Empowering large litters: can enhanced pen design with simulated udders improve piglet growth in their early life?

Abstract

Hyper-prolific sows produce large litters, which results in greater birth weight variation and an increased need for nurse sows. This study explored ways to optimize expanded farrowing pen space, following European Food Safety Authority (EFSA) recommendations, to help loose-housed sows manage litter sizes that surpass their available teat capacity. The aim was to enhance piglet growth and mortality while maintaining efficient production. A total of 66 litters were randomly assigned to one of two groups at birth: a control group (CON; n = 33) or a treatment group (TRT; n = 33), and only piglets born weighing more than 750 g were included in each group. From birth until weaning, CON sows nursed 16 piglets per litter, while TRT sows nursed 20 piglets. During the first three days postpartum, an enclosure setup was present in the creep area of pens housing TRT litters. In this enclosure, the four smallest piglets (S) were confined for six hours with access to milk replacer (MR), followed by the next four smallest (M) for another six hours, while the four largest piglets (L) were confined overnight for 12 hours. The last eight piglets from the TRT group were reared by the sow (SRT). In contrast, piglets in the CON group remained with the sow without intervention. Piglet weights were tracked daily from day one to three, with additional measurements taken on days 11, 18, and 25. No difference in litter weight gain was observed between the two groups from day one to three. However, litter weight gain was higher in CON compared with TRT in the entire study period (2.529 vs. 2.034 kg/day; P = 0.005). No difference in mortality was observed between TRT and CON from days 1–3. However, it was significantly higher in TRT than CON from day 3–24 (18.4% vs. 9.0%; P = 0.0006). Within TRT, L piglets maintained the highest body weight, while S piglets had the lowest (6.674 vs. 4.104 kg at day 25; P < 0.0001). ADG differed only in the first three days, where SRT piglets grew fastest (108 g/day; P < 0.0001). Milk disappearance was consistently highest in L piglets (297 g/6 h corrected; P < 0.0001). Mortality was highest in S piglets (60%), intermediate in M (30%), and lowest in L (11%) and SRT piglets (18%; P < 0.0001).

The enclosure setup might assist hyper-prolific sows in rearing large litters; however, no direct benefits on growth performance were observed. These findings suggest that further interventions are necessary to support piglets throughout the entire suckling phase.

Introducing a novel simulated udder during periods of temporary physical enclosure that preserves sow–piglet interactions can improve early-life growth, survival, and welfare, addressing challenges associated with large litters in modern pig production.

Introduction

Danish pig production has significantly improved in efficiency over the years, particularly with a notable increase in litter size. The average number of liveborn piglets per litter has increased to 18.5 in 2024 (Hyttel 2025), resulting in a surplus of piglets beyond the sow’s natural rearing capacity. Larger litters are often associated with an increased number of low birth weight (LBW) piglets (Riddersholm et al. 2021), greater birth weight variation, and increased pre-weaning mortality (Milligan et al. 2002). On average, birth weight decreases by approximately 20 g for each additional piglet, with nearly 30% of newborn piglets weighing less than 1 kg (Riddersholm et al. 2021).

In response to these challenges, the use of nurse sows has increased. While nurse sows enable managing surplus piglets, their use requires intensive management and presents several welfare concerns (Sørensen et al. 2016). Piglets reared by nurse sows often experience poorer growth performance, increased competition, and express a greater level of aggression (Robert and Martineau 2001). At the same time, nurse sows can face an increased risk of physiological stress, muscular and cardiovascular impairments (Pedersen et al. 2013), teat and leg lesions (Sørensen et al. 2016), and complications related to delayed milk letdown (Kobek-Kjeldager et al. 2020). These issues can negatively affect piglet survival and overall production efficiency. Furthermore, there is a risk that milk yield from nurse sows may be insufficient to support the growth potential of piglets, further exacerbating the challenges for especially LBW piglets.

