| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
ANIMAL PRODUCTION |
,2,3,
,,
* Departament de Ciència Animal i dels Aliments, Universitat Autònoma de Barcelona, 08193 Bellaterra, Spain;
and
Animal Nutrition, Management, and Welfare Research Group, 08193 Bellaterra, Spain;
and
Unitat de Remugants-Institut de Recerca i Tecnologia Agroalimentàries, 08140 Caldes de Montbui; and
Institució Catalana de Recerca i Estudis Avançats, 08010 Barcelona, Spain
 |
Abstract |
|---|
 Top Abstract INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
0.16) but the pen’s maximum concentration responded quadratically (P < 0.001), being greatest in T8, with dominant heifers being the most affected. Serum haptoglobin concentration increased linearly (P = 0.05) with competition, particularly within the most subordinate heifers. Increased competition reduced (P < 0.05) ruminal pH only in periods 1 and 2 and increased ruminal lactate (P = 0.02). Increasing competition at the concentrate feeders increased the variability in final BW but performance was not affected. Detrimental effects on animal welfare might be deduced from the altered feeding behavior, reduced resting time, and increased aggression. Ruminal lactate and blood haptoglobin indicate that the risk of rumen acidosis might increase with competition, whereas liver abscesses increased at 8 heifers per feeder.
Key Words: behavior • competition • heifer • ruminal acidosis • welfare
|
INTRODUCTION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). However, Gonyou and Stricklin (1981) reported a complete loss of daily feeding patterns when 15 beef cattle per feeder were fed a 55% concentrate diet. Competition among pen mates also increases aggression, avoidance, and intimidation behaviors, perhaps leading to social stress (Miller and Wood-Gush, 1991). Consequently, animal welfare and production may also decrease when competition for feed increases. Low-ranking animals are more often interrupted while feeding, have shorter feeding times, and have a greater frequency of visits compared with high-ranking animals (Stricklin and Gonyou, 1981; Huzzey et al., 2006). Eating time decreases and eating rate increases with competition (Nielsen, 1999; Olofsson, 1999), with subordinate cows exhibiting up to a 3-fold increase (Harb et al., 1985). Therefore, increased competition in the feeders might disrupt the normal or preferred feeding patterns, limiting the ability of cattle to self-regulate their digestive function. These observations led some researchers to suggest that reduced feeding space might be a risk factor for ruminal acidosis (Stone, 2004; Krause and Oetzel, 2006) and, consequently, of liver abscesses (Nagaraja and Chengappa, 1998). Subacute ruminal acidosis (SARA) results in reduced and variable feed intake and performance (Britton and Stock, 1989; Schwartzkopf-Genswein et al., 2003). Ruminal pH and the profile of fermentation products are commonly used as indicators of SARA, but acute phase proteins, such as haptoglobin, may also be used as markers (Gozho et al., 2006; Nagaraja and Titgemeyer, 2007). The objective of the present study was to determine the long-term effects of increasing the number of heifers at each place in the concentrate feeders on performance, behavior, welfare, and rumen function. |
MATERIALS AND METHODS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
All experimental procedures were approved by the Animal Care and Use Committee of the Institut de Recerca i Tecnologia Agroalimentàries.
The present study is the continuation of another that assessed the adaptability of newly received calves in the feedlot under the same treatments. The design of the facilities, animals and feeding management, and analytical methods were previously reported by González et al. (2008). All animals remained in the same pen with the same pen mates assigned at arrival. However, 1 heifer in the treatment of 8 heifers per concentrate feeder (T8) was replaced at the end of the 4-wk arrival period by a heifer with the same BW as the mean of the pen because it had failed to adapt to eating concentrate (a "noneater"). A second noneater heifer under T8 was stimulated for several days to eat concentrate and recovered a modest ADG after the arrival period.
Briefly, 72 Friesian heifers were assigned to a factorial arrangement of treatments in a randomized complete block design with 3 treatments and 3 BW blocks. Treatments consisted of 2 (T2), 4 (T4), or 8 (T8) heifers per concentrate feeder. Thus, each of the 9 experimental pens contained 8 heifers, and the number of concentrate feeding places per pen was 4, 2, and 1 for T2, T4, and T8, respectively. Details of the housing conditions, feeders, and videorecording system are described in González et al. (2008).
Performance and Intake
All measurements began after an arrival adaptation period of 4 wk. Heifers were weighed for 2 consecutive days after this adaptation period and also before slaughter. The experiment consisted of 28-d periods. Intermediate BW were taken after withdrawal of refusals on d 1 of each experimental period for the calculation of ADG and G:F ratio. The within-pen SD of BW was calculated to assess the variability of growth among heifers sharing the same pen. All animals were fed up to 380 kg of slaughter BW, which required 6, 7, and 8 periods for the low-, medium-, and high-BW blocks, respectively. Animals were slaughtered in a commercial abattoir, where HCW and the number of abscessed livers were registered. Livers were considered as abscessed when 1 or more pus-filled, encapsulated (active) abscess was present, independently of their size (Nagaraja and Chengappa, 1998).
All heifers received the same commercial concentrate and barley straw. The concentrate was formulated to meet NRC (1996) requirements and contained 90.6% DM, 16.6% CP, 19.2% NDF, 5.3% ether extract, 5.9% ash, and 53.0% nonstructural carbohydrates, on a DM basis. Ingredients of the concentrate (DM basis) were 31.0, 29.0, 15.0, 14.0, 6.6, 2.1, and 0.9% of barley, corn, corn gluten feed, soybean meal, soybean hulls, palm oil, and sodium bicarbonate, respectively, with the remainder being minerals and vitamins. The straw contained 90.7% DM, 6.7% ash, 73.4% NDF, 43.8% ADF, 8.7% ADL, and 4.84% CP, on a DM basis. Feeding management allowed ad libitum consumption of both components, which were fed daily at 0830 h. Fresh water was available at all times. One composited sample of the offered concentrate and straw was taken during each experimental period for DM determination and chemical analysis. Straw and concentrate DMI were calculated weekly by weighing the amount of feed offered each day and subtracting the amount refused at the end of the week. However, intake data were pooled for each 28-d period for statistical analyses.
