Peer reviewed Original research - AASV

Journal of Swine Health and Production — Volume 12, Number 3 111

LFS, ADB, JFP, HWG: Prairie Swine Centre Inc, Saskatoon, Saskatchewan, Canada

RDB: Pig Improvement Co, Franklin, Kentucky

Corresponding author: Dr A. Denise Beaulieu, Prairie Swine Centre, Inc, PO Box 21057, 2105 8th

St E, Saskatoon, Saskatchewan, Canada S7H 5N9; Tel: 306-667-7441; Fax: 306-955-2510; E-mail:[email protected].

This article is available online at http://www.aasv.org/shap.html.

Smith LF, Beaulieu AD, Patience JF, et al. The impact of feeder adjustment and group size-floorspace allowance on the performance of nursery pigs. J Swine Health Prod. 2004;12(3):111-118.

Original researchPeer reviewed

The impact of feeder adjustment and group size-floor spaceallowance on the performance of nursery pigsLaura F. Smith, BSA; A. Denise Beaulieu, PhD; John F. Patience, PhD; Harold W. Gonyou, PhD; R. Dean Boyd, PhD

SummaryObjectives: To determine the effects ofgroup size and feeder trough opening onweanling pig performance, and to developfeeder adjustment recommendations forthe pork industry.

Materials and methods: Weanling pigletswere randomly assigned to three groupsizes (16, 20, and 24 pigs), with corre-sponding floor space allowances of 0.35m2, 0.28 m2, and 0.23 m2 per pig.Crumbled feed was provided in trough-typedry feeders using five different feeder gapopenings (range 9.2 - 31.5 mm) for 42

days, beginning 8 days postweaning. Threereplicates were performed (N = 716).

Results: During Days 21 to 42, bodyweight, ADG, and average daily feed intakeincreased with feeder gap size and floorspace allowance (P < .05). By day 42, pigsin pens with the widest feeder gap weighed10% more than pigs in pens with thesmallest gap (P < .001). Pigs with the great-est space allowance (0.35 m2) were 5.6%heavier than pigs with the least amount ofspace (0.23 m2; P < .05).

Implications: Under the conditions of this

study, increasing space allowance from0.23 m2 per pig to 0.35 m2 per pig wasassociated with increased body weight atapproximately 10 weeks of age. Optimalgrowth and feed efficiency were observedwhen the feeder gap opening allowed 25 to60% of the feeder trough to be consistentlyclear of feed. The capacity of a nurseryfeeder space is 11 pigs when the feeder gapopening allows 25 to 60% of the feedertrough to be clear of feed.

Received: October 2, 2002Accepted: March 25, 2003

Nursery management is a particu-larly critical area of pig produc-tion. Achieving a smooth transi-

tion from farrowing through weaning isessential. Moreover, growth during thenursery period affects final weights at market.In one report, pigs that were 1 kg heavierat 11 weeks of age were 1.5 to 1.8 kgheavier at 17 weeks of age.1

Two important aspects of nursery manage-ment are feeder adjustments and groupsize. Adjustment of the feeder trough opening(feeder gap) may influence both feed con-sumption and feed wastage. Moreover, ifaccessing feed becomes difficult, individualpigs spend more time at the feeder, and thenumber of pigs able to obtain sufficientfeed diminishes.2 Group size is directlyrelated to floor space allocation, which isparticularly important during this periodof rapid growth. Crowding has negativeimpacts on feed intake and growth,3–6 andmay exacerbate social vices such as tail biting,side nudging, and ear chewing.7

Few reports exist concerning floor spaceallowance and feeder gap opening, especiallyfor the current high-lean genetics andnewer housing systems. This study focusedon the interaction between feeder gap ad-justment and floor space allowance.

Materials and methodsAnimals and dietsPiglets (C15 female × line 65 boar; PigImprovement Canada, Acme, Alberta,Canada) weaned at an average age of 18.2days and a mean weight of 6.4 kg were se-lected for the experiment on the basis ofgender and weight uniformity. The pigletswere obtained from groups weaned in 3consecutive weeks at Prairie Swine Center(PSC) Elstow Research Farm in Saskatchewan,Canada. A commercial feeding programbased on wheat and soy meal (formulatedby PSC Elstow, manufactured by FederatedCo-operatives Ltd, Saskatoon, Saskatchewan,

Canada) was employed. A phase 1 diet wasfed for the initial 4 days postweaning, fol-lowed by a phase 2 diet for 3 weeks, and aphase 3 diet for the remainder of the 42-day experiment. All three diets were formu-lated to appropriate energy and amino acidspecifications for pigs of this age, and thephase 1 and phase 2 diets contained moreanimal protein. The feed was provided as acrumble except during the fourth week forthe first group of pigs, when it was mistak-enly manufactured and fed as a short-cutpellet.

