Stall Usage Differences of Thirteen Different Freestall Base Types by xiuliliaofz

VIEWS: 26 PAGES: 13

									                                       The Professional Animal Scientist 20 (2004):470–482




                Stall UsageFreestall BaseofTypes
                 Different
                            Differences Thirteen
                           W. K. FULWIDER and R. W. PALMER1
                           Department of Dairy Science, University of Wisconsin, Madison 53706


                                                one RMATR, and a RMAT were infe-          tivity, lying or standing in stalls, or
Abstract                                        rior. Waterbeds were the stall base of    feeding, via videotape images, has
                                                choice during the coldest weather. In     been done in several studies (Sonck
   The objective of this study was to           both pens, cows preferred the exterior    et al., 1999; Anderson, 2002;
compare percentages of cow preference           (EXTR) row during evening and early       Overton et al., 2002; Gaworski et al.,
measures for factors affecting use of           morning. The addition of a foam layer     2003; House et al., 2003; Wagner-
stalls with different stall bases. Stall sta-   increased use of mattresses and mats.     Storch et al., 2003). Using videotape
tus was recorded four times each day in                                                   images to record cow activity re-
two pens with different stocking densi-         (Key Words: Stall Use, Freestall Base,    moves the potential influence of hu-
ties for a 6-mo period starting June 19,        Cow Preference, Stocking Density.)        man presence on cow activity.
2002. Six factors were analyzed: stall                                                       A few studies show stall base, stall
base, distance to water (WDIST), stall                                                    design, length of exposure time, stall
location within stall base section              Introduction                              or row location, stocking density
(STLLOC), stalls on interior or exterior                                                  (SD), and temperature as effects on
row (RLOC), inside barn temperature,               The high production levels ex-
                                                                                          stall use. Different stall base studies
and length of time cows were exposed            pected of a modern dairy cow in-
                                                                                          report conflicting stall usage results.
to stall bases (XPOSR). The 101%                crease the need to provide a safe,
                                                                                          This study was conducted to deter-
stocking density (SD) pen had eight dif-        comfortable environment for her.
                                                                                          mine cow preference for 13 different
ferent bases: two foam mattresses               Cows spend 10 to 14 h/d lying
                                                                                          freestall bases, as well as five other
(FMATR), three rubber mattresses                down (Gaworski et al., 2003; McFar-
                                                                                          factors affecting stall use: distance to
(RMATR), and three rubber mats                  land, 2003). The stall surface must
                                                                                          water (WDIST), stall location within
(RMAT). The 66% SD pen had seven                be non-abrasive, to prevent hock le-
                                                                                          a stall base section (STLLOC), stalls
different stall bases, two identical to the     sions, and compressible, to prevent
                                                                                          on interior or exterior row (RLOC),
first pen: a cork-filled mattress                 knee swelling and injury. Cows with
                                                                                          inside barn temperature, and length
(CMATR), two FMATR, two RMATR,                  the advantage of a softer base are
                                                                                          of time cows were exposed to stall
a waterbed, and a RMAT. Each pen                known to stand up and lie down
                                                                                          bases (XPOSR).
was analyzed separately. Stall bases            twice as often as cows on concrete
were grouped with 3 to 7 stalls per sec-        (Rushen and de Passille, 1999). A
tion and randomly placed in each row.           cow should be able to perform natu-       Materials and Methods
In the 101% SD pen, FMATR and                   ral movements when getting up and
RMATR had the greatest lying and occu-          lying down without injuring herself          This study was conducted at the
pied percentages, and two RMAT had              (Rushen and de Passille, 1999). Evalu-    University of Wisconsin-Madison Ar-
the least lying and occupied percent-           ating cow preference for different fre-   lington Agricultural Research Station
ages. In the 66% SD pen, one FMATR              estall bases is a useful means for de-    freestall barn from June 19, 2002 to
and one RMATR had the greatest lying            termining cow comfort. Providing a        December 17, 2002. The facility is a
and occupied percentages, waterbeds             comfortable environment enhances          naturally ventilated, four-row, 104-
were intermediate, and the CMATR,               health, production, longevity, and        stall, tail-to-tail barn that is not insu-
                                                overall profitability of a dairy           lated. The barn has the following di-
                                                business.                                 mensions: 30.5 m (100 ft) × 36.6 m
                                                   Preference testing is an effective     (120 ft) long; the roof pitch is 4/12.
1
 To whom correspondence should be ad-           means of measuring cow comfort.           The barn is oriented East to West.
dressed: rwpalmer@facstaff.wisc.edu             Observation of stall use and cow ac-      Eave sidewall height is 3.7 m (12 ft).
                                  Stall Usage Differences of Thirteen Different Stall Base Types                           471