Maintenance Behavior
Behavior of the animals was recorded for 24 h on d 16 and 17 of experimental periods 1, 3, and 6. Videotapes were processed by continuous recording of the activities performed by all animals. Recorded activities (time spent eating concentrate and straw, drinking, lying, and standing) were registered together with their beginning and ending times (to the nearest second) and the animal identification (González et al., 2008). Eating rates of concentrate and straw were calculated as the daily average of the pen DMI during the week divided by the average sum of the time spent eating by all animals in the pen.
Social Behavior
Displacements among calves from concentrate feeders, straw feeders, and water bowls were the events recorded. They were registered with the time of occurrence, the animal identification, the type of event, and the activity being performed when the event occurred. The criteria used to register aggressive behavior and the construction of hierarchy matrices were reported in a previous paper (González et al., 2008). Dominance order was assessed by calculating the angular dominance value (ADV) as the arcsine square root transformation of the average proportion of time that the individual displaced each pen mate (González et al., 2008).
Blood, Fecal, and Rumen Fluid Sampling
Sampling was performed on d 23, 24, and 25 of each experimental period for the low-, medium-, and high-BW blocks, respectively. All samples were taken at 1630 h (8 h after feeding) because, under this management, ruminal pH usually reaches the nadir at this time (Nagaraja and Titgemeyer, 2007). One whole-blood sample from each heifer was taken by jugular venipuncture (10-mL Vacutainer tubes, Becton Dickinson, Plymouth, UK). Serum was separated within 4 h (3,000 x g, 20 min, 4°C) and stored at –20°C until analyses of haptoglobin, β-hydroxybutyrate (β-HBT), and NEFA were completed. Fecal samples were taken from the rectum of each heifer and frozen at –20°C until analyses of glucocorticoid metabolites (GM) to assess the adrenal response of the heifers. Approximately 10 mL of ruminal liquid was taken by rumenocentesis after skin disinfection with iodine and use of a sterile i.v. catheter needle (Abbocath-T 14-gauge x 140 mm, Abbott, Sligo, Ireland). A 4-mL aliquot of ruminal fluid was subsampled and frozen at –20°C until chemical analysis of organic acids. The pH of the remaining ruminal fluid was measured immediately with a glass electrode pH meter (model 507, Crisson Instruments SA, Barcelona, Spain). Blood and ruminal fluid analyses were conducted for all experimental periods, but fecal GM was measured only in periods 1, 3, and 6, when behavior was also recorded.
Chemical Analyses
One portion of the offered feed was analyzed for DM content for 24 h at 103°C in a forced-air oven according to AOAC (1990; method 950.01). Another portion of feed samples was dried in a forced-air oven at 65°C for 48 h, ground in a hammer mill through a 1-mm screen, and stored for later chemical analysis. Samples were analyzed for DM and ash (AOAC, 1990; method 950.05). Nitrogen content was determined by the Kjeldahl procedure (AOAC, 1990; method 976.05). Ether extract analysis was performed according to AOAC (1990; method 920.39). The NDF and ADF contents were determined sequentially following the procedure of Van Soest et al. (1991) by using a thermostable 
-amylase (Ankom Technology, Macedon, NY) and sodium sulfite (Panreac Química S.A., Castellar del Vallès, Barcelona, Spain).
Fecal GM determinations were performed by using a commercially available 125I RIA kit (Rats and Mice Corticosterone kit, ICN Pharmaceuticals, Orangeburg, NY), as described by Morrow et al. (2002) and González et al. (2008). The intra- and interassay CV of the RIA were 12.2 and 15.6%, respectively. The estimated detection limit of GM in feces was 5.7 ng/g of DM. Haptoglobin was determined by the hemoglobin binding method with the use of a commercial haptoglobin assay (intra-and interassay CV of 1.4 and 6.9%, respectively; Assay Phase Range, Tridelta Development Limited, Maynooth, Ireland); D-3-hydroxybutyrate was determined by a kinetic enzymatic method (intra- and interassay CV of 3.7%; Ranbut D-3-hydroxybutyrate, Randox Laboratories Ltd., Crumlin, UK); and NEFA was determined by the colorimetric enzymatic test ACS-ACOD method (intra- and interassay CV of 2.7%; NEFA C, Wako Chemicals, Neuss, Germany).
Volatile fatty acids and lactic acid concentrations of ruminal fluid were analyzed through a modification of the capillary gas chromatography method described by Richardson et al. (1989). Modifications consisted of sample conservation and preanalysis treatment. To preserve the sample, 4 mL of ruminal fluid was added to 1 mL of a solution made up of 1% (wt/wt) mercuric chloride (Panreac Química S.A.), 2% (vol/vol) orthophosphoric acid (Panreac Química S.A.), and 0.2% (wt/ wt) 4-methylvaleric acid (Sigma Chemical Co., Steinheim, Germany) as an internal standard in distilled water and frozen at –20°C. As a preanalysis treatment, samples were thawed and centrifuged at 15,000 x g for 15 min and diluted 1:1 in distilled water.