HousingPigs were housed in three identical nurseryrooms at PSC Elstow Research Farm, a600-sow farrow-to-finish farm in centralSaskatchewan. Each nursery room contained16 pens, 3.05 m × 1.83 m (5.58 m2), withtotally slatted plastic flooring and solidPVC pen dividers and front gates. The 12noncorner pens of each room were used forthe experiment. Each pen was equippedwith one multi-space, trough-type dryfeeder (Generation 3; Staco, Shafferstown,Pennsylvania), which was divided into sixeating spaces by nose barriers 15 cm apart.Wooden blocks were inserted into thefeeder to provide one functional feederspace per four pigs regardless of the numberof pigs in the pen. Feeders were checked

!Beaulieu 3/31/04, 4:16 PM111

112 Journal of Swine Health and Production — May and June, 2004

two or three times daily to ensure that theywere functioning properly and that feedwas available. Water was provided by asingle nipple-type drinker fastened to thewall at the rear of the pen. A 12-hour light,12-hour dark lighting regime was used,with lights on at 7:00 AM.

TreatmentsGroup sizes were chosen to provide specificfloor space allowances. A group size of 16pigs per pen provided 0.35 m2 per pig,which approximates the floor space allow-ance specified in the Canadian Code ofPractice.8 A group size of 24 provided only0.23 m2 per pig, which is considered verycrowded. An intermediate group size of 20pigs per pen allowed 0.28 m2 per pig. Sinceboth group size and floor space allowancechanged with each treatment, they are con-founded, and we refer to these treatmentsas group size-density.

Feeder gap openings were selected duringa preliminary trial to determine the manu-facturer’s setting required to achieve thefour desired gap openings. Observationswere made of pigs of different ages housedfor 1 week in three rooms similar to thoseused in the study, and fed the same crumbleddiet. A lever device is provided on the sidewall of the Generation 3 feeders used inthis experiment, to adjust the openingwhich allows the feed to flow into thetrough. Level one was defined as the ad-justment that allowed only a small bead offeed to be present in the feeder tray next tothe delivery slot. Level four was defined asthe adjustment that allowed the feeder trayto be constantly full, and levels two andthree were intermediate. At the completionof the preliminary period, it was concludedthat these four levels were achieved withgap openings of 9.2, 11.8, 17.9, and 24.8mm. However, during the first replicate ofthe experiment, it was found that the gapopening of 9.2 mm restricted access to feedto a degree that compromised animal wel-fare. In subsequent replicates, the smallestgap opening used was 11.8 mm and thelargest was 31.5 mm. The openings usedcorresponded to the manufacturer’s settingsof 3 (9.2 mm), 4 (11.8 mm), 6 (17.9 mm),8 (24.8 mm), and 10 (31.5 mm).

Experimental designThe experiment was initially designed withthree replicates, comparing four feeder ad-justment gap sizes within three group sizetreatments, in a 4 × 3 factorial arrangement.In each replicate, four pens of each group

size (16, 20, and 24 pigs) were housedwithin a single room. One pen of eachgroup size was then assigned to one of fourfeeder gap openings. As gap opening sizeswere modified after the first replicate, theentire experiment was conducted using a5 × 3 fractional factorial arrangement oftreatments. Numbers of pens and pigs pertreatment are described in Table 1.

Experimental protocolThree replicates were performed. Pigletswere randomly assigned to the group size-density treatments at weaning. Each pencontained an equal number of females andcastrated males, and litter mates were allo-cated across treatments. Eight dayspostweaning (Day 0), each pen was assignedto a feeder-adjustment treatment.

Pigs were weighed individually on the dayof weaning (Day -8) and on Days 0, 21,and 42. The entire pen was weighed weeklyas a group, and feeders were also weighedat this time to allow calculation of feed dis-

appearance. Photographs of feeders weretaken every Tuesday at 10:00 AM as an il-lustration of the trough area clear of feed.The area clear of feed was measured using ameasuring tape; this area was converted toa percent based on the total feeding-surfacearea of the trough. If the area clear of feedwas circular or triangular, appropriateequations were used to estimate the cleararea. Feed depth at three equally spacedlocations within the feeder trough was de-termined when area measurements weremade.

During Days 3 to 6 and 39 to 42, thefeeder area of each pen was videotaped forone 24-hour period, using a time-lapse re-corder, to determine the time spent eating.Four pigs in each pen were used as focalanimals to estimate the time spent eating.No attempt was made to estimate otherbehaviors, such as dominant or subordinatebehavior at the feeder. Four pens were re-corded onto a single tape by means of a‘quad’ (screen splitting device).

Table 1: The impact of feeder gap and group size on mean body weight innursery pigs from approximately 26 to 68 days of age1

)mm(pagredeeF m(ezispuorG 2 )gip/

2.9 8.11 9.71 8.42 5.13 42)32.0(

02)82.0(

61)53.0(

snepfo.oN 3 9 9 9 6 21 21 21

sgipfo.oN0yaD

06 081 971 771 021 482 932 391

sgipfo.oN24yaD

06 871 571 271 611 972 432 881

thgiewydobnaeM 2 )gk(

gninaeW 43.6 33.6 23.6 24.6 04.6 93.6 63.6 33.6

0yaD 90.7 01.7 21.7 21.7 69.6 50.7 21.7 70.7

24yaD 3 52.72 79.82 55.92 05.92 19.92 30.82 93.92 96.92

1 A 42-day study was conducted (three replicates) using a total of 716 pigs weaned at18 days of age and housed in groups of 16, 20, or 24 per pen in identical pens (5.58m2) in three nursery rooms. One of the four pens of each group size in each nurseryroom was assigned on Day 0 (8 days postweaning) to one of four feeder gap sizes(three pens per feeder gap size). Feeder gap was defined as the size of the openingthat allowed feed to flow into the trough of a dry feeder. Feeder gaps used in the firstreplicate were 9.2, 11.8, 17.9, and 24.8 mm. As the 9.2-mm gap restricted access tofeed to a degree that compromised animal welfare, the feeder gaps used in thesecond and third replicates were 11.8, 17.9, 24.8, and 31.5 mm. Data were analyzedby analysis of variance with pen the experimental unit, using a 5 × 3 factorialarrangement of treatments. The model included the effects of replicate, group-sizedensity, feeder gap adjustment, and the interaction of group-size density and feedergap adjustment.