The barn has a wood-post frame             north pen, beginning with a high              Part 1 research, stall numbers 13, 14,
structure with 15.2-cm × 15.2-cm (6-       76% SD and ending in December                 38, and 39 in the south pen and
in × 6-in) wood posts situated at the      with a low 44% SD, for an average             stall numbers 17, 18, 19, 50, 51, and
feedbunk line and at the front of the      66% SD. The south pen averaged                52 in the north pen were RMATR1
internal row of stalls. Ventilation is     101% SD, with a relatively steady             stalls. Stall numbers 15, 16, and 37
controlled by a 16.5-cm (6.5-in) eave      SD, having a high 105% SD in July             in the south pen and stall numbers
opening and 2.7-m (9-ft) adjustable        and a low 92% SD in October. Cows             14, 15, 16, 53, and 54 in the north
curtains. The barn has a 61-cm (24-        on the 66% SD side had to pass                pen were RMATR2 stalls. In Part 2 re-
in) ridge opening. Fans and sprin-         through the robot to access the feed          search, stall numbers 21 to 23 and
klers were not used. Stalls were 1.2       alley and through a one-way gate on           48 to 50 in the south pen were filled
m (46 in) wide and 2.4 m (8 ft)            the west end to return to the fre-            with small rubber chunks (0.63-cm
long. Brisket boards were 3.8-cm ×         estall area. The STLLOC was defined            in diameter) (RBRC1), and stall num-
19.1-cm (1.5-in × 7.5-in) boards           as END or NOTEND. Stalls on row               bers 25 to 27 and 44 to 47 in the
attached to and below stall dividers,      ends or next to a different stall base        south pen were filled with small rub-
approximately 1.7 m (66 in) from           were END stalls; all others were clas-        ber buffings (RBRC2) stalls. Observa-
the rear curb, and extended down to        sified as NOTEND.                              tions were taken four times/d at
the top of the stall surface. Neck            The number positioned directly be-         1400, 2000, 0400, and 0900 h from
rails were mounted 1.14 m (45 in)          neath every stall is the WDIST, mea-          June 19, 2002 until December 17,
above the top of the rear curb or ap-      sured in water units; 1 water unit            2002. The 1400-h observation time
proximately 1.07 m (41 in) above           was approximately one stall width.            allowed for possible delays in the reg-
the stall surface. All stall bases had a   The WDIST was measured to the                 ularly scheduled 1200-h tape
concrete base with a 7.6-cm (3-in)         nearest water site. The XPOSR term            change. Regularly scheduled dark
upward slope toward the brisket            indicates the amount of time cows             times were from 2130 to 0330 h.
board. Stalls were bedded with saw-        were exposed to their pen; these              The 0400-h observation time al-
dust twice per week, and soiled bed-       were summarized by month of obser-            lowed stall use observation after the
ding was removed following the             vation.                                       lights were turned on. The 2000 and
scheduled milking (parlor system) or          Earlier research in this barn              0900-h observation times allowed
human intervention time (automatic         (Wagner-Storch et al., 2003) evalu-           for observations 2 h after scheduled
milking system). Stall bedding dates       ated stall use for six different stall        parlor milking times and allowed for
were not recorded, so stall usage by       base types (concrete, sand, a foam            variation in milking times and times
days since bedding was not an-             mattress [FMATR], a rubber mattress           for cows to eat after returning from
alyzed.                                    [RMATR], a rubber mat [RMAT], and             the parlor.
   A closed circuit monitor camera         a waterbed). It was later discovered             When viewing videotape images
system, with one camera on each            that one manufacturer (Pasture Mat)           and recording observations, the fol-
side, from ADT Security Systems,           had installed two different versions          lowing steps were used. Stalls were
Inc. (Menlo Park, CA) was used to          of a mattress (RMATR1 and                     observed in sequential order on the
observe cow activity 24 h/d. A video       RMATR2). The mattresses both had              south side of the barn and assigned
recorder (Pelco TLR2096; Clovis, CA)       the same type of bottom layer and             a status of empty; cow lying in stall;
and Simplex Monochrome Multi-              top cover, but the center layer was           cow standing half-in, half-out; cow
plexer (Pelco MX4016MS) were used          different. One had a 2.5-cm (1-in)            standing in stall; or unsure. Unsure
to record digital images. Stalls were      layer of foam rubber, and the other           was the designation for stalls that
panned in a specific order by video         had a 1.9-cm (0.75-in) layer of felt.         could not be accurately recorded as
cameras. Stall numbers were as-            For this research, mattress types with        any of the aforementioned. Stalls in
signed accordingly for identification       high stall preference values were re-         the north pen were then similarly
purposes. The RLOC was defined as           tained (i.e., RMATR1, RMATR2, a               observed.
stalls on interior (INTR) or exterior      FMATR, and a waterbed), and nine                 Five HOBO H8 Pro Series (Onset
(EXTR) rows of the barn. Interior          new stall base types were installed.          Computer Corporation, Pocasset,
rows were located near the feed            Stall status was recorded for a cork-         MA) sensors recorded temperature ev-
alley, whereas stalls on the outside       filled mattress (CMATR), 4 FMATR, 3            ery 15 min. Individual sensors were
of the building were EXTR. There           RMATR, 4 RMAT, and a waterbed                 positioned in each quadrant of the
were two pens in the barn: the             (Table 1).                                    barn [northeast (NE), northwest
north pen, which housed cows                  The freestall barn layout and obser-       (NW), southeast (SE), southwest
milked with a robotic milker, and          vation recording sheet for part 3 of          (SW)] 11.6 to 12 m (38 to 40 ft)
the south pen, which housed cows           this research (not to scale) is shown         from each end wall and 2.7 to 3.0 m
milked 2× in a conventional parlor.        in Figure 1. Stalls were identified by         (9 to 10 ft) above the alley, over the
The SD decreased steadily in the           stall number and stall base type. In          front of the INTR row of stalls. The
472                                                   Fulwider and Palmer




