If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
The effect of stocking density and a blind on the behavior of Holstein dairy cows in group maternity pens. Part II: Labor length, lying behavior, and social behavior
Department of Agricultural Sciences/Animal Science, University of Helsinki, PO Box 28, FI-00014 Helsinki, FinlandDepartment of Production Animal Medicine, Research Centre for Animal Welfare, University of Helsinki, PO Box 57, FI-00014 Helsinki, FinlandHelsinki One Health, University of Helsinki, FI-00014 Helsinki, Finland
In natural settings, dairy cows separate from the herd to give birth. When kept indoors, seeking isolation before calving may be restricted and may depend on space and resources provided in maternity housing. The effect of group maternity pens on behavior around calving and labor progress is unknown. Therefore, the objective of this study was to determine the effects of stocking density and provision of a blind in group bedded pack maternity pens on lying and social behavior as well as length of labor of preparturient dairy animals. The study was conducted as a complete randomized block design with a 2 × 2 factorial arrangement of treatments including stocking density and presence or absence of a blind, resulting in a total of 4 treatments: (1) high stocking density (7.7–12.9 m2 lying space/cow) with a blind, (2) low stocking density (15.4–25.8 m2) with a blind, (3) high stocking density without a blind, and (4) low stocking density without a blind. A total of 127 primiparous heifers and 247 multiparous cows were housed in mixed-parity groups from approximately 3 wk before and immediately after calving. During the 4 h before calving, lying behavior (lying time and bouts) was collected automatically using accelerometers, and social behavior (agonistic interactions, allogrooming, and attention from other cows), stage II labor duration, and frequency of position change during stage II labor were collected using video. Lying behavior was collected with accelerometers. Regardless of treatment, lying time and bouts increased as calving approached. Cows and heifers performed more lying bouts in low stocking density pens compared with high stocking density pens. Agonistic interactions and allogrooming were not different between treatments. Other cows spent more time paying attention to focal animals regardless of stocking density as calving approached, but time spent paying attention was reduced by the presence of a blind during h −2 before calving. The hazard of calving unassisted was greater for cows and heifers in low stocking density pens with a blind compared with all other treatments. Further, animals in pens with a blind tended to change positions fewer times during stage II labor. These results suggest that providing a blind in group maternity pens may improve the calving environment for cows and heifers and, in combination with low stocking density, may reduce the amount of time spent in labor.
A wide variety of maternity housing is used for dairy cattle. On dairy farms in the United States that have designated maternity areas, 64.3% of births reportedly occur in group maternity pens (
). A main advantage to group maternity pens is that dairy producers do not need to disturb and move cows into an individual pen while they are in labor. However, depending on the space and resources available, cows in group maternity pens may not be able to express some of their normal behaviors before giving birth. For example, on pasture or range, many dairy cattle seek a secluded location away from other cows with natural cover (e.g., trees or tall grasses) in which to give birth (
). This behavior is thought to be driven by an instinct to both evade predators and separate from other herdmates to allow for the dam and calf to form a strong bond before re-entering the herd. Although it is not realistic to allow cows to perform all of these behaviors in an indoor setting, this information can be used to help create indoor environments where cows can find some seclusion when giving birth (
Researchers have shown that some indoor-housed cows prefer using a physical barrier or hide at calving when it is provided in the pen and will distance themselves from a neighbor in the pen when kept in pairs (
), but achieving separation could be difficult as pregnant cows may be attracted to peri- and parturient cows. For example, pregnant cows are attracted to the scent of amniotic fluids (
). However, it is unclear whether other cows spend time interacting with parturient cows leading up to calving and whether parturient animals perceive attention from other cows negatively or positively.
One factor that may affect a cow's ability to avoid attention from neighbors in group calving pens is space allowance. Research using transition cows (e.g., 3 wk before to 3 wk after calving) has found that limiting space per cow by increasing stocking density increases agonistic interactions, likely due to competition for resources (
found that when stocking density was increased from 80% to 100% at the feed bunk and lying stalls during the few weeks before calving, cows performed more displacements at the feed bunk and spent less time lying down. Additionally, overstocking feeding spaces (2 cows:1 feeder) during the week before calving almost doubled the number of feed bunk displacements (
). Less work has explored the effect of stocking density in maternity pens around the period before birth, despite it being a time when cows may desire increased space to separate from herdmates.
Cows that are not provided the opportunity to express their natural behavior may experience disturbances during labor. The effect of disturbance from other cows during labor is unclear, but researchers have found that human disturbance can affect behavior and labor progress (
found that cows that were moved from a group pen to an individual pen during late stage I and early stage II labor spent half as much time lying during the hour before calving compared with cows moved during early stage I and before labor. In addition, the duration of stage II labor for these cows was approximately 30 min longer than that for cows moved during the earlier stages. Without disturbance from caretakers, cows become recumbent at the start of stage II labor and remain in this position until birth (
). Providing cows with the opportunity to lie down undisturbed before and during calving may be important to ensure that labor progresses normally.