2 Mean body weights determined weekly by weighing the pen as a group.3 Body weight at Days 7, 14, 21, 28, and 35 was not affected by feeder gap opening or

group size (P > .05). Day 42 body weight was influenced both by feeder opening (P <.001) and group size (P < .01). The treatment effect of the feeder opening was linear(P < .02).

!Beaulieu 3/31/04, 4:16 PM112


On Day 42, each pig was individuallyobserved and scored for evidence of tail bit-ing, side nudging, ear chewing, and generalsigns of aggression. All procedures em-ployed in this experiment were approvedby the Protocol Review Committee of theCommittee on Animal Care and Supply atthe University of Saskatchewan.

CalculationsIndividual pig weights on Days -8, 21, and42 were used to determine the coefficientof variation (CV) within each pen. Averagepig weight calculated from weekly penweights was used to calculate ADG. Aver-age daily feed intake (ADFI), estimated byfeed disappearance, was calculated fromweekly feeder weights. Previous experienceat the same facility showed that feed wast-age with these feeders was minimal.

Instantaneous scan samples of the numberof pigs eating at the feeder, sampled at 2-minute intervals, were used to determinetime spent eating for the entire day, byconverting the mean percentage of pigsaccessing the feeder per 2-minute intervalto the number of minutes per day or houreach pig spent eating. For example, if anaverage of 8% of pigs accessed the feederon scan samples, then the feeder was usedfor 115 minutes per day (0.08 × 1440minutes per day). A pig was considered tobe eating if its head was over the trough,unless it was sleeping with its head in thefeeder.

The body was divided into five regions forscoring of lesions: abdomen, legs, ears,back, and tail. Injuries were graded using ascale of 0 to 2: 0, no apparent injury; 1,evidence of a lesion, but no broken skin; 2,an open wound. The same scale was usedto score evidence of joint swelling andlameness: 0, no joint swelling or lameness;1, some joint swelling; 2, joint swellingwith lameness.

Statistical analysisPen was considered the experimental unitfor all criteria except the within-pen CV ofbody weight, which was calculated fromindividual pig weights. The experiment wasanalyzed as a randomized complete blockdesign, with a fractional factorial arrange-ment of treatments (because of differentfeeder adjustments used in the first replicatecompared to the second and third replicates).Overall ADG, ADFI, and feed efficiency(gain:feed) were also analyzed, excepting

Week 4 of the first replicate when the feedwas mistakenly pelleted.

Data were analyzed by analysis of varianceusing the GLM procedure of SAS (SAS,Version 8; SAS Institute, Cary, NorthCarolina). The model contained the effectsof replicate, group size-density, feeder ad-justment, interaction of feeder adjustmentand group size-density, and initial weight(Day 0), which was used as a covariate.Except for the two variables for which themodel also included the effect of age (CVof body weight and time spent eating), datawere analyzed within discreet time periods.Variables included body weight, ADG,ADFI, and feed efficiency, weekly and overthe entire experimental period. Statisticalanalyses of feeder-trough coverage and le-sion-score data were performed on penmeans. Orthogonal polynomials were usedto determine linear or quadratic trends offeeder adjustment. An effect was consideredsignificant if P < .05 and was considered atrend if P > .05 but < .10.

ResultsFeeder gap openingLarger feeder gap opening was not associatedwith increased ADG during the initial halfof the experiment (Days 0 to 21; P > .05;data not shown). However, an effect of di-minishing feeder gap opening on ADG wasobserved during the last week of the experi-ment (Days 36 to 42; P = .01), and overthe entire experimental period (Days 0 to42; P = .02). Average daily gain over theentire experimental period was 480 gramsper day when the feeder gap opening was9.2 mm and 532 grams per day when thefeeder gap opening was 31.5 mm. A largerfeeder gap opening was associated withgreater body weight at Day 42 (Table 1).The data were less well described by a qua-dratic curve (P = .08). An inflection pointof 25.1 mm was calculated.

Feeder gap opening impacted feed intakeonly during the fourth week (Days 22 to28; P < .05) and during the second half ofthe experiment (Days 22 to 42; P < .02).From Day 22 to 28, daily feed intake was0.93 kg per pig when the feeder gap openingwas 9.2 mm and 1.07 kg per pig when thefeeder gap opening was 31.5 mm. Duringthe fourth week of the first replicate of thestudy, feed was mistakenly pelleted. Theoverall effect of feeder gap on feed intakewas significant (P = .047) if the data from

Week 4 in the first replicate are included,but not when these data are excluded (P =.06).