  TABLE 1. Descriptive code, product name, supplier name and address, product classification, and
  description of all stall base types tested.
  Item                    Product namea                   Product supplier          Supplier address        Product class
  CMATRb               Cow Comfort Corkmat               Amorim Solutions               Trevor, WI        Cork-filled mattress
  FMATR1c              Comfy Cow Mattress               Sikkema’ Equipment          Byron Center, WI      Foam-filled mattress
  FMATR2d            Foamat Bedding System™              Foxworthy Supply             Kent City, MI       Foam-filled mattress
  FMATR3e          DeLaval Cow Mattress M100                DeLaval, Inc.           Kansas City, MO       Foam-filled mattress
  FMATR4f              Comfy Cow Mattress               Sikkema’ Equipment          Byron Center, WI      Foam-filled mattress
  RMATR1-Newg      Pasture Mat with foam insert        Dodgeland Ag Systems           Platteville, WI    Rubber-filled mattress
  RMATR1g          Pasture Mat with foam insert        Dodgeland Ag Systems           Platteville, WI    Rubber-filled mattress
  RMATR2h           Pasture Mat with felt insert       Dodgeland Ag Systems           Platteville, WI    Rubber-filled mattress
  RMATR3i               Ulti-Mat All Rubber             Zartman Farms Cow
                           Cow Mattress                   Comfort Systems              Ephrate, PA       Rubber-filled mattress
  RMAT2j               Supreme Comfort Pad               Humane Mfg. LLC               Baraboo, WI           Rubber mat
  RMAT3k                Kraiburg Softbed II                   Agromatic              Fond du Lac, WI         Rubber mat
  RMAT4l                    Heavy Cush                       J & D Mfg.               Eau Claire, WI         Rubber mat
  RMAT5m              Berg-Simplex Dairy Pad        Berg Equipment Corporation        Marshfield, WI          Rubber mat
  WATRn                   Cow Waterbed                 Relative Products, LLC         Reedsburg, WI           Waterbed