The objective of this study was to determine the effect of stocking density and provision of a blind in group bedded pack maternity pens on duration of stage II labor and position changes during stage II labor during the 4 h before calving, the social behavior between parturient animals and other cows in the pen (attention from other cows, agonistic behaviors, and allogrooming), and lying behavior. We hypothesized that animals in low stocking density pens with a blind would have the shortest duration of labor with the fewest position changes during labor, the fewest social interactions, and the least attention from other cows during labor compared with those kept in higher stocking densities without a blind.
MATERIALS AND METHODS
The study took place at the William H. Miner Agricultural Research Institute (Chazy, NY) from June 2016 to June 2017 and used the same experimental design and animals as described in
The effect of stocking density and a blind on the behavior of Holstein dairy cattle in group maternity pens. Part I: Calving location, locomotion, and separation behavior.
. All procedures were approved by the William H. Miner Agricultural Research Institute's Animal Care and Use Committee (no. 2016AUR05).
Animals, Housing, and Experimental Design
A total of 374 Holstein dairy cows, including 127 primiparous heifers and 247 multiparous cows, were included in the study [average parity = 1.4 ± 1.4 (mean ± SD); range = 0–6 lactations]. Throughout this paper, we refer to animals that are giving birth for the first time as primiparous cows. Primiparous and multiparous animals were housed together in mixed-parity groups beginning approximately 3 wk [23.6 ± 5.8 d (mean ± SD); range = 4–41] before their calving date. Groups of 1 to 6 animals (mean ± SD = 2.2 ± 1.1) entered each treatment pen on a weekly basis based on their expected calving date. Pens were equipped with 12 individual feeding bins (Calan Broadbent Feeding System, American Calan Inc.) and 1 waterer and were bedded with kiln-dried sawdust. Information on training the animals to use individual feeding bins is outlined in
. Animals were fed a close-up TMR once daily at approximately 0900 h. Additional feed was added to individual feeding bins as needed in 4.5-kg increments to ensure ad libitum access to feed. Within 8 h after calving, cows were moved to a postpartum freestall pen with group feeding.
A complete randomized block design with 2 × 2 factorial arrangement of treatments was used with 2 factors: stocking density and the provision of a blind in the pen. A total of 4 treatments resulted: (1) low stocking density without a blind, (2) low stocking density with a blind, (3) high stocking density without a blind, and (4) high stocking density with a blind. Experimental pens were created by dividing a large sawdust-bedded pack pen into 4 areas using metal gates. The treatment pens were replicated 4 times (replicate 1–4) during the study approximately every 3 mo (mean ± SD: 12 ± 3 wk). This design resulted in a total of 16 pens (4 replicates × 4 treatment pens per replicate). A total of 23.4 ± 5.9 animals (mean ± SD) were enrolled in each treatment pen per replicate. At the start of each new replicate, the location of each treatment pen was moved to a new location in the bedded pack to avoid location bias. All treatment pens were located in all areas of the bedded pack, and pen location was never repeated.
The blind constructed for this study was a 1-sided structure (3.7 m long × 0.6 m wide at the base, narrowing to 2.5 cm at the top × 1.5 m high) fabricated out of 2 plastic road barriers (Traffic Safety Store) filled with water and a plywood extension. The blind was located 4.0 m from the back gate and 3.1 m from the side gate of the pen. Treatment pens containing a blind were outfitted with shade cloth around the gate separating the bedded pack from the transfer alley (1.5 m high × 9.8 wide) and a small metal gate covered with plywood (1.5 m high × 1.2 m long × 7.6 cm wide). Treatment pens without a blind had an unaltered bedded pack space (for a more detailed pen description, see
Different stocking densities were created by altering pen size and keeping a consistent number of animals per pen (6–10). Low stocking density pens contained 154.1 m2 of bedded pack space (lying area dimensions: 7.9 m × 19.6 m lying space), and high stocking pens contained 77.4 m2 of bedded pack space (lying area dimensions: 7.9 m × 9.8 m lying space). Desired lying space was 15.4 to 25.8 m2/cow in the low stocking density treatment (actual mean ± SD: 22.7 ± 5.6 m2/cow) and 7.7 to 12.9 m2/cow in the high stocking density treatment (actual mean ± SD: 11.6 ± 3.6 m2/cow).
Exclusion Criteria
This study is part of a larger experiment that investigated the effects of stocking density and a blind in prepartum bedded pack maternity pens on periparturient dairy cattle. The complete list of exclusion criteria is presented in
The effect of stocking density and a blind on the behavior of Holstein dairy cattle in group maternity pens. Part I: Calving location, locomotion, and separation behavior.
. For all cows that met our inclusion criteria (n = 204; primiparous = 58 and multiparous = 146), data including lying behavior, moves during stage II labor, and duration of stage II labor were assessed.