Throughout the entire experiment, feedblockage was a problem when the feedergap adjustment was set for an opening of9.2 mm or 11.8 mm. Feeders with thesegap openings required blockages to be cor-rected two to 10 times per week, whilefeeders with a larger gap size remainedtrouble free, allowing for a constant supplyof feed. Feed intake increased numericallybut not significantly up to a gap size of ap-proximately 18 mm, but not with largergap sizes.

Increases in feed intake were generally ac-companied by an improvement in gain,and feed efficiency was not affected by thefeeder gap opening (P > .05; data notshown).

The CVs of body weight at weaning (Day - 8)and on Days 21 and 42 are presented inTable 2. There was an age effect on averageCV, which was > 20% at weaning, but haddecreased to 12.7% by Day 42 (P < .001).Increased variability of body weight at Day42 was associated with a smaller feeder gapopening (P = .04).

A smaller feeder gap opening was associatedwith a smaller area of the feeder troughcovered with feed (P < .001). Average feeddepth increased (P < .001) as feeder gapopening increased (Table 3). This effect isillustrated in Figure 1.

The amount of time the young pigs spenteating (Days 3 to 6) was longer when thefeeder gap was smaller and feed wasdifficult to access (P = .013; Table 4). Theeffect of the smallest feeder gap was espe-cially marked relative to the next increment(142 minutes per pig per day for the 9.2mm gap opening; 118 minutes per pig perday for the 11.8 mm opening). Feeder gapopening had no effect on the time spenteating when the pigs were older (Days 39to 42; P = .29).

Eating activity followed a marked circadianrhythm, with one peak at approximately8:00 AM, shortly after the lights came on,and another at 2:00 PM, before the lightswent off. The feeder gap opening did notaffect this pattern of eating.

Feeder gap adjustment was not associatedwith any change in the severity or numbersof skin lesions or with joint swelling scores(P > .05; data not shown).

!Beaulieu 3/31/04, 4:16 PM113


Group size-densityDuring the last 2 weeks of the experiment,ADG tended to be lower in pens withlarger group sizes, and the cumulative effectwas such that ADG from Day 0 to Day 42was lower in pens which contained thelarger groups. In pens with groups of 24pigs, ADG was 495 grams per day; in penswith groups of 20 pigs, ADG was 518 gramsper day; and in pens with groups of 16pigs, ADG was 531 grams per day (P < .01).Consequently, Day 42 body weight wasgreater in pens with a smaller group size(P < .01; Table 1). Increased pig densitywas associated with decreased feed intake(ie, feed disappearance, ADFI) during the

final half of the experiment (Days 22 to42). During this period, ADFI was 975grams with 24 pigs per group; 1029 gramswith 20 pigs per group, and 1050 gramswith 16 pigs per group (P = .01). This ef-fect was also apparent over the entire ex-perimental period (Days 0 to 42): overallADFI was 719 grams with 24 pigs pergroup; 771 grams with 20 pigs per group;and 774 grams with 16 pigs per group (P =.01). Increases in feed intake were generallyaccompanied by an improvement in gain,and feed efficiency was not affected by thenumber of pigs in a pen (P > .05; data notshown).

The CV of body weight was greater in penswith larger group sizes at Day 21, but notat Day 42 (Table 2).

The number of pigs per pen did not affecteither the average proportion of the feedertrough clear of feed or feed depth (Table 3).Increasing the number of pigs per pen hadno effect on the time spent eating in youngor older pigs (Table 4). The pattern of eat-ing was not affected by the group size (datanot shown).

The mean skin lesion score of 0.06 with 24pigs per pen decreased to 0.05 and 0.03

yaDdehgiew

VC 2 )%( VC 2 )%(P eulav)mm(pagredeeF ytisned-ezispuorG

m( 2 )gip/ 3

2.9 8.11 9.71 8.42 5.13 MES 4 P eulav 42)32.0(

02)82.0(

61)53.0(

MES 4 ytisneD 5 noitcaretnI 6

gninaeW 3.91 5.02 5.02 6.02 6.02 9.0 79. 3.12 0.02 7.91 7.0 23. 61.12yaD 8.71 8.61 4.51 6.51 2.61 8.0 14. 5.71 7.61 9.41 6.0 40. 01.24yaD 2.51 5.21 3.11 2.21 3.21 6.0 40. 3.21 4.21 5.31 5.0 12. 70.

elbairaV )mm(pagredeeF ytisned-ezispuorGm( 2 )gip/ 2

P eulav

2.9 8.11 9.71 8.42 5.13 MES 3 P eulav 42)32.0(

02)82.0(

61)53.0(


raelcaerA 6 )%( 4.49 7.78 3.65 9.52 4.7 39.3 100.< 5.25 0.35 5.75 40.3 215. 655.htpeddeeF 7 )mc( 30.0 30.0 42.0 86.0 62.1 40.0 100.< 34.0 94.0 34.0 30.0 903. 520.

Table 2: The impact of feeder gap, group size, and pig age on variability (CV) in nursery pig body weight1

1 A 42-day study was conducted (Table 1). Pigs were weighed individually at weaning (Day -8) and on Days 0, 21, and 42. Mean bodyweights were determined weekly by weighing the pen as a group. Data were analyzed by analysis of variance with pen theexperimental unit, using a 5 × 3 factorial arrangement of treatments. The model included the effects of replicate, group-size density,feeder gap adjustment, and the interaction of group size-density and feeder gap adjustment.