To address the challenges associated with nurse sow use, alternative rearing strategies have been explored. The research show that artificial rearing methods like use of temporary physical enclosure and complete artificial rearing in rescue decks assist to manage surplus piglets (Amdi et al. 2022). While these strategies can potentially lower pre-weaning mortality, they do not consistently improve piglet growth (Baxter et al. 2020). Studies comparing artificially reared piglets to those raised by sows indicate that sow-reared piglets achieve higher body weight (BW), feed intake, and greater adipose tissue reserves, muscle mass, and bone mineral content at weaning (Amdi et al. 2022). Artificially reared piglets have been shown to exhibit lower immunoglobulin levels, higher inflammatory markers, and elevated cortisol concentrations, all of which may negatively impact immune system function (Han et al. 2024). Behaviourally, these piglets display reduced exploratory and play behaviours, along with an increase in abnormal social interactions such as belly nosing (Han et al. 2024). Moreover, early-life microbial interactions are critical for the proper development of immune responses, suggesting that artificial rearing can fundamentally disrupt immune system maturation (Stokes 2017). Overall, artificial rearing may adversely affect immune system development, social behaviour, and long-term physiological outcomes, as artificially reared piglets lack access to essential bioactive compounds, microbial exposure, and maternal interactions necessary for optimal early-life development. Despite vast amount of research, there is still no optimal solution that effectively balances piglet survival, growth, welfare, and production efficiency. Existing strategies fail to replicate the biological and behavioural benefits of natural rearing, highlighting the need for novel approaches. Complete artificial rearing is not allowed in several European countries under EU Council Directive 2008/120/EC 2008 (Council Directive 2008/120/EC 2008/120/EC 2008).

To overcome the limitations of current rearing strategies, this study investigated the use of a novel simulated udder during periods of temporary physical enclosure of piglets, designed to maintain visual and social contact with the sow and littermates while promoting natural suckling behaviour. The study explored the potential to enhance piglet survival and growth within a system that accommodates larger litters while maintaining production efficiency.

Materials and methods

This study was conducted in accordance with the Danish Animal Experiments Act, implementing EU Directive 2010/63/EU, and was approved by the Danish Animal Experiments Inspectorate (license no. 2022-15-0201–01175). The experiment was conducted on a Danish conventional herd in Denmark.

Animals, experimental design, and diets

At farrowing, 66 litters (Duroc × (Danish Landrace × Yorkshire)) from 66 sows (Danish Landrace × Yorkshire) of parity two to seven were housed in farrowing pens equipped with a sow area and a creep area. Sows were confined during farrowing and for the first three to four days post-partum, after which they were loose housed for the remainder of lactation. Litters were eligible for inclusion in the trial provided that farrowing had been completed within the 12 hours preceding group allocation. The litters were then randomly allocated to either a control group (CON; n = 33) or a treatment group (TRT; n = 33). Only piglets weighing >750 g at birth were included in the study. Piglets added to litters (for standardization purposes) were required to meet the same inclusion criteria, including completion of farrowing within the previous 12 hours and a birth weight > 750 g. Piglets weighing ≤750 g were fostered to nurse sows and were not included in the trial. If necessary, TRT and CON litters were supplemented with piglets of similar age from another litter, which was not included in the study to reach 20 or 16 piglets. From farrowing to weaning, CON sows reared 16 piglets per litter, which reflects the standard number of piglets typically managed under conventional conditions in Denmark, whereas TRT sows reared 20 piglets per litter.

From day one to three post-partum, a custom-made enclosure setup was installed in the covered creep area of pens housing TRT sows and litters. The enclosure was constructed from wire fencing and consisted of three sides (Fig. 1). The side of the enclosure facing the sow, where the simulated teats were mounted, measured 55 cm; the back wall measured 90 cm, and the third side measured 85 cm. The enclosure contained four simulated teats mounted on the front side, facing toward the sow, each positioned 15.5 cm above the creep area floor (measured to the nipple tip). Although the simulated teats were positioned opposite the sow, piglets could see the sow’s udder while drinking from the simulated teats. Each simulated teat consisted of two components: a teat body (supporting structure) and a nipple (used for piglet suckling). (Hvid et al. 2023). The teat bodies were fabricated by casting silicone (body silicone, Shore 15, Makemake, DK) in custom molds produced via 3D printing. Each teat body measured 12.2 cm in width and 14.4 cm in height and was dyed light pink to resemble natural tissue. A commercially available baby-bottle nipple (MAM, size 2 tip) was fixed to the front of each teat body (Hvid et al. 2023). Milk replacer was supplied to each nipple through an 8 mm diameter tube connected to a 0.5 L plastic bottle. The tubing was sealed with transparent silicone sealant, and the bottles were secured to the enclosure walls using steel wire.