  a
    Products listed were donated and installed by the supplier listed. Listing of products by product name is not an
  endorsement for the product.
  b
    The CMATR-based stall was a multi-celled, cork-filled mattress with top cover.
  c
    The FMATR1-based mattresses were comprised of cross-linked, closed-cell, non-absorbent, polyethylene foam and vinyl
  from the automobile industry and a Tafcoat waterproof cover. The Tafcoat cover is a composite material composed of two
  layers of bonded polyester, a non-woven fabric derived from recycled soft drink bottles. It is an impervious membrane of
  olefin copolymer resin, providing a chemical resistant, waterproof barrier.
  d
    The FMATR2-based stalls contained a 7.5-cm (3-in) waterproof soft foam mat with a waterproof, premium rubber top
  cover.
  e
    The FMATR3-based stalls consisted of composite foam in PVC-sealed envelope with a PVC cover.
  f
    The FMATR4-based stalls were made of cross-linked, closed-cell, non-absorbent polyethylene foam and a geo-textile cover
  with urethane topcoat.
  g
    The RMATR1-based stalls and RMATR1-new based stalls were comprised of a multi-celled, rubber crumb-filled mattress with
  a 2.5-cm (1-in) foam pad and needle-punched polypropylene top cover impregnated with wax to increase water shedding.
  h
    The RMATR2-based stalls consisted of a multi-celled, rubber crumb-filled mattress with a 1.9-cm (0.75-in) felt pad and
  needle-punched polypropylene top cover impregnated with wax to increase water shedding.
  i
    The RMATR3-based stalls were 6.8-cm (3-in) thick and were made of recycled tire rubber, which was mixed with pliable
  polyurethane adhesive mold-formed under pressure. The top cover was non-woven and water-resistant.
  j
    The RMAT2-based stalls were 4.4-cm (1.75-in) thick and were comprised of a foam cushion completely encased in rubber.
  k
    The RMAT3-based stalls were comprised of a solid rubber anti-slip cover over a 2.2-cm (0.88-in) foam pad. The rear one-
  third of the mat was sloped to enhance drainage.
  l
    The RMAT4-based stalls were 3.1-cm (1.25-in) thick vulcanized rubber with a pebble top and large corrugated
  undersurface.
  m
     The RMAT5-based stalls consisted of 3.2-cm (1.4-in) vulcanized rubber with a textured surface and deeply grooved
  bottom.
  n
    The WATR-based stalls were made of two layers of 100% natural vulcanized rubber. Each layer had a fiber layer and six
  plies of rubber, and each stall was individually filled with 11.4 L (3 gal) of calcium chloride and 49.2 L (13 gal) of water.




fifth sensor was in a radiation shield        Temperature data from the nearest         0 to 10 min. Stalls 1 to 13 and 40 to
on the north side of the barn 6.4 m        sensor were assigned to each stall for      50 in the south pen were assigned
(21 ft) from the barn and 1.1 m (3.5       each observation. Because of differ-        the temperature recorded from the
ft) above the ground. Approximate          ent systems for collecting barn tem-        sensor SE. Stalls 14 to 39 in the
location of the four inside sensors        perature data and stall occupancy,          south pen were assigned the tempera-
(NE, NW, SE, and SW) is illustrated        the actual stall’s status observation       ture recorded by sensor SW. Temper-
in Figure 1.                               time differs from the temperature by        atures recorded from sensor NW
                                    Stall Usage Differences of Thirteen Different Stall Base Types                               473




Figure 1. Freestall barn layout and observation recording form (not to scale; 13 stall base types). WU = Water unit. See Table 1 for
manufacturer information.


were assigned to Stalls 1 to 16 and           statements were used to determine             Storch et al. (2003) to see if there
43 to 54 at the 66% SD. Finally, tem-         significant differences between per-           was a difference between the two
peratures recorded from sensor NE             centages (Stokes et al., 1995). Each          Pasture Mat mattress types (RMATR1
were assigned to Stalls 17 to 42 at           side was analyzed separately because          and RMATR2; Table 2). The RMATR1
the 66% SD.                                   of different SD in the south and              was found to have significantly
   Percentages of each status (empty;         north pens. Stocking densities were           higher stall usage values than
lying; standing; half-in-half-out; un-        a function of herd management and             RMATR2. The RMATR1 was equal to
sure) were calculated as the number           were not set intentionally as part of         washed mason sand for lying per-
of stall-day status observations di-          the study’s design.                           centage. Both sand and RMATR1
vided by total number of stall-day               The CATMOD Procedure in SAS                had larger lying percentages than
observations for the different catego-        was used to model lying and occu-             RMATR2, and RMATR2 was superior
ries or factors tested across the 6-mo        pied as binary outcomes. Indepen-             to the other stall base types tested.
study. Stall occupied percentages (ly-        dent variables considered continuous          Ranking of stall base types by per-
ing; standing; half-in-half-out) were         were XPOSR, WDIST, and tempera-               centage occupied was the same ex-
calculated using stall-day lying;             ture. Independent variables consid-           cept RMATR1 was higher than
standing; standing half-in, half-out;         ered categorical were stall base,             RMATR2. Because cow preference for
and unsure observations and divided           RLOC, STLLOC, and side. Any two-              the RMATR1 and RMATR2 base
by the total number of stall-day ob-          way interactions deemed biologically          types appeared to be different, these
servations. The unsure percentages            significant or meaningful were also            stall base types were analyzed sepa-
were very small and had little im-            analyzed.                                     rately for the remainder of the re-
pact on other stall statuses.                                                               search.
   Percentages for different factors an-      Results and Discussion                          For Part 2 of our research, two
alyzed using logistic regression with                                                       crushed rubber (RBRC1 and RBRC2)
the GENMOD procedure in SAS (SAS                Part 1 of this research was to re-          types were used to replace sand in
Inst., Inc., Cary, NC) and contrast           analyze the data from Wagner-                 the sand stalls on the south side of
474                                                 Fulwider and Palmer