Statistical Analyses
All statistical analyses were carried out by considering the 6 consecutive experimental periods when the 3 BW blocks were present, except for final BW, which was measured before slaughter. A logarithmic transformation was applied to blood haptoglobin and fecal GM concentrations. The same was applied to the ruminal lactic acid concentration (+1 because of the presence of zero concentrations in some animals). All individual data were averaged to give pen means at each sampling point over time. Pen was the experimental unit for all statistical analyses (n = 3). Normally distributed variables were analyzed by a mixed-effects regression model by using PROC MIXED (SAS Inst. Inc., Cary, NC). The model contained the fixed linear and quadratic regression coefficient of the number of heifers per feeding place (treatment), BW block, linear and quadratic treatment x period, and block x period interactions. The random effects were modeled through the correlations among the repeated measure of time (period or time interval) subjected to pen, and through the random effect of pen. To analyze the eating patterns within day and among periods, double-repeated measures were considered (time interval of the day subjected to the period nested within pen). The same effects were used in a Poisson regression model to analyze the number of displacements per pen and day, and in a logistic regression model to analyze the proportion of animals showing liver abscesses and ruminal pH below the threshold of 5.6. In both cases, the GLIMMIX procedure of SAS was used. An analysis of the social hierarchy was undertaken to study the trends of GM and haptoglobin of each social category with increasing competition. Each heifer within a pen was classified as very dominant (the 2 heifers at the top of the social hierarchy showing the greatest ADV), dominant, subordinate, and very subordinate (the 2 heifers showing the lowest ADV, and therefore being the most subordinate of the group). A mixed-effects regression model was also used for this purpose, where the correlations were modeled at 3 levels of random effects: the repeated measure of time, the individual animal nested within the pen, and the pen nested within treatment, with all animals within a pen being correlated (St-Pierre, 2007). This model contained the fixed categorical effects of period, BW block, and dominance category of each heifer, plus the linear and quadratic effects of the number of heifers per concentrate feeder and the appropriate 2- and 3-way interactions. In all models, the interaction between the linear and quadratic treatment by any of the categorical effects (period, time of day, dominance category) tested the null hypothesis that regression coefficients of the number of heifers per feeder were equal among all levels of the categorical effect. If this interaction was significant, then regression coefficients at each level were estimated and tested for their difference from zero. The choice of the best covariance structure was based on fit statistics, where the model that minimized either Akaike’s information criterion corrected or Schwarz’s Bayesian information criterion was preferable. Significance was declared at P 
0.05 and tendencies were discussed at 0.05 < P 0.10 unless otherwise noted.
|
RESULTS |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
At the start of the current study, initial BW showed a quadratic response (P < 0.01) to increasing the number of heifers per concentrate feeder as a consequence of treatment effects during the adaptation period, being lowest in T8 (Table 1). The initial BW was used as a covariate to test for carryover effects on all variables tested, but it was not significant in all the variables presented (P > 0.10) and was therefore eliminated. However, this statistical difference was not maintained until the end of the experiment (quadratic P = 0.12). The within-pen SD of final BW tended to increase linearly as social pressure in the concentrate feeders increased (P = 0.06). There were no treatment effects on ADG, within-pen SD of ADG, or G:F ratio (P > 0.10). Concentrate DMI decreased linearly (P = 0.05) with an increasing number of heifers per concentrate feeder, but no effects were observed on straw or total DMI (P > 0.10). There were no treatment x period interactions in any performance or intake characteristics (P > 0.05). Hot carcass weight was not affected (P = 0.36), but the percentage of abscessed livers followed a quadratic pattern as the number of heifers per feeder increased, being greatest in T8 (P = 0.03; Table 1).
|
The main effect of treatments on maintenance behavior is presented in Table 2. The average time spent eating concentrate per day decreased linearly (P < 0.001) as social pressure in the concentrate feeders increased. This resulted in a linear increase (P = 0.05) in concentrate eating rate as the number of heifers per feeder increased. The intrapen SD of concentrate eating time responded quadratically (P = 0.001), with the T4 and T8 groups showing the greatest variability among heifers sharing the same pen. The average time spent eating barley straw was 2 times greater than that spent eating concentrate, and showed a quadratic response (P = 0.03) as competition increased, with the lowest time in T2 pens. The intrapen variation of time at the straw feeders responded quadratically (P < 0.001) as the number of heifers per concentrate feeder increased, with the greatest homogeneity between pen mates in T2. Regardless of treatment, time spent eating straw increased from 97.8 ± 3.6 min/d in period 1 to 115.8 ± 5.7 min/d in period 6 (data not shown). However, animals under T2 did not show an increase in the time spent eating straw, whereas those under T4 and T8 did from period 1 to 6 (P < 0.05; data not shown). Barley straw eating rate and time spent drinking water were not affected by treatments (P > 0.10). The time spent lying down decreased linearly (P = 0.02) as the number of heifers per concentrate feeder increased, although it also showed a quadratic response (P = 0.005). In contrast, the time spent standing increased linearly (P < 0.01). The period main effect (P < 0.05) indicated that concentrate eating rate, time spent eating straw, straw eating rate, time spent drinking water, and time spent standing increased as the age of the heifer increased, whereas time spent eating concentrate and lying down decreased (data not shown). Finally, the period x treatment interaction for concentrate eating time (P = 0.001) and rate (P = 0.05), and time spent eating straw (P < 0.05) indicated that the treatment differences became greater as heifers grew (data not shown).