Stage II Labor Duration and Position Changes During Labor
Individual cows were distinguished using their own markings, which were identifiable from the cameras. A total of 12.7 ± 4.2 animals (n = 204 total) were included per treatment pen per replicate. Video data were reviewed by 2 trained observers to determine the start of rhythmic abdominal contractions (defined as the first time the cows' abdominal muscles contracted and released in a rhythmic motion). Each animal was watched by 1 trained observer. Interobserver repeatability between the 2 observers was tested on a subset of animals (n = 20) and was high (R2 = 0.93). Length of stage II labor was estimated by subtracting the time the first rhythmic abdominal contractions were recorded from the time of calf delivery (
The number of times cows changed positions was recorded from the first abdominal contraction to the time of calf delivery. This behavior was of interest because we predicted that cows may have been more disturbed during labor in high stocking density pens and those without a blind compared with the other treatments. We opted to create a new behavioral variable to estimate how disturbed cows were by looking at their position changes during stage II labor. To our knowledge, no other studies have measured this behavior, although
measured position changes between 3 areas of a pen using a preference test. We chose to define position changes during labor as the cow standing from a lying position and taking at least 4 steps (1 step per foot) before lying down again. We came to this definition because taking at least 1 step per foot would capture an animal turning around or readjusting her body position without changing her location in the pen.
Social Behavior
Video data were collected continuously on all animals during the study using 12 digital video cameras (model IQ511, IQinvision) mounted above the bedded packs (4 cameras per low stocking density pen and 2 cameras per high stocking density pen). A Genetec Security Center (v. 5.2 SR6) IP video surveillance system (Genetec Inc.) was used to continuously record video. Cameras were mounted 4.0 and 6.4 m above the bedded packs on the east- and west-facing walls, respectively, to provide complete visual coverage of each treatment pen. Video data were saved on external hard drives throughout the duration of the study.
From the sample of cows that met the inclusion criteria, a subset of animals (n = 62; primiparous = 30 and multiparous = 32) were selected using stratified random sampling with the SURVEYSELECT procedure in SAS (version 9.4; SAS Institute Inc.) for detailed social behavior analysis. The subset of animals was stratified by replicate (1–4), both factors (stocking density and blind), and parity (primiparous and multiparous; for additional details, refer to
The effect of stocking density and a blind on the behavior of Holstein dairy cattle in group maternity pens. Part I: Calving location, locomotion, and separation behavior.
. A single trained observer recorded all behaviors, with the exception of stage II labor duration, which was recorded by 2 trained observers, using continuous video observation. Behaviors were measured during the 4 h before calving (Table 1). Head butts, chases, threats, and displacements occurred infrequently and thus were combined into a single category referred to as “agonistic behaviors.” For each agonistic interaction, it was recorded whether the focal animal was the actor (initiator) or reactor (receiver).
Table 1Description of continuously recorded social behaviors of the focal and nonfocal cows during the 4 h before calving
Item
Definition
Attention from other cows
While standing, any nonfocal cow (1 or more at a time) standing within close proximity to the focal cow (within 1 head's length) with ears and muzzle focused on any part of the focal cow
The total amount of time parturient animals received attention from other animals in the pen (Table 1) was defined as “attention from other animals.” Attention bout duration (s) began when another cow oriented its body, head, and ears toward the focal cow and was within 1 head length of the focal cow. The bout ceased when the other cow moved its head or body away from the focal cow. If more than 1 other cow was paying attention to the focal animal at a time, it was considered 1 attention bout. In this latter case, the bout began with the other cow that initiated the bout and ended when the last remaining other cow ended the bout.
Social behaviors (allogrooming and agonistic interactions) were summarized as bouts (no./h). Attention from other cows and allogrooming were also summarized as durations (s/h) at the individual animal level. For bout durations, we used a bout criterion of 5 s, whereby the events were compiled as 1 bout when breaks in the performance of a behavior were <5 s (
Behaviour of dairy cows kept in extensive (loose housing/pasture) or intensive (tie stall) environments. III. Grooming, exploration and abnormal behaviour.
). Allogrooming was observed infrequently and was thus summarized as a binary variable before analysis (i.e., the focal animal did or did not perform or receive allogrooming).
Lying Behavior
Lying behaviors were recorded using electronic 3D accelerometers (Hobo Pendant G Acceleration Data Logger, Onset Computer Corporation). Loggers were attached to the medial side of 1 hind leg of the cow using Vet Wrap (Co-Flex, Andover Coated Products Inc.). The logger was changed weekly to alternating legs. Loggers were set to record the g-forces of the y-axis at 1-min intervals. Lying data were processed using the cut point reported by
). Lying data from individual animals were summarized retrospectively by hour relative to the animals' actual calving time. Measures of lying behavior included total time (min/h) and bouts (no./h) during the 4 h before calving. Data loss due to logger malfunction or loss resulted in removal from the lying behavior data set (n = 17 of 204 animals with unassisted calvings).
Statistical Analysis
All analyses were performed using SAS software (version 9.4, SAS Institute Inc.). Raw data were screened for data distribution and outliers using the UNIVARIATE procedure in SAS. Extreme outliers (3 interquartile ranges beyond the first or third quartile of the data set) were checked against video recordings. Pen (n = 16) was considered the experimental unit for all analyses. For all analyses, except labor length (analyzed using a Cox proportional hazard ratio) and allogrooming (analyzed using a logistic regression), data were first summarized by pen and parity (n = 32; primiparous = 16 and multiparous = 16).