  TABLE 2. Stall status percentagesa by baseb for each barn sidec (five types plus RMATR split)d.
  Item                  CONC       FMATR1      RMATR1       RMATR2        RMAT1         SAND       WATR       X       SEf
  100% SD (South)g
   Empty                 61.2t        15.5x      10.5z        13.3y       34.7v           19.8d      38.2u   29.3    0.007
   Half-in, half-out      7.0uv        6.0w       6.3vw        5.8w        7.3u            7.0bc      8.3t    7.0    0.003
   Lying                 22.8y        57.4v      68.4t        60.9u       32.9x           68.7a      45.4w   51.0    0.008
   Occupiedh             38.7z        84.1v      89.4t        86.8u       64.8x           79.0d      61.6y   70.1    0.007
   Standing               8.8w        20.7u      14.7v        20.0u       24.6t            3.3e       7.9w   12.2    0.005
   Unsure                 0.2v         0.4u       0.1v         0.0v        0.5u            1.2a       0.2v    0.5    0.001
   Total numberi       7688         6727       3844         2883        6727          13,454       6727
  Low SD (North)j
   Empty                 83.7t        44.7x      31.9y        50.5w        81.0u                     65.4v   61.4    0.007
   Half-in, half-out      2.9v         4.2t       3.7tu        3.1uv        2.2w                     25.6w    3.4    0.002
   Lying                 10.3x        39.1u      53.4t        35.6v        11.8x                     25.6w   27.6    0.003
   Occupied              38.7z        84.1v      89.4t        86.8u        64.8x                     61.6y   70.1    0.007
   Standing               5.1v        11.9t      11.0tu       10.6u         5.1v                      3.9w    7.3    0.003
   Unsure                 0.1u         0.1u       0.0u         0.2u         0.0u                      1.1t    0.3    0.001
   Total number        7130       10,812       5898         4915           10.813                  9830

  a
    Percentage = number of stall-day-status observations/total number of stall-day observations by stall base by side.
  b
    CONC = concrete, FMATR1 = Comfy Cow Mattress with Cow-Flex Top Cover (Sikkema’s Equipment, Byron Center, MI),
  RMATR1 = Pasture Mat with foam layer (Promat Ltd., Seaforth, Ontario, Canada), RMATR2 = Pasture Mat with felt layer
  (Promat Ltd.); RMAT1= Comfort Zone-Milk Mat (Alfa Laval-Agri, Kansas City, MO); SAND = sand stalls, and WATR = Cow
  Waterbed (Relative Products, Reedsburg, WI).
  c
    Data collected from May 9, 2001 to February 9, 2002.
  d
    RMATR1 and RMATR2 were found to be significantly different from one another for both sides. In the following studies,
  they are treated as two different stall base types. Columns other than RMATR1 and RMATR2 were published in a previous
  study (Wagner-Storch et al., 2003).
  e
    Mean percentages are total stall-day status observations/total number stall-day observations by side.
  f
    Pooled SE for each status by side.
  g
    South side of the freestall barn with 100% stocking density.
  h
    Occupied is defined as lying + standing + half-in, half-out.
  i
    Total number of stall-day observations by stall base by side.
  j
    North side of the freestall barn with low stocking density.
  t,u,v,w,x,y,z
                Means within row with different letters are significantly different (P<0.05).