|
. The linear treatment effect (P = 0.55) was not significant, but the linear treatment xtime interval of day (P = 0.01) demonstrated that linear regression coefficients of competition level were different among different times of the day. This was the result of linear coefficients being different from zero (P < 0.05) during some time intervals but not during others (P > 0.10), as shown in Figure 1a. Accordingly, the time spent eating concentrate decreased linearly (P < 0.05) between 0700 to 0800 h, 0800 to 0900 h, 1200 to 1300 h, and 1700 to 1800 h as the number of heifers per feeder increased. The quadratic treatment main effect (P = 0.01) and the quadratic treatment x time interval interaction (P = 0.14) indicated that, regardless of the time of day, the time spent eating concentrate responded quadratically, but quadratic regression coefficients were different among time intervals. In fact, the average concentrate eating time per animal was 1.92, 2.08, and 1.82 ± 0.16 min/h for T2, T4, and T8, respectively, and the quadratic regression coefficient was significant at 1100 h (Figure 1a). However, there were also differences in the patterns between periods because peak eating times differed (data not shown). Straw eating patterns were also affected by a treatment main effect (linear P < 0.0001; quadratic P = 0.001) and their interaction with the time of day (linear P = 0.003, and quadratic P = 0.07 treatment x time). Regardless of the time of day, the average time spent eating straw was 2.86, 3.93, and 4.01 ± 0.16 min/ h for T2, T4, and T8, respectively. Figure 1b shows that the time spent eating straw increased in all daylight intervals as the number of heifers per concentrate feeder increased (P < 0.05).
|
The number of displacements among pen mates at the concentrate feeders increased linearly as the number of heifers per feeder increased (Table 3; P < 0.001). In contrast, the number of displacements from the straw feeders decreased linearly as the number of heifers per concentrate feeder increased (P = 0.01), but it also showed a tendency for a quadratic effect (P = 0.09). In addition, the linear (P = 0.01) and quadratic (P = 0.004) effect of a treatment x period interaction indicated a linear increase in period 1 (P = 0.01) and quadratic responses in period 3 (P = 0.03) and 6 (P = 0.08), when T2 heifers showed the greatest number of displacements from the straw feeders (data not shown). Displacements from the water bowls followed a quadratic pattern (P = 0.005) as the number of heifers per feeder increased, being greatest in T4 (Table 3). The total number of displacements per pen and day increased linearly as the social pressure in the concentrate feeders increased (P < 0.05). Regardless of treatment, the average number of displacements from straw feeders, the displacements from water bowls, and the total sum increased across periods (P < 0.05; data not shown). However, the number of displacements from the concentrate feeders showed a nonsignificant numerical decrease from 36.5 at period 1 to 30.1 at period 6 (P = 0.33; data not shown).
|
). As expected, there were differences among all dominance categories, with mean proportions of displacements won of 68, 55, 45, and 32%, for very dominant, dominant, subordinate, and very subordinate heifers, respectively (data not shown). The quadratic treatment x ADV category (P < 0.001) showed that each category responded differently to treatments by changes in the average proportion of displacements won by the heifers (Table 3). Therefore, increasing the number of heifers per concentrate feeder resulted in a quadratic response (P < 0.001) of the ADV of very dominant and very subordinate heifers, with T4 showing the lowest ADV in very subordinate heifers but the greatest ADV in very dominant heifers. Fecal and Blood Measurements
The average fecal GM concentration of the pen was not affected (P 0.16), whereas the individual maximum concentration of the pen responded quadratically (P < 0.001) as the number of heifers per concentrate feeder increased, with T4 pens showing the lowest values (Table 4). The linear treatment x dominance category interaction (P = 0.12) did not indicate significant linear regression coefficients, but the quadratic effect of treatment x dominance class interaction (P = 0.01) indicated that quadratic regression coefficients were different among dominance categories. Therefore, fecal GM showed a tendency for a quadratic response in dominant animals (P = 0.06), with T8 showing the greatest values. Serum haptoglobin concentration increased linearly (P = 0.05) as the number of heifers per concentrate feeder increased (Table 4). Average haptoglobin level was 219 ± 8 mg/L in period 1 and 229 ± 8 mg/L in period 2. It then increased at period 3 to 270 ± 8 mg/L (P < 0.05) and decreased sharply from period 3 to 4 (98 ± 8 mg/L) to remain at a low level thereafter, 89 and 106 ± 8 mg/L in period 5 and 6, respectively (data not shown). The linear treatment x dominance category interaction (P = 0.13) in blood haptoglobin concentration indicated that there was a weak tendency for linear regression coefficients to differ among dominance categories. Therefore, increasing the number of heifers per concentrate feeder resulted in a linear increase of haptoglobin in very subordinate heifers (P = 0.02) and tended to increase linearly in very dominant and dominant heifers (P < 0.10; Table 4). Moreover, very subordinate heifers under T8 had greater (P < 0.05) haptoglobin levels than the remainder of their pen mates, which did not happen within T2 and T4. Serum concentration of NEFA was not affected (P > 0.10), but β-HBT increased linearly (P < 0.001) as social pressure increased.
|
Treatments did not affect the average ruminal pH at 8 h after feeding (linear P = 0.25; Table 5), but a treatment x period interaction was observed (linear P < 0.001; quadratic P = 0.08), indicating that the regression coefficients of the number of heifers per feeder against ruminal pH were different among sampling periods. Indeed, ruminal pH showed a quadratic response in period 1 (P = 0.04; Figure 2), with T4 and T8 showing low values. In period 2, ruminal pH decreased linearly as the number of heifers per feeder increased (P = 0.03). Regardless of treatment, ruminal pH increased gradually from period 1 (5.37 ± 0.04) to 5 (5.92 ± 0.06), although period 6 (5.66 ± 0.10) was not different from the other periods (P > 0.05) because a numerical decrease was observed from period 5 to 6. The proportion of heifers with ruminal pH below 5.6 tended to increase linearly (P = 0.08) as the number of heifers per concentrate feeder increased. The average acetate, propionate, and valerate molar proportions, as well as total VFA concentration, were not affected by treatments (P > 0.10; Table 5). However, butyrate molar proportion increased linearly (P = 0.05), whereas the branched-chain VFA decreased linearly (P < 0.01) as the number of heifers per concentrate feeder increased. The linear treatment x period interaction (P < 0.01) in total VFA showed a positive linear regression coefficient of the number of heifers per concentrate feeder in period 2 (P < 0.01), with values being 167, 180, and 210 ± 10 mM for T2, T4, and T8, respectively (data not shown). Increasing the number of heifers per concentrate feeder resulted in a linear increase in ruminal lactic acid concentration (P < 0.01). In addition, the linear treatment x period interaction (P = 0.07) in ruminal lactic acid concentration indicated that regression coefficients were different among periods. Hence, linear increases were observed in periods 1, 2, and 3 (P 0.01).