Descriptive statistics were generated using the FREQ and MEANS procedures in SAS. Attention from other cows was included in the analysis as a covariate for lying behavior and position changes during stage II labor. Length of stage II labor was included as covariate for social behaviors (attention from other cows and agonistic interactions), lying behaviors (bouts and time), and position changes during labor. Covariates were retained in the model if they were associated with the outcome of interest (P < 0.10). Additional information regarding analysis is described in
The effect of stocking density and a blind on the behavior of Holstein dairy cattle in group maternity pens. Part I: Calving location, locomotion, and separation behavior.
. Parity (primiparous vs. multiparous) was a predictor of interest and was included in all models regardless of significance. Interactions between predictor variables were investigated and kept in the models if significant (P < 0.05). No significant interactions between the fixed effects were detected in the models unless otherwise stated. Model fit was evaluated using Pearson chi-squared or degrees of freedom, when applicable, and the normality of residuals was visually assessed with residual plots. Significance was declared at P < 0.05, and tendencies were declared at P < 0.10. Post hoc analysis was used to determine differences between hours before calving (categorical; −4 to −1), and custom hypothesis tests between hours of interest were created using the LSMESTIMATE statement in SAS. Specifically, we compared the first hour measured (−4 h) with each of the following hours to determine whether behavior changed as calving approached (e.g., h −4 vs. −3, −2, and −1).
Stage II Labor Duration and Position Changes During Labor.
Before statistical analysis was performed, data were pooled into group means by pen (n = 16). To determine the effect of our factors and their interaction on the duration of stage II labor, a survival analysis was performed using a Cox proportional hazard ratio estimate using the PHREG procedure in SAS. The time to event variable for the model was number of minutes from the first abdominal contraction until the calf's hips were fully expelled from the cow (length of stage II labor). Fixed effects in the model included stocking density (high vs. low), blind (yes vs. no), and their interaction. To help with the interpretation of results, we also generated descriptive statistics for labor length between treatments and parities using the MEANS procedure in SAS.
The number of position changes during labor was assessed using a generalized linear mixed model with Gaussian distribution using the GLIMMIX procedure in SAS. A mixed effects model included length of stage II labor as a covariate, stocking density (high vs. low), blind (yes vs. no), parity (primiparous vs. multiparous), and all significant interactions as fixed effects. Replicate (1–4) was included as a random effect.
Social Behavior.
To determine the effect of our factors and their interaction on the attention focal cows received from other cows (duration) and agonistic behaviors (number of bouts), a generalized linear mixed model was performed using the GLIMMIX procedure in SAS. Agonistic behaviors were analyzed multiple ways including interactions performed and received by focal animals (total number of bouts), performed by focal animals (number of actor bouts), and received by focal animals (number of reactor bouts). A mixed effects model included hour before calving (categorical; −4 to −1), stocking density (high vs. low), blind (yes vs. no), parity (primiparous vs. multiparous), and all significant interactions as fixed effects. Length of stage II labor was a covariate when it was associated (P < 0.05) with the outcome. Replicate (1–4) was included as a random effect. Models included a Gaussian distribution. Hour before calving was included as a repeated measure using the RANDOM statement in GLIMMIX. Attention from other cows and agonistic interactions used an autoregressive and variance components correlation structure, respectively.
For agonistic interactions, 1 primiparous animal was determined to be an outlier. The video data were reviewed and the number of agonistic interactions was confirmed to be correct. It is unclear whether the animal's behavior was an anomaly, whether it was affected by the treatment pen in which she was housed (high stocking density with a blind), or whether it was affected by social dynamics within the pen. In an attempt to assess treatment effects while not discounting this animal's behavior, analysis of agnostic interactions was performed with and without the outlier. Removal of the outlier from the data set normalized the residual distribution. For this reason, the outlier animal was removed from the data set for the final analysis reported in this paper.
Too few animals (11/62) experienced allogrooming to determine allogrooming event (bouts and duration) differences between treatments. For this reason, we aimed to determine whether our factors affected the odds that animals experienced (did or did not) allogrooming (performed, received, or both). A logistic regression with a binary distribution and logit link function was performed using the GLIMMIX procedure in SAS. Fixed effects included hour before calving (categorical; −4 to −1), stocking density (high vs. low), blind (yes vs. no), and all significant interactions. Hour before was included as a repeated measure with an autoregressive correlation structure. Replicate (1–4) was included as a random effect.
Lying Behavior.
Lying bouts and duration were analyzed with a generalized linear mixed model and Gaussian distribution using the GLIMMIX procedure in SAS. The mixed effects models included length of stage II labor as a covariate and hour relative to calving (categorical; −4 to −1), stocking density (high vs. low), blind (yes vs. no), parity (primiparous vs. multiparous), and all significant interactions as fixed effects. Hour before calving was included as a repeated measure with an autoregressive correlation structure. Replicate (1–4) was included as a random effect.