the barn. Stall bases previously filled    RMATR2 for lying percentage. Of             dictors: intercept, stall base type,
with sand were filled and main-            the stall bases tested, RBRC1 was in-       RLOC, STLLOC, side of barn,
tained the same as sand stalls. The       ferior to four, equal to one, and su-       XPOSR, WDIST, temperature, and
objective of this effort was to deter-    perior to two of the stall bases tested     base interaction with all other pre-
mine the effect on stall usage by us-     for lying percentage. Neither RBRC1         dictors, interactions of RLOC × side
ing RBRC1 or RBRC2 rather than            nor RBRC2 were found to be supe-            of barn, RLOC × temperature, and
sand. It was felt that rubber, if pre-    rior for occupied percentage. It was        side of barn × WDIST. The results
ferred by the cow, might be less          concluded that use of crushed rub-          showed WDIST, STLLOC, RLOC,
damaging to equipment and easier          ber as a stall base filler had no obvi-      temperature, and XPOSR all affect
to separate from manure. The              ous advantage over products cur-            stall use. Table 4 shows the lying
RBRC2 consisted of small rubber           rently being used.                          and occupied model analysis vari-
buffings, and RBRC1 consisted of              Part 3 of this study tested 13 differ-   ables, chi-square values, and P val-
small chunks of rubber with a diame-      ent stall base types for a 6-mo pe-         ues. In the lying model, all P values
ter of about 0.63 cm. For a 30-d pe-      riod (June 19 to December 17,               were significant with the exception
riod (March 1 to March 31, 2002),         2002). Stall use was analyzed using         of base, RLOC, side of barn, RLOC ×
eight different stall bases were ob-      the CATMOD procedure (SAS; Stokes           side of barn, and side of barn ×
served (Table 3). The RBRC2 was sec-      et al., 1995). The lying and occupied       WDIST. In the occupied model, P
ond to RMATR1 and equal to                models included the following pre-          values were significant for all pre-
                                   Stall Usage Differences of Thirteen Different Stall Base Types                            475



    TABLE 3. Stall status percentagesa by baseb for the south barn sidec (five types plus rubber chunks and
    buffings)d.
    Item                 CONC     FMATR1      RBRC1      RBRC2      RMATR1      RMATR2       RMAT1      WATR       X       SEf
    97% SD (South)g
     Empty                60.1u    12.5x      30.0vw      13.9x       3.9y          9.5x       33.2v     27.0w    27.0    0.02
     Half-in, half-out     7.0v     3.5w      12.6u        8.0v       1.8w          2.1w        3.9w      3.9w     5.8    0.01
     Lying                21.9z    57.3w      49.3x       70.0v      76.4u         67.9v       32.1y     54.5wx   50.5    0.02
     Occupiedh            39.9z    87.3v      66.8y       81.7w      96.1u         90.5v       64.4y     72.7x    71.7    0.02
     Standing             11.0x    26.5u       5.0y        4.1y      17.9vw        20.5v       28.2u     14.6wx   15.4    0.01
     Unsure                0.0w     0.3w       3.2u        4.5u       0.0w          0.0w        2.5v      0.3w     1.5    0.004
     Total numberi       872      763        763         763        436           327         763       763

    a
      Percentage = number of stall-day-status observations/total number of stall-day observations by stall base by side.
    b
      CONC = concrete, FMATR1 = Comfy Cow Mattress with Cow-Flex Top Cover (Sikkema’s Equipment, Byron Center, MI),
    RBCR1 = small rubber chunks, RBRC2 = small rubber buffings, RMATR1 = Pasture Mat with foam layer (Promat Ltd.,
    Seaforth, Ontario, Canada), RMATR2 = Pasture Mat with felt layer (Promat Ltd.), RMAT1 = Comfort Zone-Milk Mat (Alfa
    Laval Agri, Kansas City, MO), and WATR = Cow Waterbed (Relative Products, Reedsburg, WI).
    c
      Data collected from March 1, 2002 to March 31, 2002.
    d
      Rubber chunks and crumb rubber replaced sand. Cows preferred rubber buffings to rubber chunks. Rubber buffings were
    not superior to other surfaces, so no further investigation was attempted.
    e
      Mean percentages are total stall-day status observations/total number stall-day observations by side.
    f
      Pooled SE for each status by side.
    g
      South side of the freestall barn with 97% stocking density.
    h
      Occupied is defined as lying + standing + half-in, half-out.
    i
      Total number of stall-day observations by stall base by side.
    u,v,w,x,y,z
                Means within columns with different letters are significantly different (P<0.05).
.