|
|
|
DISCUSSION |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
). Regrouping was not carried out after the adaptation period because of potential effects of social disorganization (Hasegawa et al., 1997) and to study whether animals were able to revert those effects in the remaining fattening period. Performance and Intake
The initial BW of animals in the T8 pens was 4.3% lower than the average of the other 2 treatments. However, this difference was 3.4% at the time heifers were sent to the slaughterhouse. Therefore, negative effects during the arrival period were only partially recovered. The pen average and the within-pen variability of ADG were not affected by treatments in the present study. Nevertheless, variability of BW among heifers sharing the same pen tended to increase linearly as the number of heifers per concentrate feeder increased, confirming a lack of compensating mechanisms in the effects that occurred during the arrival period. Indeed, a numerical linear decrease was noted in HCW. Concentrate DMI decreased linearly as the number of heifers per feeder increased. Longenbach et al. (1999) increased the feed bunk length from 0.15 to 0.31 or 0.47 m/heifer in 3 groups of replacement heifers with different ages and observed a trend for increased within-pen SD in ADG and BW with diminishing feeding space in the older 2 groups, which was particularly greater at the end of each trial. In addition, Longenbach et al. (1999) observed a decreasing number of heifers able to eat simultaneously at a given linear feed bunk allowance as the age of heifers increased. Thus, social pressure was not constant because body size increases with age. The present study assessed social pressure in the concentrate feeders by controlling the number of heifers per pen able to eat simultaneously, and this was maintained throughout the study. The lack of long-term effects of increasing social pressure at the feeders on performance in the present experiment agrees with observations by Gonyou and Stricklin (1981), Zinn (1989), and Longenbach et al. (1999) in growing cattle. However, intake was not affected in those studies. Differences may be due to feeding management, group size, and feeder design. Concentrate and straw were fed in separate feeders in the present study and concentrate feeding places were separated by barriers. Despite the fact that increased social pressure at the concentrate feeder may affect performance during stressful situations, such as after arrival (González et al., 2008), and that concentrate DMI remained lower afterward, no long-term effects were observed. Finally, the percentage of abscessed livers followed a quadratic pattern as the number of heifers per feeder increased, with T8 showing the greatest proportion. Low ruminal pH has been regarded as one of the main predisposing factors in the etiology of liver abscesses because it produces ruminal tissue damage, allowing necrotic pathogens to enter the bloodstream (Nagaraja and Chengappa, 1998). These authors indicated that a change in feeding pattern, allowing cattle to become hungry, may have been a predisposing factor, and sufficient bunk space was recommended to reduce incidence of liver abscesses.
Maintenance Behavior
Social competition for concentrate feeding increased as the number of heifers per feeder increased. Daily time spent eating concentrate decreased linearly by 28% as the number of heifers per feeding place increased from 2 to 8. However, animals partly compensated for this shorter concentrate eating time through a linear increase in the eating rate, although this increase was of a lesser magnitude (18%). This might explain the observed decrease of concentrate DMI. Olofsson (1999) observed that increasing competition from 1 to 4 cows per feeder resulted in a tendency to increase DMI in a 50% concentrate diet, whereas eating time decreased but eating rate and number of daily meals increased significantly. The quadratic effect on the within-pen SD of concentrate eating time suggests either that feeding behavior was more homogeneous among heifers sharing T2 pens or, alternatively, that feeding behavior was altered to a greater extent within T4 and T8 groups. These 3 variables show that both group and individual concentrate feeding behavior were altered. Harb et al. (1985) showed similar effects when increasing from 1 to 2 cows per feeder, with a positive correlation between social rank and eating time. Submissive cows in this study were probably less comfortable during feeding because their silage eating rate increased 2- to 3-fold. However, greater eating rate is also an adaptation mechanism to social constraints (Nielsen, 1999). Negative correlations between social rank and intake (estimated with markers) were observed by Harb et al. (1985) and Leaver and Yarrow (1980) when competition at the feeder increased.
The daily time spent by the heifers eating barley straw was lowest in the T2 groups. This result might have been a consequence of the lower linear straw feeding space in this treatment. The concentrate feeders in the present study were designed to be adapted to the existing experimental facilities. This shorter straw feeding space in T2 pens was a consequence of placing 2 concentrate feeder units in the 3.84-m feeding front, whereas only 1 steel unit was necessary in the T4 and T8 pens. Hasegawa et al. (1997) reported variable responses in the time spent eating several feeds after mixing cows, with responses subject to the cows’ social status, but the causes were not clear. In conclusion, the present study showed that increasing competition in the concentrate feeders led to reduced eating time and to increased eating rate and number of displacements. However, increased social pressure at the straw feeders led to decreased straw eating time and increased aggressive behavior.