Sample size for labor length and behavior was determined using previous studies that looked at similar outcomes (
Short communication: Relationship between competitive success during displacements at an overstocked feed bunk and measures of physiology and behavior in Holstein dairy cattle.
), as no studies to our knowledge have measured the same outcomes in response to our treatments. We estimated needing approximately 60 animals total within the 16 pens (experimental unit: pen = replicate × stocking density × blind). As this was also part of a larger study (
), we had more animals than were needed for labor length, so we decided to measure this outcome in all available animals that met our inclusion criteria (n = 204).
RESULTS
Stage II Labor Duration and Position Changes During Labor
Stage II labor length ranged from 20 to 233 min (median = 79 min; low stocking density with a blind: 73.8 ± 30.1 min; low stocking density without a blind: 89.4 ± 46.6 min; high stocking density with a blind: 79.7 ± 38.7 min; high stocking density without blind: 82.2 ± 35.4 min; raw mean ± SD). A greater hazard of calving indicates a shorter duration of stage II labor. Hazard of calving was not affected by stocking density in pens with or without a blind [hazard ratio (95% CI) = 0.5 (0.1–2.0) and 1.3 (0.3–5.7), respectively; P = 0.33; Figure 1]. Animals in pens with a blind had greater hazard of calving (indicating shorter labor) in low stocking density pens but not high stocking density pens [hazard ratio (95% CI) = 0.1 (0.02–0.7) and 0.4 (0.1–1.8), respectively; P = 0.02]. Stage II labor duration was approximately 25 min longer for primiparous animals than for multiparous animals (mean ± SD: 98.7 ± 38.8 vs. 74.6 ± 35.9 min).
Figure 1Stage II labor duration (from the first rhythmic abdominal contraction to the time the calf's hips were fully expelled) survival curve for Holstein dairy cows and heifers kept in indoor group maternity pens at high (7.7–12.9 m2 of lying space/cow) or low (15.4–25.8 m2 of lying space/cow) stocking density and with or without a blind. This study had 16 experimental units (4 replicates × 4 treatments per replicate).
Animals in pens with a blind tended to change position fewer times during stage II compared with those in pens without a blind (LSM ± SE: 3.1 vs. 4.2 ± 0.4; F1,13 = 3.6; P = 0.08). We did not detect a difference in the number of position changes during stage II labor between high and low stocking densities (3.4 vs. 3.9 ± 0.4; F1,13 = 1.2; P = 0.29). Multiparous animals performed more position changes during stage II labor than primiparous animals (4.6 vs. 2.7 ± 0.4; F1,14 = 9.0; P = 0.01). Animals with a longer duration of stage II labor made more position changes labor (estimate = 0.05, SE = 0.01; F1,14 = 16.54; P < 0.001).
Social Behaviors
A total of 94% of focal cows (58/62) received attention from other cows during the 4 h before calving. A blind × hour interaction was detected for the amount of time (s/h) focal animals received attention from other cows (F3,55 = 3.3; P = 0.04; Figure 2). Animals in pens without a blind received more attention from other cows −2 h before calving (P = 0.01), but there was no difference in attention received during −3 and −1 h before calving (P = 0.53 and P = 0.16, respectively). However, there was no effect of stocking density (F1,55 = 0.2; P = 0.63) or parity (F1,58 = 0.2; P = 0.70).
Figure 2Least squares means (±SE) of the amount of time (s/h) parturient Holstein dairy cattle received attention from other cows during the 4 h before calving in (A) high or low stocking density pens and (B) pens with or without a blind. Animals were housed in indoor group maternity pens at high (7.7–12.9 m2 of lying space/cow) or low (15.4–25.8 m2 of lying space/cow) stocking density and with or without a blind. This study had 16 experimental units (4 replicates × 4 treatments per replicate). Asterisk indicates a difference between treatments (P < 0.05).
No differences were detected in agonistic interactions (total no./h) between high and low stocking density pens (F1,58 = 0.02; P = 0.89; Figure 3), pens with and without a blind (F1,58 = 0.2; P = 0.65), or parities (F1,58 = 0.2; P = 0.63). Total agonistic interactions varied by hour before calving (F3,58 = 3.7; P = 0.01). Using a post hoc analysis, we found that more agonistic (total no./h) interactions occurred during −2 than −4 h before calving (P = 0.03); there were no differences between −3 versus −4 h (P = 0.02) or −1 versus −4 h (P = 0.30) before calving. Agonistic interactions performed by focal animals (actor no./h) were not different between high and low stocking density pens (F1,58 = 0.4; P = 0.54), pens with and without a blind (F1,58 = 0.01; P = 0.91), or parities (F1,58 = 0.03; P = 0.85). Agonistic interactions performed by focal animals tended to vary by hour before calving (F1,58 = 2.4; P = 0.08). Post hoc analysis found that focal animals tended to perform more agonistic interactions during −2 than −4 h before calving (P = 0.054), but number of agonistic interactions was not different between −3 versus −4 (P = 0.28) or −1 versus −4 h before calving (P = 0.66). The number of agonistic interactions received by focal animals (reactor, no./h) did not vary between high and low stocking density pens (F1,58 = 0.5; P = 0.48), pens with and without a blind (F1,58 = 0.5; P = 0.47), parities (F1,59 = 2.4; P = 0.13), or hour before calving (F3,58 = 1.9; P = 0.14).
