the possibility that repeated access to SYN (e.g., FOS‑linked post‑ingestive reward) might be selectively sought by birds prone to SFP. Notably, visits to SYN jugs did not decline during the stress phase even though SYN intake did. This discrepancy suggests that at least some jug visits may have reflected visits to the available jug rather than the preferred jug, potentially due to resource competition, displacement, or queueing, particularly during periods of high flock activity. Because only one hen could access a jug at a time, hens may have temporarily visited the alternative (non‑preferred) jug if the preferred one was occupied. Competitive interactions of this type may have altered resource‑use patterns and may therefore have contributed to the divergence between visit frequency and actual consumption. Social influences from nearby hens at the apparatus could also have shaped visitation patterns independently of preference. Moreover, because the RFID system recorded presence at the jug but not confirmed ingestion, and water consumption was measured at the pen‑level rather than the individual level, we cannot determine whether visits corresponded to actual drinking. Future hypothesis‑driven work should pair individual RFID events with individual intake measurements and/or provide multiple simultaneous access points to disentangle true preference from competition‑driven jug selection and to better characterize how social dynamics shape drinking behavior.

Notably, preference for SYN water may have unintended consequences. Repeated beak contact with the sweet SYN solution can transfer FOS and probiotic residue onto conspecifics’ plumage, inadvertently reinforcing feather-directed pecking. Such a feedback loop is supported by prior findings that hens can be conditioned to preferentially peck at sucrose-soaked feathers, with subsequent increases in SFP observed in group-housed conditions (Harlander-Matauschek et al., 2008). Moreover, pecking serves, not only as a mechanism for interaction and exploration, but also for microbial exchange. Beak contact with feed, water, feathers, and environmental surfaces enables the transfer of particulate matter, microorganisms, and nutrient compounds. In mammals, social grooming predicts microbial similarity (Sherwin et al., 2019). Similar horizontal transmission may occur in birds via behaviors like GFP and SFP, preening, and shared feeding (van Staaveren et al., 2021). While speculative, these findings raise the possibility that behavioral interactions, especially pecking, may facilitate the spread of prebiotic and probiotic components from SYN formulations, potentially influencing gut microbiota and associated behavioral outcomes.

This study aimed to determine whether stressed laying hens would voluntarily consume a SYN solution enriched with palatable components such as fructooligosaccharides (FOS). The hens consistently preferred SYN-enriched water across all periods, even during unpredictable social stress, highlighting the importance of palatability in promoting voluntary intake. This finding is particularly encouraging given that the current synbiotic, when mixed into feed, has demonstrated health benefits in other physiological domains in chickens (Mohammed et al., 2018; 2019). Moreover, under conditions of stress or disease, birds need to be encouraged to drink before they eat, making the effectiveness of additives delivered via drinking water especially critical. Despite stable rates of SFP, plumage damage progressed. These findings highlight a crucial point: while palatability fosters intake, resolving chronic stress-linked behaviors like SFP may require precision-targeted synbiotic formulations with strains that influence neuroimmune and integumentary pathways associated with SFP.

Funding

This study was funded by the Ontario Ministry of Agriculture, Food & Rural Affairs (OMAFRA), and the Natural Sciences and Engineering Research Council of Canada (NSERC).

CRediT authorship contribution statement

Anna Naim: Writing – review & editing, Writing – original draft, Visualization, Validation, Methodology, Investigation, Formal analysis, Data curation. Nienke van Staaveren: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Formal analysis, Data curation, Conceptualization. Emily M. Leishman: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Project administration, Methodology, Investigation, Formal analysis. Anna Lea Nicklas: Writing – review & editing, Writing – original draft, Visualization, Methodology, Investigation, Formal analysis, Data curation. Dan Tulpan: Writing – review & editing, Writing – original draft, Supervision, Software, Formal analysis, Data curation. Paul Forsythe: Writing – review & editing, Writing – original draft, Validation, Supervision, Project administration, Methodology, Funding acquisition, Data curation, Conceptualization. Alexandra Harlander: Writing – review & editing, Writing – original draft, Visualization, Validation, Supervision, Resources, Project administration, Methodology, Investigation, Funding acquisition, Conceptualization.

Disclosures

The authors declare the following financial interests/personal relationships which may be considered as potential competing interests:

Alexandra Harlander reports equipment, drugs, or supplies was provided by Ontario Ministry of Agriculture Food and Rural Affairs. Alexandra Harlander reports article publishing charges and equipment, drugs, or supplies were provided by Natural Sciences and Engineering Research Council of Canada. Alexandra Harlander reports equipment, drugs, or supplies was provided by dsm-firmenich. If there are other authors, they declare that they have no known competing financial interests or personal relationships that could have appeared to influence the work reported in this paper.