Figure 1.

Illustration of the pen and enclosure. A) Photograph of the entire farrowing pen with the temporary enclosure positioned in the corner. B) Illustration of the temporary physical enclosure: a = 55 cm (front side facing the sow), b = 85 cm (side facing the neighboring pen), c = 90 cm (back side facing the aisle). Four artificial teats were mounted on the front side, each positioned 15.5 cm above the creep-area floor. Each teat consisted of a silicone teat body cast in custom 3D-printed molds, fitted with a MAM size-2 baby-bottle nipple.

During the three-day period, the four smallest piglets (S) in each litter, based on birth weight, were kept in the enclosure and given semi–ad libitum access to milk replacer (MR) for the first 6 hours in the morning (08:00–14:00). This was followed by the next four smallest piglets (M) for another 6 hours (14:00–20:00), while the four largest piglets (L) were kept in the enclosure overnight for 12 hours (20:00–08:00). The rationale for separating the piglets into these specific groups was to facilitate milk intake for the smallest piglets by reducing competition. By temporarily separating the largest piglets, the remaining piglets had improved access both to the milk replacer and to the sow. The remaining eight piglets in the treatment group were reared exclusively by the sow (sow-reared treatment; SRT) during the experimental period. Piglets in the control group (CON) were reared by the sow without intervention.

Individual piglet body weight (BW) was recorded daily from day one to three, with additional measurements on days 11, 18, and 25. Mortality and health status of the piglets were monitored daily from day one to three and weekly from day four to 25.

In the enclosure from day one to three of age, the TRT group received semi ad libitum MR (Danmilk Rescue 2, AB Neo, Videbæk, Denmark) consisting of milk and whey-based ingredients and vegetable fat. The MR was prepared by mixing 1700 g of milk powder with 68 g of blood plasma and adding water to reach a total volume of 10 L. The MR was optimized to resemble sow milk in terms of protein, energy, and fat content (Hojgaard et al. 2020). When switching piglet groups in the enclosure, the remaining MR volume in the bottles was weighed back, and the bottles were refilled with freshly mixed MR. The CON group was reared by the sow and only received sow milk. The ingredients and chemical composition are shown in Table 1. All piglets were offered creep feed from day 10 of age.

Table 1.

Ingredients and calculated nutrient composition of the milk replacer provided in the enclosure.

Item	Content	Item	Content
Fat filled whey powder	38.16	Vitamin A added	31.10 (1000) IU
Dairy bi-products	45.91	Vitamin D3 added	3.50 (1000) IU
Vit/Min + probiotic mix	2.3	Mn added	0 mg
Whey protein concentrate (WPC)	13.63	Fe added	171 mg
Fat filled whey powder	38.16	Cu added	131 mg
		Zn added	116 mg
Calculated nutrient composition		Iodine added	2 mg
Dry matter	97.45 %	Se added	0 mg
Crude protein	22.20 %	Bio plus YC	400 mg
Crude fat	32.00 %	Bacillus subtilis	1 mia
Crude ash	5.06 %	Aroma	100 mg
Crude fiber	0.00 %
Lactose	365 g
Lysine	18.59 g
Methionine	5.40 g
Calcium	6.85 g
Phosphorous	5.58 g
Sodium	4.48 g