Figure 3Least squares means (±SE) of total agonistic interaction bouts (A and B; total no./h), agonistic interactions performed by the focal animal (C and D; actor, no./h), and agonistic interactions received by the focal cow (E and F; reactor, no./h) during the 4 h before calving for Holstein dairy cattle housed in indoor group maternity pens at high (7.7–12.9 m2 of lying space/cow) or low (15.4–25.8 m2 of lying space/cow) stocking density and with or without a blind. This study had 16 experimental units (4 replicates × 4 treatments per replicate).
During the 4 h before calving, only 17.7% of animals (11/62) performed or received allogrooming (6/31 high stocking density, 5/31 low stocking density, 4/32 no blind, 7/30 blind, 4/30 primiparous, 7/32 multiparous). Allogrooming events were infrequent, ranging from 0 to 10 bouts (mean = 0.5, median = 0, SD = 1.5) during the 4 h before calving. Time spent allogrooming was also low, ranging between 0 and 228 s (mean = 15.6, median = 0, SD = 40.5) per animal. The likelihood of experiencing allogrooming was not different between high or low stocking density pens [odds ratio (95% CI): 1.7 (0.5–5.8); F1,58 = 0.9; P = 0.35], pens with or without a blind [odds ratio (95% CI): 0.4 (0.1–1.3); F1,58 = 2.4; P = 0.12], or hour before calving [odds ratio (95% CI): −4 to −3 h, 0.3 (0.1–1.7); −4 to −2 h, 0.6 (0.2–2.5); −4 to −1 h, 0.2 (0.02–1.4); F3,58 = 1.4; P = 0.28].
Lying Behavior
There was no difference in lying time between high or low stocking density pens (F1,58 = 2.4; P = 0.15; Table 2) or pens with or without a blind (F1,58 = 2.1; P = 0.17). An hour by parity interaction was detected (P < 0.001), whereby primiparous animals spent less time lying than multiparous animals at −4 (20.6 vs. 31.9 ± 2.1 min/h; F3,59 = 6.4; P < 0.001) and −3 h (21.8 vs. 30.9 min/h; P = 0.005) before calving, tended to spend less time lying −2 h before calving (25.7 vs. 31.7 min/h; P = 0.06), and tended to spend more time lying −1 before calving (45.3 vs. 39.6; P = 0.07). Regardless of treatment, lying time (F3,58 = 27.1; P < 0.001) increased before calving (−4 vs. −3 h, P = 0.96; −4 vs. −2 h, P = 0.28; −4 vs. −1 h, P < 0.001). Animals with longer durations of stage II labor (included as a covariate) spent more time lying (min/h) during the 4 h before calving (estimate = 0.09, SE = 0.03; F1,59 = 0.5; P = 0.01).
Table 2Least squares means (±SE) for lying behavior (lying time and bouts) of dairy animals (n = 187) during the 4 h before calving in group bedded pack maternity pens
Animals were housed in indoor group maternity pens at high (7.7–12.9 m2 of lying space/cow) or low (15.4–25.8 m2 of lying space/cow) stocking density and with or without a blind. This study had 16 experimental units (4 replicates × 4 treatments/replicate).
Item
Factor
Treatment
Hour before calving
SE
P-value
−4
−3
−2
−1
Lying time (min/h)
Stocking density
High
24.1
24.7
27.9
42.4
2.2
0.15
Low
28.3
28.0
29.5
42.5
Blind
Yes
25.0
23.6
28.2
42.4
2.2
0.17
No
27.4
29.1
29.2
42.5
Lying bouts (no./h)
Stocking density
High
1.2
1.2
1.8
2.0
0.1
0.02
Low
1.5
1.5
1.9
2.3
Blind
Yes
1.3
1.3
1.7
2.1
0.1
0.26
No
1.5
1.4
1.9
2.1
1 Animals were housed in indoor group maternity pens at high (7.7–12.9 m2 of lying space/cow) or low (15.4–25.8 m2 of lying space/cow) stocking density and with or without a blind. This study had 16 experimental units (4 replicates × 4 treatments/replicate).
Animals in low stocking density pens performed more lying bouts than those in high stocking density pens (F1,58 = 5.8; P = 0.02; Table 2); however, no difference in lying bouts was detected between pens with and without a blind (F1,58 = 1.3; P = 0.26). Multiparous animals had more lying bouts than primiparous animals (LSM ± SE: 2.0 vs. 1.3 ± 0.1; F1,62 = 43.9; P < 0.001). Regardless of treatment, the number of lying bouts (no./h) increased with time (F3,58 = 15.6; P < 0.001) before calving (−4 vs. −3 h, P = 0.7; −4 vs. −2 h, P = 0.002; −4 vs. −1 h, P < 0.001). Animals with longer durations of stage II labor (included as a covariate) had more lying bouts (estimate = 0.006, SE = 0.002; F1,62 = 7.2; P = 0.01).