2 Coefficient of variation of body weight calculated using individual body weights.3 Floor space allowance per pig at the respective group sizes.4 Calculated using the average number of pens per feeder gap.5 P value for the group size-density treatments.6 P value for the interaction of feeder gap size and group size-density treatment.

Table 3: The impact of feeder gap and group size on the average percent area of the feeder trough clear of feed and onaverage feed depth for nursery pigs beginning 1 week after weaning at approximately 18 days of age1

1 A 42-day study was conducted (Table 1). Data were analyzed by analysis of variance with pen the experimental unit, using a 5 × 3factorial arrangement of treatments. The model included the effects of replicate, group size-density, feeder gap adjustment, and theinteraction of group size-density and feeder gap adjustment.

2 Floor space allowance per pig at the respective group sizes.3 Calculated using the average number of pens per feeder gap.4 P value for the group size-density treatments.5 P value for the interaction of feeder gap opening and group size-density treatment.6 The trough area clear of feed was measured weekly and converted to a percent based on the total feeding-surface area of the trough.

Data from Week 4 in the first replicate, when feed was mistakenly provided as pellets instead of crumbles, was excluded.7 Calculated as the average depth of feed at three locations in the tray after the lip area was wiped clear, excluding data for Week 4.

!Beaulieu 3/31/04, 4:16 PM114


with 20 and 16 pigs per pen, respectively(P =.02). However, this effect could not beattributed to any specific body region (P >.05 for individual body regions).

Feeder gap size by group size-density interactionsThe effect of group size on final bodyweight (Day 42) was accentuated at smallerfeeder gap openings (interaction of feedergap opening and group size-density, Figure2). The effect of group size on the totaltime spent eating was most evident at thesmallest feeder gap opening (Figure 3).

DiscussionThe effects of feeder gap adjustment andgroup size-density treatments were not al-ways apparent on a weekly basis, but hadprogressive effects, and became significantin the latter half of the experiment. Wewould expect the negative effects of eitherto have been exacerbated if the experimenthad continued. Floor space allowance inthis experiment decreased with increasinggroup size, while pen size was constant.Thus, group size and floor space allowancewere confounded. Group size, independentof floor space allowance, may have affectedthe results. As group size increases, the freeor unused space in a pen increases.9,10

Studies examining the impact that variousgroup sizes have on performance of weanlingpigs are limited. Generally, performanceeffects are seen only when group size variesto a large extent. After reviewing the litera-ture, Kornegay and Notter3 concluded thatthe effects of increasing group size fromfive to about 30 pigs per pen had minimaleffects on the daily gain of growing andfinishing pigs. Increasing the group sizefrom two to 15 pigs per pen had only mod-est effects on the growth of weanling pigs.3

When pigs were housed in groups of three,five, six, seven, 10, or 15 pigs per pen, atthree different space allowances, increasinggroup size had a negative effect on perfor-mance, and the greatest effect of group sizeoccurred between the very small and theintermediate-sized groups.11 Increasinggroup size from 10 to 15 animals did notaffect performance.11 Conversely, in otherstudies, ADG and ADFI were negativelyaffected during the nursery phase whenspace allowance was decreased from 0.25m2 to 0.16 m2 by increasing the group sizefrom 12 to 18 pigs per pen.5 Similarly,when weanling pigs were housed in groupsof four, but with pen sizes adjusted to pro-vide 0.28 m2 to 0.14 m2 of floor space per

Figure 1: The effect of feeder gap opening on the trough area clear of feed inmultispace, trough-type dry feeders (Generation 3; Staco, Shafferstown,Pennsylvania). Feeder gaps representing manufacturer’s settings of 3 (9.2 mm;Figure 1A), 6 (17.9 mm; Figure 1B), and 10 (31.5 mm; Figure 1C) are shown.

A

B

C

!Beaulieu 3/31/04, 4:16 PM115


pig, the lowest ADG and ADFI wereobserved in groups with the least floorspace.6 We propose, therefore, that the det-rimental effects of the group size-densitytreatments in our study are a result of de-creasing floor space allowance rather thangroup size per se.

It has been suggested that space allowanceshould be based on body surface area, whichis proportional to body weight.12 For grow-ing pigs maintained on fully slatted floors,the Canadian Code of Practice8 recommendsusing the formula Area = k × BW.667,where Area is the recommended floor spaceallowance in meters, BW is body weight inkg, and the coefficient k is .035. Others9

have used .030 as the constant for k. In astudy in which floor space allowances wereadjusted as the pigs grew, productivity wasaffected by crowding with coefficients be-tween .030 and .039.11 In an earlier study,movable partitions were used to adjustfloor space allowances in pens containingseven pigs.4 As a result of this work, aminimum floor space allowance on slattedfloors of 0.27 m2 per pig was proposed3 forpigs weighing 11.5 to 18.0 kg, and 0.37m2 per pig for pigs weighing 18 to 45.5 kg(corresponding to the group sizes of 16 and20 pigs per pen in our experiment). OnDay 0 of our experiment, pigs were allowedbetween 300% (16 pigs per pen) and al-most 190% (24 pigs per pen) of the spaceallowance recommended in the CanadianCode of Practice for the Care and Handlingof Pigs.8 However, by Day 42, when bodyweight had increased to almost 30 kg, thefloor space allowance provided only 105%

Table 4: The impact of feeder gap and group size-density on the time nursery pigs spent eating (minutes/day) during thefirst and sixth weeks after feeder gap adjustment1

1 A 42-day study was conducted (Table 1). Time spent eating was determined by videotaping the feeder area of each pen for one 24-hour period during days 3 to 6 and 39 to 42. Data were analyzed by analysis of variance with pen the experimental unit, using a 5 × 3factorial arrangement of treatments. The model included the effects of replicate, group size-density, feeder gap adjustment, and theinteraction of group size-density and feeder gap adjustment.