The daily concentrate eating patterns showed that the decrease in the time spent eating concentrate with increasing social pressure was particularly evident at peak eating times, such as after feed delivery and sunset. However, concentrate eating time at sunrise and just after feed delivery was greatest in T2, but it was numerically greatest in T4 during the afternoon and at the sunset feeding period. Meanwhile, T8 showed the lowest concentrate eating time during both the sunrise and sunset feeding periods. Therefore, heifers under the greatest social pressure did not show a shift in the concentrate eating pattern toward times of the day when competition was lower. In contrast, Gonyou and Stricklin (1981) and Olofsson (1999) reported increased eating activity during nighttime with diets containing greater forage proportions than the proportion selected by the heifers in the present study. The availability of straw in different feeders may be the cause of this difference compared with TMR-fed cattle. This view may be supported by the significant treatment effect observed in the straw eating patterns. Heifers under the greatest social pressure at the straw feeders (T2), but under the lowest at the concentrate feeders, spent the shortest time eating straw during, and between, the 2 major periods of eating. Therefore, reduced feeding space either in the concentrate or straw feeders may have resulted in a restriction in the heifers’ natural feeding behavior, which usually leads to synchronization of feeding at peak eating times (Miller and Wood-Gush, 1991). In addition, heifers showed a reluctance to change both feeding patterns during less busy times of the day.
Lying time decreased, whereas standing time increased with an increasing number of heifers per feeder. This may suggest that subordinate animals did not greatly change their daily feeding patterns but spent more time standing without eating, perhaps waiting for less competition in the feeders. This was reported previously when 15 beef cattle per feeder were used (Gonyou and Stricklin, 1981) or when the number of cows per feeder increased from 1 to 4 (Olofsson, 1999). Friend (1991) concluded that deviations of behaviors such as feeding, resting or rumination patterns, and feed consumption rates are very sensitive indicators of the animal’s internal state. Therefore, eating patterns, lying time, and standing time in the present study indicate that the welfare of the animals may decrease as feeding space is reduced. However, results depend on other factors such as feeding management and facility design. For instance, Huzzey et al. (2006), using post-and-rail compared with headlocks, observed that daily feeding time was greater but that aggressions increased to a greater extent as competition increased.
Social Behavior
As expected, a clear positive linear relationship was observed between the number of heifers per feeder and the number of displacements among pen mates at the concentrate feeders. The stability over time observed in the number of displacements from the concentrate feeders may indicate that a social order for access to this resource was established and no further aggression was required. However, stabilization of the social hierarchy required maintenance of high levels of aggression. The number of aggressions during the present study was greater than the numbers observed in the first and third week after arrival at the feedlot (González et al., 2008). This is contrary to observations made by Kondo and Hurnik (1990), who reported that the number of aggressions decreased after regrouping, remaining at low levels after a social stabilization was achieved at approximately 1 wk. Nevertheless, the number of displacements from straw feeders increased over time and treatment differences became larger as heifer age increased. This was particularly noticeable in T2 pens because of the shorter straw feeding space, which led to greater aggression. The number of displacements from the water bowls responded quadratically to an increasing number of heifers per concentrate feeder and showed the greatest aggression among T4 heifers. This treatment had average social pressure at both straw and concentrate feeders. There are no obvious reasons for this result because the number and size of water bowls were identical in all pens. Although cattle spend a small proportion of the daily time on drinking activity, Andersson et al. (1984) showed that dominance is also exerted in the water bowls and completely modifies drinking behavior in a way similar to feeding. The number of displacements from concentrate feeders was greatest with the maximum number of heifers per feeder (T8), and the number of displacements from straw feeders was greatest with the shortest linear space per animal (T2). Therefore, we hypothesize that the T4 groups redirected part of the aggressive interactions, which were needed to establish or maintain priority of access to resources, toward water bowls. Finally, the sum of the number of displacements from all 3 resource containers also followed a linear trend with an increasing number of heifers per concentrate feeder, because of the greater relative importance of displacements from concentrate feeders. Nevertheless, the total number of displacements showed a treatment x period interaction because the T2 and T8 groups showed the same number of aggressions at period 6. Moreover, aggressions in T2 were greater at the straw feeders than at the concentrate feeders, but the opposite was true for the T8 groups.
No treatment effects were observed in the pen average ADV. Nevertheless, the treatment x ADV category indicated that heifers responded to treatments by modifying the aggressive behavior of the group’s social categories. Therefore, T4 were subjected to the lowest social pressure at both feeders, and this led to very subordinate heifers (weakest social category) either not needing to be aggressive or suffering the least aggression. On the other hand, very dominant T4 heifers were able to win the greatest proportion of displacements, either by being more aggressive or by suffering less aggression (fewer bidirectional displacements). Altogether, increasing social pressure at both concentrate and straw feeders led to differences in social organization within the group and increased aggressive behaviors.