DISCUSSION
The objective of this study was to investigate the effects of lying space and a blind in group maternity pens on social and lying behavior during the 4 h before calving, duration of stage II labor, and position changes during labor. Providing cows and heifers with a blind and more space reduced the duration of stage II labor and tended to reduce position changes during labor. Regardless of treatment, there was a low occurrence of social behaviors including agonistic interactions and allogrooming in the few hours before calving. Social behaviors, including attention from other cows and agonistic behaviors, and lying behavior (time and bouts) were most affected by the time relative to calving and not treatment.
Length of stage II labor was shorter for animals in pens that contained a blind compared with animals in pens that did not and was shortest for those in low stocking density pens with a blind compared with the other treatment pens. Additionally, animals tended to change position fewer times during stage II labor in pens that contained a blind compared with those without a blind. However, it is not clear whether the reduction in labor length was biologically relevant. Reducing the amount of time animals are in stage II labor may benefit the cow and calf as prolonged labor is associated with dystocia (
found that cows moved to individual maternity pens during late stage I labor have prolonged labor durations and higher standing time during labor; the authors speculated that human disturbance and the novelty of a new pen could have contributed to a disruption of normal labor. The blind and additional space may have provided cows with a greater feeling of safety during labor, although it may not have protected the cows from attention from others. Despite no difference in attention from other cows,
The effect of stocking density and a blind on the behavior of Holstein dairy cattle in group maternity pens. Part I: Calving location, locomotion, and separation behavior.
found that many cows and heifers calved using a blind when it was available and were able to physically distance themselves from other cows more in low stocking density pens compared with high stocking density pens. The combination of visual separation from penmates with a blind and physical distance from penmates in low stocking density pens may have contributed to less overall disturbance during stage II labor.
The majority of focal cows received attention from other cows during the 4 h before calving (94%). Parturient cows in pens without a blind received more attention from other cows than those with a blind 2 h before calving, but no other treatment differences were detected. Limited treatment differences may imply that our reduction in stocking density and provision of a blind were insufficient to offer parturient animals protection from disturbances by other cows leading up to and during labor. Attention from other cows increased from approximately 9 s to 4 min from −4 to −1 h before calving, but it should be noted that, overall, this represented a short time frame within each hour. Other cows were likely attracted to the smells of the newborn and amniotic fluid as calving approached (
). A limitation with this measurement is that we were unable to determine whether 1 or more animals were paying attention to the focal cow at the same time. There may have been differences between our treatments had we also included the number of animals that performed this behavior at the same time, and this may be interpreted differently from the focal cow's perspective. As this was the first study to our knowledge to measure attention from other cows, we encourage more research to help determine cows' and heifers' perception of attention from other animals before calving.
Agonistic interactions between focal and other cows were not different between treatments. In a study of commercial compost bedded pack barns,
found that lactating cows performed approximately 3.5 agonistic behaviors (including chase, push, and head butt) per hour, which is much higher than the number of agnostic behaviors observed in our study (approximately 0.3–1.7 bouts/h). The authors included farms with variable stocking densities (3.5–10.8 m2 of bedded space/cow), which was consistent with our low stocking density pens. It is unclear why we found much fewer agonistic behaviors, but this may be driven by the time periods that were assessed;
measured agonistic behavior during 4-h periods shortly after milking, when agonistic behavior may have been highest. In contrast, our measurements occurred throughout the day and night, when activity may not have been as high. Additionally, stage of lactation may have played a role in the differences between the studies (mid lactation vs. preparturient period). Cows may have had lower overall agonistic interactions as calving approached compared with other stages of lactation, as they distance themselves from other animals at calving (
The effect of stocking density and a blind on the behavior of Holstein dairy cattle in group maternity pens. Part I: Calving location, locomotion, and separation behavior.
found that cows spent less than 100 s within 4 m of any other cow on the day of calving, which was less than the 50 d before and after calving.
Interestingly, we found that parturient heifers and cows tended to perform more agonistic interactions during −2 compared with −4 h before calving. It is unclear why parturient cows and heifers increased their performance of agonistic behaviors in the 2 h before calving. However, we speculate that cows and heifers may have had higher motivation to protect themselves and their newborns as calving approached, as maternal aggression is common in cattle after birth (
). Additionally, aggressive behaviors may have been driven by pain associated with labor, as other mammals have been found to direct aggressive behaviors toward their conspecifics following painful experiences (
); thus, it is possible that parturient dairy cattle were frustrated by an inability to find an undisturbed calving location or attention from other cows, resulting in greater agonistic behaviors. Further research assessing agonistic behaviors in dairy cows during labor is needed.