2 Floor space allowance per pig at the respective group sizes.3 Calculated using the average number of pens per feeder gap size.4 P value for the group size-density treatments.5 P value for the interaction of feeder gap opening and group size-density treatment.

yaDydutS )yad/nim(emitgnitaE )yad/nim(emitgnitaE P seulav)mm(pagredeeF ytisned-ezispuorG

m( 2 )gip/ 2

2.9 8.11 9.71 8.42 5.13 MES 3 P eulav 42)32.0(

02)82.0(

61)53.0(


6-3syaD 241 811 521 611 3 3 310. 221 721 121 3 095. 640.24-93syaD 79 09 58 97 57 6 582. 28 58 88 5 436. 793.

24

26

28

30

32

9.2 11.8 17.9 24.8 31.5

Feeder gap opening (mm)

Body

wei

ght D

ay 4

2 (k

g)

16 pigs/pen 20 pigs/pen 24 pigs/pen

Figure 2: Impact of feeder gap opening and group size-density on final bodyweight of 716 pigs weaned at approximately 18 days of age (Day -8) andhoused in groups of 16, 20, or 24 per pen in identical pens (5.58 m2). Crumbledfeed was provided in multispace, trough-type dry feeders. On Day 0, one pen ofeach group size was assigned to one of four feeder gap sizes (three pens perfeeder gap size). Feeder gap was defined as the size of the opening thatallowed feed to flow into the trough of the feeder. Feeder gaps used in the firstreplicate were 9.2, 11.8, 17.9, and 24.8 mm. As the 9.2-mm gap restricted accessto feed to a degree that compromised animal welfare, the feeder gaps used inthe second and third replicates were 11.8, 17.9, 24.8, and 31.5 mm (Table 1).Pigs were weighed individually at weaning and on Days 0, 21, and 42. Datawere analyzed by analysis of variance with pen the experimental unit andusing initial weight as a covariate. The interaction of feeder gap opening andgroup size-density was significant (P < .05).

(16 pigs per pen) to 70% (24 pigs per pen)of the Canadian recommendations.8 Themost crowded treatment in our experiment(24 pigs per pen, 0.23 m2 per pig) providedless than 100% of the Canadian Council

on Animal Care recommendation by Day28 and less than 100% of the recommenda-tion cited by Wolter et al9 by Day 35. Ac-cording to these recommendations, thepigs in our experiment were crowded by

!Beaulieu 3/31/04, 4:16 PM116


Figure 3: Impact of feeder gap opening and group size-density on total timespent eating in 716 nursery pigs in the study described in Figure 2. During Days3 to 6 and 39 to 42, the feeder area of each pen was videotaped for one 24-hourperiod, using a time-lapse recorder. Instantaneous scan samples were used todetermine total duration of eating for the entire day. Data were analyzed byanalysis of variance with pen the experimental unit, using a 5 × 3 factorialarrangement of treatments. The model included the effects of replicate, groupsize-density, feeder gap opening, and the interaction of feeder gap openingand group size-density. The interaction of feeder gap opening and group size-density was significant (P < .05).

0

50

100

150

200

9.2 11.8 17.9 24.8 31.5

Feeder gap opening (mm)

Eati

ng

tim

e (m

in/d

ay)

16 pigs/pen 20 pigs/pen 24 pigs/pen

the second half of the trial, which is indeedwhen detrimental effects of crowdingbecame apparent.

In agreement with what has been observedby others,3,5,6,13–15 reducing the floor spaceallowance by increasing animal density wasassociated with lower ADG and ADFI, butthere was no effect on feed efficiency. Thissuggests that the decrease in gain was pri-marily caused by the reduction in feed in-take. In fact, Kornegay and Notter3

concluded that 75% of the decrease indaily gain could be explained by a reduc-tion in feed intake. Similar to what wasobserved with ADG, this effect is probablyassociated with crowding (ie, reduced floorspace allowance), not group size per se.When group sizes of 6, 18, and 36 weremaintained at a constant floor space allow-ance, group size had no effect on feed intake,but feed efficiency was reduced in thelarger groups.16 These authors observedhigher activity levels, and therefore greaterenergy expenditures, in the larger groups.In our experiment, as the pigs increased insize and weight, their space available formovement and other activities became morerestricted. In other studies, concentratingeither the energy17 or lysine6 content of the

diet did not overcome the decrease in per-formance associated with crowding. Thisindicates that the problem is not a restric-tion in ability to consume feed, but a re-duction in appetite, as suggested by others,18

or is perhaps due to increasing aggressivebehaviors. We did not measure dominantor subordinate behavior; however, we didobserve an increase in overall lesion scorewhen the pigs were more crowded, suggest-ing increased aggression.

