S.E. van Nieuwamerongen

A review of sow and piglet behaviour and performance in group housing systems for lactating sows.

Commercial use of group housing systems for lactating sows is limited, but the recent transition to group housing during gestation in the EU may result in a renewed interest in such systems. Therefore, this review aims to identify key factors that may contribute to the success or failure of group housing of lactating sows in comparison with individual housing by describing the variety in group housing systems and discussing animal behaviour and performance compared with individual housing. Group housing systems can be divided in multi-suckling (MS) systems, in which sows are grouped with their litters, and get-away (GA) systems, which include a separate communal area accessible to sows only. These systems differ in many aspects regarding management and layout but, compared with individual housing, generally provide more environmental complexity, more freedom of movement for the sows and more freedom to express behaviours related to, for example, maternal care and social interactions. Group housing poses several risks, such as disrupted nursing and an increased level of crushing during the MS phase, and in the GA systems there is a risk for early cessation of nursing. On the other hand, pre-weaning mingling of litters clearly benefits piglet social development and may improve adaptation to the post-weaning situation. In addition, group-housed sows may show lactational ovulation, which provides opportunities for insemination during an extended lactation period, which benefits the piglets. Gradual transitions in social and physical environment around gestation, farrowing, grouping and weaning seem to be key success factors for group housing systems during lactation. In addition, selection of suitable sows and quality of stockmanship seem important.

Development of piglets raised in a new multi-litter housing system vs. conventional single-litter housing until 9 weeks of age

This study compared the development until 9 wk of age of piglets raised in either a multi-litter (ML) system or a conventional single-litter (SL) system. The ML system consisted of a multi-suckling system with 5 sows and their litters before weaning, followed by housing in a pen with enrichment in a group of 40 piglets after weaning. In the SL system, piglets were housed with a crated sow before weaning, followed by postweaning housing in groups of 10 littermates in a standard pen. Fifty litters were used in 5 batches and piglets were weaned at 4 wk of age. Preweaning mortality was higher in the ML system than in the SL system (3.22 ± 0.42 vs. 1.52 ± 0.25 piglets per litter, < 0.01), mainly due to crushing before grouping of litters. Litter size at grouping did not differ between systems. ML piglets showed more feed-directed behavior at 2 wk of age (6.80 ± 0.96 vs. 2.35 ± 0.59, < 0.01), suggesting an earlier start of feed exploration, possibly due to social learning from the floor-fed sows and other piglets. Moreover, before weaning, ML piglets showed less damaging oral manipulation (e.g., tail biting) than SL piglets (1.4 ± 0.2 vs. 3.6 ± 0.3 freq/h, < 0.001), which was likely related to the more enriched environment in the ML system. After weaning, ML piglets ate 81% more feed between d 1 and 2 (0.29 ± 0.02 vs. 0.16 ± 0.03 kg/piglet, < 0.01) and had an 82% higher weight gain until d 5 than SL piglets (1.35 ± 0.21 vs. 0.75 ± 0.17 kg, < 0.05) despite a similar weaning weight (ML: 8.4 ± 0.2 kg, SL: 8.3 ± 0.2 kg). Within the first 2 wk after weaning, ML piglets had a lower fecal consistency score (0.27 ± 0.03 vs. 0.39 ± 0.03, < 0.05), indicating a lower occurrence of diarrhea compared with SL piglets. Over the entire 5-wk postweaning phase, ML piglets had a 24% higher weight gain ( < 0.05) and showed more play behavior (4.0 ± 0.3 vs. 2.8 ± 0.3 freq/h, < 0.05) and less damaging oral manipulation (1.8 ± 0.3 vs. 3.5 ± 0.4 freq/h, < 0.01) than SL piglets. These results are probably explained by a combination of the differences in preweaning development, early postweaning performance, and postweaning environment, with a larger and more diverse social group and more physical enrichment in the ML pen. To summarize, provided that preweaning mortality can be reduced, the ML system seems promising for raising robust piglets with better welfare, indicated by a better preweaning behavioral development, improved transition to the postweaning phase, and better postweaning performance.

Post-weaning performance of piglets raised in a multi-suckling system vs farrowing crates

Carbohydrates are not the only macronutrients triggering insulin release following food intake. Besides other components, leucine (Leu) has been demonstrated to induce insulin secretion in humans and in laboratory rodents. We hypothesized that Leu stimulates the release of insulin in dairy cows, and thus our objective was to test the effects of a single-dose of Leu infused intraduodenally on the concentrations of insulin, glucose and free amino acids (AA) in blood plasma as compared to infusions with either glucose or saline. Six duodenum-fistulated Holstein cows were studied in a replicated 3 × 3 Latin square design with 3 periods of 7 d, in which the treatments were applied at the end of each period. The treatments were duodenal bolus infusions of Leu (DIL; 0.15 g/kg BW), glucose (DIG; at Leu equimolar dosage) or saline (SAL). Blood samples were taken at -15, 0, 10, 20, 30, 40, 50, 60, 75, 90, 120, 180, 210, 240 and 300 min relative to the infusion and the concentrations of insulin, glucose and AA were determined in plasma. In DIG, insulin and glucose concentrations peaked at 30-40 and 40-50 min after the infusion, respectively. Insulin concentrations were greater (P<0.05) from 30-50 min in DIG than DIL and SAL. In DIG, glucose concentrations were greater (P≤0.01) from 30-75 and 40-50 min than in DIL and SAL, respectively. In DIL, Leu concentrations peaked 50-60 min after infusion, reaching 20 and 15-fold greater values than that in DIG and SAL, respectively. The plasma concentrations of total AA minus Leu were affected by treatment (P<0.0001), resulting in lowest mean concentrations of total AA minus Leu in DIL, followed by DIG and SAL. The data suggest that Leu infusion did not elicit an apparent insulin response, but may stimulate the tissue uptake of AA by mechanisms yet to be elucidated.Breeding goals of Holstein dairy cattle in New Zealand (NZL) focus on traits reflecting adaptation to pasture-based systems. Due to high prices for concentrates, and advantages for cattle health and welfare, low input grazing systems become important in regions of North-West Germany. The present study addresses the hypothesis that progeny of NZL-Holstein sires are superior in 30 German grassland farms compared to offspring of Holstein sires being progeny tested in German indoor systems. A research design was implemented to create three different genetic groups (F1 generation) within herds on the basis of a German Holstein cow (GHC) population: Group 1=GHC x NZL sires, group 2=GHC x GH sires, group 3=GHC x GH_pasture sires. GH_pasture sires represent bulls of German origin with high breeding values for the traits being important in NZL. Group comparison focused on the full set of production and functional traits from official recording systems, and in addition on further novel traits reflecting cow health and welfare: Scores for body condition (BCS), locomotion (LOCS), hock lesions (HLS), and total hygiene (THS). Linear and generalized linear mixed models were applied to estimate group differences for a longitudinal data structure (2998 observations). In parity 1, daughters of NZL sires had the highest LSMean for BCS (2.56), and lowest LSMean for LOCS (0.10), HLS (0.25) and THS (3.67), reflecting an enhanced robustness of HF strains from NZL. Within breed genetic analyses simultaneously considering all genetic groups revealed a moderate genetic background with the following heritabilities: BCS (0.25), LOCS (0.12), HLS (0.04), and THS (0.17). Genotype by environment interactions between pasture-based and indoor systems were studied by applying multiple trait models. For all traits, rg were smaller than 0.80, especially indicating GxE interactions for low heritability functional traits.