Statistical analysis

All analyses were performed in SAS (version 9.4; SAS Institute Inc., Cary, NC, USA), and all illustrations were created in GraphPad Prism (version 10.4.1 (627); GraphPad Software, La Jolla, CA, USA). All data were checked for normality and homoscedasticity of residuals prior to formal analyses. For litter-level parameters, the litter was used as the experimental unit. To account for differences in the number of pigs per litter between groups, litter weights were normalized by dividing by the adjusted number of pigs. Average daily gain (ADG) was analysed using linear mixed-effects models, with treatment group and sow age (parity 2–3 vs. ≥4) included as fixed effects, and litter included as a random effect. Mortality from day 0 to day 3 was analysed using a generalized linear mixed model with treatment group as a fixed effect, a logit link function, and litter included as a random effect. For analyses within the treatment group comparing the four subgroups (defined by birth weight and treatment), the individual pig was used as the experimental unit for growth from day 1 to 3 and from day 1 to day 24, with subgroup included as a random effect nested within litter, and litter included as a random effect. P-values below 0.10 were considered a tendency, and P-values below 0.05 were considered statistically significant.

Results

Growth performance of the treatment and control group

Normalized Litter weight gain did not differ between groups during the first three days postpartum (1.418 kg/day in CON vs. 1.311 kg/day in TRT) (Fig. 2). However, over the entire experimental period (day 1–25), litter weight gain was greater in the CON compared with TRT group (2.529 kg/day vs. 2.034 kg/day; P = 0.005) (Fig. 2).

Figure 2.

Normalized litter daily weight gain. Normalized litter daily weight gain of the two groups; Control (CON) and Treatment (TRT) during the enclosure period (days 1–3) and over the entire study period (days 1–25). ^*Indicate P < 0.05.

Mortality in the treatment and control group

From day one to three mortality was numerically higher in the TRT group compared with CON (6.6% vs. 4.2%) (Fig. 3). From day 3–24, mortality was higher (P = 0.0006) in the TRT group (18.4%) compared with the CON group (9.0%) (Fig. 3). Across the entire pre-weaning period (day 1–24), cumulative losses were 173 piglets in the TRT group compared with 73 in the CON group.

Figure 3.

Mortality between CON and TRT. Mortality of the two groups; Control (CON) and Treatment (TRT) during the enclosure period (days 1–3) and the whole study period (days 3–25). ^*Indicate P < 0.05.

Growth performance of birth weight categories within the treatment group

Within the TRT group, piglet BW differed (P < 0.0001) between subgroups at all time points (Fig. 4). On day one, the L piglets were the heaviest (1.737 kg), followed by the SRT piglets (1.471 kg), M piglets (1.229 kg), and S piglets (1.023 kg). These differences persisted throughout the pre-weaning period. By day 25, L piglets maintained the highest BW (6.674 kg), compared with 5.820 kg in SRT piglets, 4.701 kg in M piglets, and 4.104 kg in S piglets (P < 0.0001). Average daily gain, corrected for birth weight and sow, differed (P < 0.05) only in the first three days postpartum. Sow-reared-treatment piglets achieved the highest ADG (108 g/day), followed by S piglets (105 g/day), M piglets (94 g/day), and L piglets (73 g/day; P < 0.0001). No differences were observed between groups for ADG during the entire experimental period (day 1–25). Across the pre-weaning period, all piglets maintained steady growth rates of approximately 114–180 g/day.

Figure 4.

Average TRT subgroup BW. Daily weight gain of the four subgroups within the treatment group; The four smallest (S), next four smallest (M), four largest (L) the treatment group were entirely reared by the sow (Sow-reared-treatment; SRT) during the entire study period (days 1–25). ^*P < 0.05 indicates a significant overall treatment effect across all study days (days 1–25).

Milk replacer disappearance within birth weight categories of the treatment group

Marked differences in milk disappearance were found between treatment groups. On day one, milk disappearance in L piglets were higher (P < 0.0001) MR (453 g) compared with S (222 g) and M piglets (130 g) (Table 2). This difference persisted on day two (648 g vs. 192 g and 123 g; P < 0.0001) and day three (681 g vs. 221 g and 170 g; P < 0.0001) (Table 2).

Table 2.

Average daily milk replacer disappearance in the enclosure within the treatment group (S, M, and L) from day one to three post-partum.