dictors and interactions except side           The stall base effect had signifi-           The RLOC was significant at P<0.01
of barn × WDIST. All stalls were of         cant (P<0.0001) interactions with              in the occupied model, and stall
the same design, so this was not a          RLOC, STLLOC, XPOSR, and                       base and side were significant at
factor.                                     WDIST. Stall base interaction with             P<0.05. Stall base had significant
   According to chi-square values,          temperature was significant at                  (P<0.0001) interaction with RLOC,
side effect, RLOC, and stall base ex-       (P<0.05) in the lying model. This in-          STLLOC, XPOSR, WDIST, and tem-
plained the least amount of varia-          dicated that a cow’s preference to             perature. The RLOC × side was sig-
tion in the lying model, but were           lie on a particular stall base varied          nificant at P<0.05.
significant at (P<0.05) in the occu-         for each of these factors. The RLOC               Table 5 shows the percentage of
pied model. This result indicated           had significant (P<0.0001) interac-             stalls empty; half-in, half-out; lying;
that side and RLOC had the least            tion with temperature, which indi-             standing; or unsure for each barn
impact on whether cows lie in               cated lying percentages on INTR                side for the 13 different stall base
stalls. The side variable was in-           and EXTR rows were different for               types tested. It also shows the per-
cluded in the model to determine            temperature intervals. Base, RLOC,             centage of occupied stalls, which
the effect of SD and barn orienta-          and side were not significant                   combines the three statuses of ly-
tion. Stocking density was not in-          (P<0.05) predictors for lying, but             ing, standing, and half-in, half-out.
cluded as a variable because of its         were kept in the model because of              The 13 different stall base types
confounding with side. Tempera-             interaction with other factors. The            were categorized by type of con-
ture and XPOSR best explained the           RLOC was significant at (P<0.01) in             struction for comparison purposes,
variation in lying status, and              the occupied model, and base and               but overlap in the components of
STLLOC and WDIST were other sig-            side were significant at (P<0.05).              each makes direct comparisons by
nificant (P<0.0001) predictors of ly-           The XPOSR effect best explained             category difficult. Mattresses
ing status. The effects of tempera-         the variation in stall occupied                (RMATR and FMATR) were defined
ture and XPOSR on lying status sug-         (P<0.0001), but STLLOC, WDIST,                 as any stall base having a cover
gested preference for a particular          and temperature effects were also              over some type(s) of interior prod-
base changed over time.                     significant (P< 0.0001) predictors.             uct, whereas RMAT were those stall
476                                                   Fulwider and Palmer




                                                                                    FMATR2, and on the 66% SD side,
  TABLE 4. Lying and occupied model analysis variables, χ2 values,                  FMATR3, were equal to the
  and P values.                                                                     RMATR1 stall base type for lying
                                                                                    and occupied percentages. The
                                Lying model                  Occupied model
                                                                                    RMATR3 was also equal to RMATR1
  Item                     χ2                 P             χ2               P      and FMATR2 for lying percentage
                                                                                    on the 101% SD side. On the 101%
  Intercept               10.87           0.001           26.91           <0.0001
                                                                                    SD side, RMATR2, FMATR1, and
  Basea,b                 20.17           0.0639          24.74            0.0161
  RLOCa,c                  0.74           0.3888           7.72            0.0055   RMAT2 were intermediate, and
  STLLOCa,d               44.42          <0.0001          30.82           <0.0001   RMAT3 and RMAT4 were inferior.
  Sidea,e                  0.22           0.6425           5.70            0.0170   On the 66% SD side, FMATR4 and
  XPOSRf,g                87.45          <0.0001          75.38           <0.0001   waterbed were intermediate, and
  WDISTf,h                29.11          <0.0001          28.28           <0.0001   RMATR2, CMATR, and RMAT5 were
  TEMPf,i                107.69          <0.0001          49.33           <0.0001   inferior. In general, it was con-
  Base × RLOC             42.79          <0.0001          72.91           <0.0001   cluded that FMATR and RMATR
  Base × STLLOC           92.28          <0.0001          80.17           <0.0001
                                                                                    were preferred to RMAT and that
  Base × XPOSR            47.00          <0.0001          44.27           <0.0001
  Base × WDIST            51.68          <0.0001          67.11           <0.0001
                                                                                    the addition of a foam layer in-
  Base × TEMP             31.72           0.0015          44.47           <0.0001   creased the use of mattresses and
  RLOC × Side              0.04           0.8459           5.26            0.0219   mats. Stall lying percentages for
  RLOC × TEMP             25.68          <0.0001          59.84           <0.0001   RBRC1 and RBRC2 were similar to
  Side × WDIST             0.22           0.6383           2.14            0.1436   those reviewed by (Rodenburg and
                                                                                    House, 2000) and those reported by
  a
     Independent variables were considered categorical in model analyses.           (Sonck et al., 1999). The waterbed
  b
     Base represents 13 different freestall bases in this study.                    percentage (40.2%) was comparable
  c
    RLOC (row location) represents interior and exterior row categories.            with that reported by (Wagner-
  d
     STLLOC represents END and NOTEND of stall section categories.                  Storch et al., 2003) (45.4%) and
  e                                                                                 that reported by (Sonck et al., 1999)
     Side represents north or south pen in the barn.
  f
    Independent variables were considered continuous in model analyses.             (45.95%). The four RMAT in this
  g
     XPOSR is time cows were exposed to bases.                                      study were consistently the least pre-
  h
     WDIST is distance from stall base to waterer.                                  ferred stall bases for percentages ly-
  i
    TEMP is temperature in degrees Fahrenheit in four sections of the barn and      ing and occupied, which was simi-
  outside northern exposure of barn.                                                lar to findings reported by Sonck et
                                                                                    al. (1999) and Wagner-Storch et al.
                                                                                    (2003).
                                                                                       Table 6 shows the percentage of
base types without a cover. Mattress       the barn (101% vs 66%). The              stalls with cows lying for each side
interiors composed mainly of rub-          RMATR1 and RMATR2 were allowed           by month of year. The XPOSR term
ber were referred to as RMATR              to remain on both sides of the barn      influenced lying percentages. Lying
types, and interiors composed              for this experiment because they         percentage varies month to month
mainly of foam where referred to as        had high cow preference values           for stall bases. The stall base ×
FMATR. Designations RMATR1,                from previous research work              XPOSR interaction is significant
RMAT2, and RMAT3 are examples              (Wagner-Storch et al., 2003). The        (P<0.0001), indicating stall use for
of overlap in definitions, as               FMATR1 was allowed to remain on          particular stall bases was different
RMATR1 was rubber filled with a             the 101% SD side, and the waterbed       for different levels of XPOSR. The
foam interior layer and RMAT2 and          was allowed to remain on the 66%         FMATR4 on the 66% SD side of the
RMAT3 were rubber mats with foam           SD side because they had intermedi-      barn exhibited the most variation by
interiors.                                 ate cow preference results pre-          month (41.6% in July to 17.8% in
  Values for lying time percentage         viously. It was felt that retaining      December). The RMATR1 and
and percentage of time stall was oc-       some stall base types from previous      RMATR2, the only stall bases repre-
cupied in Table 5 were chosen to           work would allow comparisons to          sented on both sides of the barn,
characterize cow preference for dif-       be made between the two exper-           had the most variation by month
ferent stall base types. The average       iments.                                  (51.4% in July to 72.9% in Decem-
lying and occupied percentage val-           The RMATR1 ranked highest on           ber and 41.7% in June to 62.4% in
ues were higher on the south side          both sides of the barn for both ly-      September, respectively) on the
than on the north side because of          ing percentage and occupied per-         101% SD side and the least variation
the higher SD of the two sides of          centage. On the 101% SD side,            by month (45.7% in August to
                                   Stall Usage Differences of Thirteen Different Stall Base Types                              477