Fecal and Blood Measurements
Despite the fact that average fecal GM concentration of the pen did not differ (P = 0.16) among treatments, a quadratic response (P < 0.001) was observed in the pen’s maximum fecal GM concentration as the number of heifers per concentrate feeder increased. This indicates that the most stressed heifers within T8 groups may have showed greater adrenal activity or suffered greater stress. Glucocorticoid metabolite concentrations of the present study are much greater than those reported by Morrow et al. (2002), who observed an increase from approximately 12 to 20 ng/g of DM of feces when dairy cows were moved to a novel environment or after a short transport. These most stressed heifers within a group may have suffered greater social pressure at the concentrate feeders in T8, as reflected in greater aggressions. In contrast, of the most stressed heifers in each treatment, the one in T4 showed the lowest stress because this treatment had the lowest concentrate and straw feeder social pressure combination. Glucocorticoids have been measured extensively as indicators of psychological and physical stressors. The results in GM of the present study reinforce the hypothesis that increasing competition at the feeders produces social stress, but only in some individuals within a group. We found no treatment effects in fecal GM concentration during the arrival period while heifers were adapting to the new conditions (González et al., 2008). Therefore, social stress may be the result of a long-term process, which may increase cumulatively over time in the most prone animals. Mench et al. (1990) reported increased blood cortisol in cows at 84 d after the introduction of new members to the pen. Chronic or long-term stress, as measured in the present study through fecal GM over a 6-mo period, is more likely to cause welfare and health problems than acute stress because of its effects on the central nervous system (Lane, 2006). Fecal GM have been considered as one of the most reliable noninvasive measurements of chronic stress in animals because they do not interfere with the stress response itself (Lane, 2006). Heifers within each pen were assigned to dominance categories by their ADV to assess which social category within a group suffered the greatest stress. Only dominant heifers showed a tendency for a quadratic response when increasing the number of heifers per concentrate feeder, with those under T8 showing the greatest fecal GM concentration. In addition, dominant heifers within T8 had greater fecal GM than very subordinate pen mates (P < 0.05) and had the greatest values of the pen. Conversely, dominant heifers within T4 had lower fecal GM concentration than very subordinate heifers (P < 0.05), and had the lowest of the pen. Within T2, there were no significant differences in fecal GM between dominance categories (P > 0.10). Therefore, dominant heifers seem to be the most benefited social class under the lowest competition but the most stressed heifers under the greatest competition. This may suggest that, with high social pressure, dominant animals either try to ascend in the social hierarchy or they are stressed because they are pushed down in the social hierarchy. The fact that subordinate heifers did not show increased fecal GM may suggest that they gave up any possibility of ascending in the social hierarchy. Other studies with cows did not find relationships between dominance rank and blood cortisol, either when dairy cows were forced to a change in dominance rank under a competitive situation (Arave et al., 1977) or in herds composed of different breeds (Adeyemo and Heath, 1982). Contrary to our results, Mench et al. (1990) observed greater blood cortisol values in subordinate heifers compared with dominant heifers at 84 d after mixing beef cows. Nonetheless, those results may have reflected differences in the response to the handling stress during bleeding. Results of the present study suggest that social status influences how each animal copes with the social environment, with dominant heifers having more difficulties as competition increases. High glucocorticoid concentrations have been associated with both subordinance and dominance in different species (Lane, 2006). In primate societies, the diversity in blood and urinary cortisol in relation to social status can be explained by different social environments experienced by individuals. Thus, subordinate animals show greater cortisol concentrations than dominant animals in overtly aggressive societies because they carry the greatest rates of physical or psychological stressors, have severe resource limitations, have less social support, and are minimally related to other members (Abbott et al., 2003). Conversely, these authors found that the dominant counterparts showed greater cortisol concentration than subordinates in nonaggressive societies, which might be the case in cattle and may even explain breed differences in response to stress. Finally, the treatment effect on the most stressed heifers within a group and on dominant heifers highlights the importance of analyzing not only the group mean values to assess stress and welfare, but also individual animals within a group.
Blood NEFA levels were not affected by treatments, but β-HBT concentration increased linearly as the number of heifers per feeder increased. Both NEFA and β-HBT were used as indicators of body fat metabolism because they ensure appropriate and coordinated fuel under high energy demands or muscular activity, or when other energy sources are scarce (Adewuyi et al., 2005). Therefore, a greater extent of lipolysis was expected with an increasing number of heifers per feeder because of the increased energy demands needed for competition and displacements (physical activity), or because of undernutrition caused by decreased DMI. In positive energy balance, ketone bodies arise, mainly from metabolism of butyrate by the rumen wall; therefore, the increase in β-HBT is not proportional to that of NEFA (Sato et al., 1999). The increase in ruminal butyrate molar proportion and concentration in the present study supports this hypothesis.
Ruminal Fermentation
Subacute ruminal acidosis is characterized by a ruminal pH below 5.6 (Britton and Stock, 1989; Owens et al., 1998) for at least 3 h/d (Gozho et al., 2006). However, acidosis encompasses an array of biochemical and physiological stresses caused by rapid production and absorption of organic acids and endotoxins, but not necessarily by low ruminal fluid pH (Britton and Stock, 1989). Therefore, the mere presence of low ruminal pH may be meaningless if it does not affect other aspects of the animal such as performance or immune response. However, when rumen pH is considered together with other indicators, then SARA can be diagnosed (Nordlund et al., 2004). Nonetheless, SARA was assessed together with other ruminal and blood measurements in the present study. The average ruminal pH throughout the 6 experimental periods was not consistently affected, but lactic acid concentration increased linearly with increasing competition. However, T4 and T8 resulted in lower ruminal pH at periods 1 and 2. Heifers under the lowest social pressure (T2) were able to maintain a more stable ruminal pH at 8 h after feeding across periods, compared with T4 and T8. Effects observed in ruminal pH during periods 1 and 2 corresponded with greater total VFA and lactic acid concentrations. These results were due, in part, to different numbers of heifers showing greater than normal concentrations. Ruminal lactate does not accumulate at concentrations greater than 5 mM under normal conditions, and a concentration greater than 40 mM is indicative of severe acidosis (Owens et al., 1998; Nagaraja and Titgemeyer, 2007). Six heifers under T8, 5 under T4, and none under T2 had ruminal lactate concentrations greater than 5 mM during periods 1 and 2 (data not shown). Therefore, a greater proportion of heifers may be at risk for, or show more severe, ruminal acidosis during these periods. In addition, the proportion of heifers with ruminal pH below 5.6 tended to increase linearly by 20 percentage units as the number of heifers per feeder increased from 2 to 8. Dairy herds have been suggested to be at high risk of acidosis if more than 25% of animals show ruminal pH below 5.5 (Nordlund et al., 2004) or when the average pH (rumenocentesis) across cows is below 5.6 (Stone, 2004). In the present study, this incidence rate of ruminal pH below 5.5 was 39, 54, and 62 ± 10%, for T2, T4, and T8, respectively. However, the proportion of heifers with ruminal pH below 5.6 decreased from an average of 85.5% in period 1 to 24.3% in period 5, regardless of treatment (P < 0.05; data not shown). The greater ruminal acidosis incidence, as well as the time spent standing and the reduction of resting time with increased competition observed in the present study may have negative effects on hoof health (Cook et al., 2004; Stone, 2004), but this issue was not assessed in the present study. The linear increase in ruminal butyrate proportion and concentration (data not shown) with an increasing number of heifers per concentrate feeder may be a result of increased lactic acid production and degradation to butyrate (Nagaraja and Titgemeyer, 2007). Branched-chain VFA come from dietary true protein degradation and microbial protein recycling, processes reduced as ruminal pH decreases (Miura et al., 1980).