	Small piglets (S)	Medium (M)	Large piglets (L)	SEM	P-value
Day 1, ml	222	130	453	29	<0.0001
Day 2, ml	192	123	648	40	<0.0001
Day 3, ml	221	170	681	43	<0.0001
Days 1–3, ml	212	141	297	22	<0.0001

When averaged across days one to three, L piglets had the highest (P < 0.0001) MR disappearance (594 g), compared with 212 g in S piglets and 141 g in M piglets. Because L piglets had 12 hours available for MR consumption compared with 6 hours for the other groups, disappearance data were corrected accordingly. Even after correction, the MR disappearance of the L piglets was higher (P < 0.0001, 297 g) than of S piglets (212 g) and of M piglets (141 g) (Table 2).

Mortality within birth weight categories of the treatment group

During the first three days, S piglets had the highest mortality rate (17%), which was higher (P = 0.007) than SRT piglets (2%) and L piglets (4%). M piglets (10%) showed intermediate values (Fig. 5).

Figure 5.

Mortality within TRT. Mortality of the four subgroups within the treatment group; The four smallest (S), next four smallest (M), four largest (L) the treatment group were entirely reared by the sow (Sow-reared-treatment; SRT) during the enclosure period (days 1–3) and over the entire study period (days 3–25). Different lowercase letters indicate P < 0.05, and different uppercase letters indicate a tendency (0.05 ≤ P < 0.10).

Over the entire period (day 1–24), mortality differed (P < 0.0001) between birth weight categories, being highest in S piglets (60%), followed by M piglets (30%), sow-reared piglets (18%), and L piglets (11%) (Fig. 5). When excluding mortality rate during the early period (day one to three), a similar pattern for the mortality rate day 3–24 within the treatment group was revealed: S piglets had the greatest losses (34%), higher than SR (15%) and L piglets (9%), with M piglets again intermediate (19%; P < 0.0001) (Fig. 5).

Discussion

This study evaluated whether a temporary physical enclosure with access to simulated udders with MR could support pre-weaning growth and survival in large litters. While overall litter growth was reduced and mortality increased in treatment litters compared with controls, the results also revealed important insights into the potential and limitations of this management strategy. Notably, no differences were observed in litter growth between TRT and CON litters during the first three days postpartum, suggesting that the temporary physical enclosure system did not negatively impact early piglet performance when taking litter size of the groups into account.

Litter growth performance

During the first three days, litter weight gain was comparable between the TRT and CON litters, indicating that the system did not compromise immediate growth despite reduced contact with the sow. However, from days 1–25, litter growth was lower in TRT litters. One possible explanation is that the high management input during the first three days in the enclosure was not sustained in the later period, leaving piglets less supported after reintegration into sow-rearing. Another factor could be the disruption of teat order caused by repeatedly separating subgroups. The establishment of teat order in terms of accessing the most functional teat and not necessarily the same teat between nursing (Pedersen et al. 2011) during the first days after farrowing can be crucial for consistent milk intake, and interruptions may have disadvantaged piglets once the enclosure period ended (Skok and Škorjanc 2014).

Mortality outcomes

The major drawback of the system was increased mortality in TRT litters, particularly after day three. Within the TRT group, mortality was highest among the S piglets, despite the intention of the system to protect them. This indicates that the design may not have provided the smallest animals with sufficient nutritional or behavioural support. In contrast, the L and sow-reared treatment piglets demonstrated much lower mortality, indicating that these piglets adapted more easily to the contemporary physical enclosure system than their smaller littermates.