Similar to what was observed with thegroup size-density treatments, effects asso-ciated with the feeder gap opening werenot apparent until the latter half of the ex-periment. There was an interaction offeeder gap opening and group size-densityon gain in the final week of the experiment.The poorest performance was observed inthe pen with the most pigs (ie, smallestfloor space allowance) and the smallestfeeder gap opening. This demonstrates thatthe pigs could tolerate some crowding ifother resources were not limiting. Typically,when experiments are designed to examinethe impact of crowding or floor space al-lowance on pig performance, feeders areadjusted so that the availability of feederspace per pig is maintained regardless of

group size.5,6,9,16 However, in a trial thatdid examine the interaction of group sizeand feeder space, there was no effect of theinteraction between the two treatments onperformance.19

The size of the feeder gap also contributedto differences in body weight, ADG, andADFI. A tighter feeder adjustment de-creased the ease and speed at which thefeed could be accessed. Feed flow in thefeeders with the smaller openings was fre-quently blocked, while those with a largergap size remained trouble free and allowedfor a constant supply of feed. Reducing thesize of the feeder gap not only increasedlabor, but restricted the pigs’ access to feedto times when the animal technician waspresent to assist with feed flow. The lowerADG, ADFI, and average weight observedin these pigs may have occurred becausetheir ability to access feed was less thantheir ability to consume it. Average dailygain and average weight were not significantlyaffected during the first few weeks of thetrial when nutrient needs were less, as thepigs could access feed as fast as they couldconsume it, regardless of treatment group.There was a positive linear relationship be-tween the feeder gap opening and finalbody weight. The shape of the responsecurve also tended to be quadratic. Aninflection point of 25.1 mm was calculated;however, we feel this should be viewed withcaution, as the final weight of pigs in penswith gaps of 17.9 to 31.5 mm varied byless than 0.4 kg (1.0%).

The optimum feeder gap size changes withthe physical consistency of the feed. There-fore, to provide practical recommendations,feeders were photographed, and the percentof the trough that was clear of feed wasdetermined. On the basis of improvementsin ADG and feed efficiency, the optimalgap size in our experiment, when feed wasprovided as a crumble, occurred at inter-mediate gap openings of 17.9 mm to 24.8mm. This corresponded to approximately56% to 26% of the feeder trough clear offeed. We propose that this optimal troughcovering is independent of the physicalconsistency of the feed.

Younger pigs spent more time eating. Theamount of time a pig spends eating beginsto decline at about 21 days postweaning.14

Older pigs spend less of their time eatingbecause they eat faster and consume largeramounts in a single meal.2 The interactionof group size-density and feeder adjustmenton eating time illustrates that pigs in the

!Beaulieu 3/31/04, 4:16 PM117


smaller groups compensated for restrictionof feed availability by spending more timeat the feeder. The effect of density-groupsize on eating time was no longer apparentat larger feeder gap openings.

The performance of older pigs was affectedby feeder adjustment because they couldnot access feed as fast as they could eat, orbecause they were not prepared to increasethe proportion of their time spent eating inorder to obtain feed. Interestingly, in ourexperiment, the pigs did not change theirdiurnal pattern of eating with differentgroup sizes or feeder gap openings. Con-versely, when more pigs were fed from asingle feeding space, the diurnal pattern offeeding was altered.18 Using electronic feedintake recording equipment, Hyun andEllis (2002)20 showed that finishing pigsmaintained intake and growth rate bychanging feeding behavior when group sizeincreased from two to 12 pigs per pen;however, this was not the case in growingpigs.21 Crowding of older pigs resulted infewer but longer feeder visits and higherfeed intake per visit.22

All pigs in our experiment had access to thesame number of feeding spaces regardlessof pen density; thus, it can be concludedthat feeders that allow for easier feed accessalso allow for more pigs to eat from eachfeeding space. Therefore, by optimizing thefeeder gap, the number of pigs able to uti-lize a given feeding space increases.Gonyou18 found that pigs with easier ac-cess to feed spent less time at the feeder,allowing for more pigs to be fed from asingle feeding space, and concluded thatfeeder capacity was achieved when it wasbeing used 90% of the time. In our experi-ment, this implies that the maximum ca-pacity is 11 pigs for a nursery feederadjusted to a 25-mm or 32-mm gap open-ing. This, of course, would change for feedwith different physical characteristics andfor different feeder designs.

Lower joint swelling scores and lesions as-sociated with fighting or biting were ob-served in pens with smaller group sizes.This is not surprising, as more floor spaceresults in less direct interaction amongpigs; however, changes in these indicatorsof aggressive behavior were not associatedwith changes in animal performance. Simi-larly, the overall lesion score tended to behigher when the feeder gap opening wasthe smallest. Spoolder et al23 showed thatthe number of aggressive interactions at thefeed trough was not affected by group size,

but did decrease with the number of feederspaces available.

Producers strive to market uniform groupsof pigs. As observed by others at our re-search facility1 and elsewhere,4 the variabil-ity in pig weight was not affected by groupsize-density. The present experiment wasconcluded when the pigs were approxi-mately 10 weeks of age. However, in anearlier study,1 in which pigs were housedeither 16 per pen (0.35 m2 per pig) or 21per pen (0.26 m2 per pig), lack of an effectof crowding on variability in body weightwas maintained at Day 140.