Sudden drops or wide fluctuations of ruminal pH, and also a greater time under suboptimal pH increases ruminal bacteria cell lyses and, consequently, the release of endotoxin lipopolysaccharides (Nagaraja and Titgemeyer, 2007). Therefore, the acute phase response is activated and, consequently, blood haptoglobin increases (Gozho et al., 2006). In the present study, serum haptoglobin concentration increased, on average from 157 to 178 mg/L, whereas ruminal pH decreased numerically from 5.66 to 5.54 as the number of heifers per feeder increased from 4 to 8, respectively. When the haptoglobin concentration ([Hapto]) of the present study was regressed against ruminal pH, the resulting equation yielded an R2 = 0.32 (P < 0.0001; pH = 5.96 (±0.08) – 2.15 (±0.44) [Hapto]). In addition, Skinner et al. (1991) suggested that blood haptoglobin concentrations greater than 200 mg/L are indicative of early or mild infection, a threshold exceeded from periods 1 to 3 of the present study, coincident with the lowest ruminal pH. Therefore, despite the fact that one spot-in-time ruminal pH sample is not an accurate indicator of increased risk of SARA, ruminal lactate, total VFA concentration, branched-chain VFA, and blood haptoglobin support the hypothesis that increasing the number of heifers per concentrate feeder may be another risk factor for ruminal acidosis (Gozho et al., 2006; Nagaraja and Titgemeyer, 2007). However, the risk seems to be high during the first 3 to 4 mo of the growing period. Finally, the linear decrease in concentrate DMI with increasing number of heifers per feeder in the present study may also be due to the effects of acidosis (Britton and Stock, 1989), but other factors such as social behavior may also be responsible.
Subordinate cattle may be at greater risk of ruminal acidosis because they have to wait to access the occupied feeders, causing anxiety; therefore, meal size and eating rate might increase once the feed is reached. On the other hand, dominant animals may be more prone to overeating because they eat first and are usually not interrupted while feeding and may therefore have larger meals. Disruption of the normal feeding patterns has been seen as increased risk of ruminal acidosis in individual animals within a pen because the ability to self-regulate ruminal function may be impaired (Schwartzkopf-Genswein et al., 2003). Feeding behavior in the present study showed that feeding characteristics were altered by increased competition. Larger meals and greater eating rates of highly fermentable diets may result in lower ruminal pH, but individual feed intake behavior was not measured in the present study. The fact that blood haptoglobin increased at a greater extent within very subordinate heifers indicates that they are likely to suffer more severe ruminal acidosis. In addition, ruminal pH was 5.45, 5.60, 5.53, and 5.45 for very subordinate, subordinate, dominant, and very dominant heifers, respectively (P = 0.09), but was not affected by treatments (data not shown). These results suggest that both very subordinate and very dominant animals are at a greater risk of ruminal acidosis.
In conclusion, an increasing number of heifers up to 8 per concentrate feeder reduces concentrate intake but does not affect performance after an adaptation period to the limited feeding space because compensating mechanisms may arise. However, an improvement in the group homogeneity of BW may be achieved with more feeding space. Heifers adapt to the increased competition by changing their behavior, although these forced changes may be detrimental for animal welfare. Fecal GM concentration indicates that sufficient feeding space should be provided at both concentrate and straw feeders. Otherwise, subordinate heifers might suffer social stress. Increased competition may result in lower ruminal pH during the initial 3 or 4 mo of the fattening period, which activates the acute phase response and increases liver abscesses. Four heifers per concentrate feeding place seems a reasonable upper limit considering several welfare and productive indicators, including within-group variability of final BW, liver abscesses, and fecal GM concentration of some individual animals within the group. However, advantages of having 4 heifers per concentrate feeding place were not apparent when considering concentrate intake, time spent eating concentrate, lying and standing time, aggressive behavior, serum haptoglobin, and ruminal acidosis incidence. Therefore, it would be interesting to further assess the possible benefits of giving more feeding space to heifers fed high-concentrate diets.
|
Footnotes |
|---|
2 Ministerio de Educación y Ciencia, Madrid, Spain (Studentship FPU AP20023344) is acknowledged. Present address: Agriculture and Agri-Food Canada, Lethbridge Research Centre, Lethbridge, Alberta T1J 4B1, Canada.
3 Corresponding author: Alfred.Ferret{at}uab.es
Received for publication October 23, 2007. Accepted for publication January 31, 2008.
|
LITERATURE CITED |
|---|
|
TopAbstractINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONLITERATURE CITED |
|---|
This article has been cited by other articles:
|
|
 |
|
  T. J. DeVries and M. A. G. von Keyserlingk Competition for feed affects the feeding behavior of growing dairy heifers J Dairy Sci, August 1, 2009; 92(8): 3922 - 3929. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 L. A. Gonzalez, L. B. Correa, A. Ferret, X. Manteca, J. L. Ruiz-de-la-Torre, and S. Calsamiglia Intake, water consumption, ruminal fermentation, and stress response of beef heifers fed after different lengths of delays in the daily feed delivery time J Anim Sci, August 1, 2009; 87(8): 2709 - 2718. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
T. J. DeVries and M. A. G. von Keyserlingk Short communication: Feeding method affects the feeding behavior of growing dairy heifers J Dairy Sci, March 1, 2009; 92(3): 1161 - 1168. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