Milk replacer disappearance and subgroup differences

Milk disappearance results suggest that L piglets consistently had a higher MR disappearance than S piglets, even after adjusting for time spent in the enclosure. This is not unlike conventional rearing, where larger littermates are likely to have consumed more colostrum, and thereby have an advantage for securing a functional teat and maintaining a weight lead over lighter littermates. This outcome illustrates that heavier piglets were better able to exploit the replacer feeding opportunity, which may explain their relatively strong growth and survival compared with smaller littermates. Instead of levelling weight disparities, the enclosure system may therefore have amplified them. However, contradicting this observation, there was more MR disappearance in the group of S compared with M piglets. Although some animals develop diurnal or circadian patterns of feed intake as they age, such patterns are generally weak or absent in very young animals and show substantial individual variation (Bus et al. 2023). Therefore, while day-to-day or within-day variation in suckling behavior may occur, it is unlikely to have introduced a systematic bias over the relatively short observation period. Thus, this observation should be interpreted with caution. Despite the higher MR disappearance observed in S compared with M piglets, this did not result in catch-up growth, as S piglets did not achieve similar daily gain to M piglets, which contradicts a previous study suggesting similar feed intake between low- and normal-birth-weight piglets (Tang et al. 2023). With these observations in mind, it could be speculated that S piglets were less efficient in metabolizing energy and nutrients, or that a relatively greater proportion of energy was directed towards thermoregulation, as S piglets were likely at greater risk of hypothermia (Caldara et al. 2014). Lastly, the observation may simply be explained by S piglets interacting more frequently with the teats, thereby increasing apparent MR disappearance.

Future refinements should focus on ensuring that weaker piglets can access the MR more effectively, perhaps through subgrouping strategies, nipple design improvements, or milk formulas that combine palatability with immunological support (De Vos et al. 2014; Sugiharto et al. 2015; Hvid et al. 2023). In addition, it is worth investigating to what degree LBW piglets are in fact able to digest and metabolize an optimized MR and how energy and nutrients are distributed between physiological processes. A recent study has suggested that low compared to normal birth weight piglets have reduced TCA cycle metabolites among others, which in general could indicate unfavourable metabolic alterations (Wellington et al. 2023), underscoring the importance of a greater understanding of LBW piglet biology.

Implications for large litter management strategies

Although overall results were not entirely promising in terms of increasing growth and decreasing mortality of the smallest litter mates, the lack of differences in growth during the first days suggests that the enclosure itself was not inherently harmful. By identifying the exact limitations related to improving survival and growth, particularly of the S piglets, during artificial rearing conditions, using the temporary physical enclosure approach could potentially be a solution to the challenge of rearing large litters from hyper-prolific sows. A more targeted use of the enclosure system may be beneficial, for example, placing the smallest piglets with a nurse sow to secure colostrum and stable udder access (Devillers et al. 2011), while moving only the largest piglets to the enclosure to relieve competition at the sow’s udder. Such a strategy could reduce disruption for vulnerable piglets while still offering larger litter mates a chance to balance milk access/consumption, as well as reducing the number of nurse sows used.

Limitations of temporary physical enclosure

The present study did not evaluate sow behaviour or piglet stress responses, which may have influenced survival and growth performance outcomes. Additionally, the composition of the MR differed from sow milk and may have limited its effectiveness, particularly with respect to sufficient development of the immune system. Future research should evaluate modified designs where the enclosure is applied more selectively, replacer supplementation is optimized for palatability and immunoglobulin content, and supplemental heating is integrated to secure thermoregulation. Ensuring the hygiene of simulated udders will also be critical to prevent pathogen spread.

In conclusion, the temporary physical enclosure setup supported hyper-prolific sows in managing large litters during the first three days postpartum. However, no direct benefits were observed in piglets’ growth performance or survival during this period. These findings highlight the need for continued support strategies throughout the entire lactation period to improve outcomes for both sows and piglets.

Glossary

List of abbreviations

(ADG)

Average daily gain

(BW)

Body weight

(CON)

Control group

(EFSA)

European Food Safety Authority

(LBW)

Low birth weight

(MR)

Milk Replacer

(after the four smallest piglets (S) in the litter)

Next four smallest piglets (M)

(SRT)

Sow-reared-treatment

(L)

The four largest piglets

(S)

The four smallest piglets

(TRT)

Treatment group

Funding

The authors would like to acknowledge financial support by the Pig levy foundation.

Author contributions

JGM contributed to conceptualization, data curation, investigation, funding acquisition, writing of the original draft, and review and editing. JMvdL contributed to conceptualization, data curation, investigation, and review and editing. VAM contributed to conceptualization, data curation, investigation, funding acquisition, and review and editing. MBN contributed to data curation. CL contributed to conceptualization, data curation, investigation, writing of the original draft, and review and editing.

Conflict of interest

None declared.