Providing more floor space per pig in thisexperiment was associated with increasedbody weight at approximately 10 weeks ofage. The increased body weight would notbe sufficient to compensate for the smallernumber of pigs housed, thus the kg of porkproduced per floor space in the nurserywould be less at the larger floor space al-lowances. However, less crowding wouldresult in fewer days to market, which mightimprove the overall cost effectiveness of theswine production unit.

Implications• Under the conditions of this study,

increasing space allowance from 0.23 m2

per pig to 0.35 m2 was associated withincreased body weight at approximately10 weeks of age.

• Optimal growth and feed efficiencywere observed when the feeder gapopening allowed 25 to 60% of thefeeder trough to be consistently clearof feed.

• The capacity of a nursery feeder spaceis 11 pigs when the feeder gap openingallows 25 to 60% of the feeder troughto be clear of feed.

References*1. Cooper DR, Patience JF, Gonyou HW. Effect ofgender and crowding on variation in days to market.Annual Research Report. Prairie Swine Centre Inc,Saskatoon, Saskatchewan. 2001.2. Gonyou HW, Lou Z. Effects of eating space andavailability of water in feeders on productivity andeating behavior of grower/finisher pigs. J Anim Sci.2000;78:865–870.3. Kornegay ET, Notter R. Effects of floor space andnumber of pigs per pen on performance. Pig NewsInfo. 1984;5:23–31.4. Gehlbach GD, Becker DE, Cox JL, Harmon BG,Jensen AH. Effects of floor space allowance andnumber per group on performance of growing-finishing swine. J Anim Sci. 1966;25:386–391.5. Brumm MC, Ellis M, Johnston LJ, RozeboomDW, Zimmerman DR, NCR-89 Committee onSwine Management. Interaction of swine nurseryand grow-finish space allocations on performance. JAnim Sci. 2001;79:1967–1972.

6. Kornegay ET, Lindemann MD, Ravindran V.Effects of dietary lysine levels on performance andimmune response of weanling pigs housed at twofloor space allowances. J Anim Sci. 1993;71:552–556.7. Bryant MJ, Ewbank R. Some effects of stockingrate and group size upon agonistic behaviour ingroups of growing pigs. Br Vet J. 1972; 128:64–70.8. Connor ML. Recommended code of practice for thecare and handling of farm animals. Pigs. Agricultureand Agri-Food Canada; 1993. Publication #1898/E.9. Wolter BF, Ellis M, Curtis SE, Parr EN, WebelDM. Group size and floor-space allowance can af-fect weanling-pig performance. J Anim Sci.2000;78:2062–2067.10. McGlone JJ, Newby BE. Space requirements forfinishing pigs in confinement: behavior and perfor-mance while group size and space vary. Appl AnimBehav Sci. 1994;39:331–338.11. Gonyou HW, Stricklin WR. Effects of floor areaallowance and group size on the productivity ofgrowing/finishing pigs. J Anim Sci. 1998;76:1326–1330.12. Hurnik JF, Lewis NJ. Use of body surface area toset minimum space allowances for confined pigs andcattle. Can J Anim. Sci. 1991;71:577–580.13. Jensen AH, Baker DH, Harmon GB, WoodsDM. Response of growing-finishing male and fe-male swine to floor space allowance on partially andtotally slotted floors. J Anim Sci. 1973; 37:629–631.14. Spicer HM, Aherne FX. The effects of groupsize/stocking density on weanling pig performanceand behavior. Appl Anim Behav Sci. 1987;19:89–98.15. Lindvall RN. Effect of flooring material andnumber of pigs per pen on nursery pig performance.J Anim Sci. 1981;53:863–868.16. Petherick JC, Beattie AW, Bodero DAV. Theeffect of group size on the performance of growingpigs. Anim Prod. 1989;49:497–502.*17. Levesque CL, Patience JF, Beltranena E, ZijlstraRT. Response to dietary energy concentration andstocking density in weaned pigs [abstract]. Adv PorkProd. 2002;13:21.*18. Gonyou H. Feeder and pen design to increaseefficiency. Adv Pork Prod. 1999;10:103–113.19. Lindemann MD, Kornegay ET, Meldrum JB,Schurig G, Gwazdauskas FC. The effect of feederspace allowance on weaned pig performance. J AnimSci. 1987;64:8–14.20. Hyun Y, Ellis, M. Effect of group size and feedertype on growth performance and feeding patterns infinishing pigs. J Anim Sci. 2002;80:568–574.21. Hyun Y, Ellis M. Effect of group size and feedertype on growth performance and feeding patterns ingrowing pigs. J Anim Sci. 2001;79:803–810.22. Hyun Y, Ellis M, Johnson RW. Effects of feedertype, space allowance, and mixing on the growthperformance and feed intake pattern of growingpigs. J Anim Sci. 1998;76:2771–2778.23. Spoolder HAM, Edwards SA, Corning S. Effectsof group size and feeder space allowance on welfarein finishing pigs. Anim Sci. 1999;69:481–489.

* Non-refereed references.

!Beaulieu 3/31/04, 4:16 PM118

Peer reviewed Original research - AASV

Documents

Transcript of Peer reviewed Original research - AASV