Allogrooming was found to be a rare event during the 4 h before calving, experienced (performed, received, or both) by only 18% of the focal cows and heifers. Our records of allogrooming are lower than what has been previously reported in lactating dairy cows (0.5 bout/4 h vs. 2 bouts/h in a bedded pack,
). Social licking, or allogrooming, in cattle is an important part of maternal care and is considered to be positive social contact between adult animals (
). The likelihood of allogrooming was not affected by proximity to calving or our treatments, perhaps due to the low occurrence of the behavior. Similar to the prediction for agonistic behaviors, it is possible that focal cows were preoccupied with finding a place to calve or physically distanced from other cows, reducing the likelihood they would engage in social licking with other cows.
Cows and heifers performed more lying bouts in low stocking density pens compared with high stocking density pens; however, the difference was minimal (approximately 0.3 bout/h or less) and may not be biologically relevant. We did not detect any effect of the blind on the number of lying bouts. Lying bouts observed in this study are similar to the number of lying bouts observed in individual maternity pens (approximately 3/h;
) before calving. It is unclear why stocking density may have affected lying bouts; however, greater space allowance may have facilitated the performance of more behaviors. Some suggest that lying bouts before calving are an indicator of pain associated with labor (
found that cows in bedded packs had higher lying bouts compared with those in freestalls, where movement may be more restricted. Although our treatments changed the pen design (7.7–12.9 m2 vs. 15.4–25.8 m2 of lying space/cow and blind vs. no blind), all cows were able to lie down at the same time, but perhaps more space in the low stocking density pens allowed for more opportunities for cows to change positions.
Cows and heifers increased their lying time and lying bout number as calving approached. This increase in lying time and lying bouts before calving is similar to previous studies (
). We also observed that animals with longer stage II labor had more lying bouts and lying time; however, we are unable to determine cause-and-effect relationship. An increase in lying bouts and lying time may indicate discomfort during prolonged labor (
Lying behavior also varied between heifers and cows during the 4 h before calving. Heifers performed more lying bouts during the 4 h before calving, spent less time lying during −4 and −3 h before calving, and tended to spend more time lying −1 h before calving compared with cows. In another study using group bedded pack maternity pens,
found that heifers spent less time lying and had more lying bouts than cows during −4 to −1 d before calving, which was similar to our study, but on the day of calving there were no parity differences. Our results likely differ from those of these authors because we specifically investigated the few hours before calving, when behavioral differences between cows and heifers may have been the greatest. Higher lying bouts and lower lying time in heifers before −1 h of calving align with our findings that heifers had more locomotor behavior before calving compared with cows (
The effect of stocking density and a blind on the behavior of Holstein dairy cattle in group maternity pens. Part I: Calving location, locomotion, and separation behavior.
). It is unclear why heifers tended to have longer lying time during the hour before calving; we speculate that because heifers had longer labor, they may have spent more of their time during the hour before calving in lateral recumbency compared with cows, potentially due to fatigue.
One of the limitations of this study is that we only assessed animals with unassisted calvings; therefore, the data interpretation only applies to animals that calved on their own. It is unclear how our treatments would have affected a cow's risk of assistance during labor or difficult birth. Thus, we encourage future research to determine the effect of features in group maternity pens on the risk of calving assistance and difficult births, including dystocia.
CONCLUSIONS
Animals in low stocking density pens with a blind had the shortest stage II labor length compared with all other pens. Parturient cows tended to change position fewer times during labor when a blind was provided. Lying and the majority of social behaviors were not affected by differences in stocking density or the provision of a blind; however, these behaviors varied based on the proximity to calving. One exception was that cows provided a blind received less attention from other cows in the pen 2 h before calving compared with those in pens without a blind. The addition of a blind in group maternity pens may improve the calving environment, but additional steps should be taken to offer a calving site that is free of disturbances by other cows.
ACKNOWLEDGMENTS
We sincerely thank the research and farm staff of the William H. Miner Agricultural Research Institute (Chazy, NY) for their help with animal care and data collection throughout the study; without them this project would not have been possible. This study was funded by USDA-NIFA (Washington, DC) grant no. 2016-67015-24734. The authors have not stated any conflicts of interest.
The effect of stocking density and a blind on the behavior of Holstein dairy cattle in group maternity pens. Part I: Calving location, locomotion, and separation behavior.
Short communication: Relationship between competitive success during displacements at an overstocked feed bunk and measures of physiology and behavior in Holstein dairy cattle.
Behaviour of dairy cows kept in extensive (loose housing/pasture) or intensive (tie stall) environments. III. Grooming, exploration and abnormal behaviour.
In a more natural setting, dairy cows separate from herdmates and seek a secluded area to give birth. However, on many dairy facilities, cows calve in barren group pens with limited space, which may limit their ability to perform these behaviors. The objective of this study was to determine the effect of stocking density and provision of a blind (a single-sided solid partition) in group bedded pack maternity pens on the behavior of dairy cattle. The study included 4 experimental treatments: (1) high stocking density (7.7–12.9 m2 of lying space/cow) with a blind, (2) low stocking density (15.4–25.8 m2) with a blind, (3) high stocking density without a blind, and (4) low stocking density without a blind.
00454-9&id=gr1.jpg){kind=link}
00454-9&id=gr2.jpg){kind=link}
00454-9&id=gr3.jpg){kind=link}