  TABLE 5. Stall status percentagesa by baseb for each barn sidec (13 stall base types).
  Item                 RMATR1      RMATR2      FMATR1       FMATR2     RMATR3      RMAT2      RMAT3       RMAT4       Xd     SEe
  101% SD (South)f
   Empty                12.1yz      19.4x        18.4x         9.2z      13.7y       26.6w      34.1v       33.1v    21.6   0.014
   Half-in, half-out     6.1v        6.2v         4.6v         5.5v       6.2v        5.6c       6.1v        6.7v     5.8   0.008
   Lying                59.7v       52.9w        52.1w        62.4v      58.9v       50.5w      43.1x       41.9x    52.2   0.016
   Standing             22.2vw      21.6vw       24.6v        22.5vw     19.2wx     17.1xy      14.9y       16.5xy   19.5   0.012
   Occupiedg            87.9vw      80.6x        81.3x        90.5v      84.3wx      73.2y      64.1z       65.1z    77.5   0.014
   Unsure                0.0w        0.0w         0.3w         0.3w       1.9v        0.2w       1.9v        1.9v     0.9   0.003
   Total numberh       736         552         1288         1288       1288        1472       1288        1288

                       RMATR1       RMATR2        CMATR        FMATR3       FMATR4        RMAT5          WATR         X      SE

  66% SD (North)i
   Empty                 40.0z        71.0w        75.5v         38.0z       50.0y          76.7v         59.1x      58.5   0.015
   Half-in, half-out      2.9wxy       2.1y         2.5xy         4.0vw       3.4vwx         2.5xy         4.5v       3.2   0.005
   Lying                 46.8v        20.3x        18.3xy        48.9v       35.3w          15.7y         31.8w      31.1   0.014
   Standing              10.3vw        6.4x         3.8y          8.9w       11.2v           5.1xy         4.0y       7.1   0.007
   Occupied              60.1v        28.7y        24.5yz        61.7v       50.0w          23.3z         40.2x      41.4   0.015
   Unsure                 0.0w         0.3vw        0.0w          0.2w        0.1w           0.1w          0.07v      0.2   0.001
   Total number        1134          945         1512          1341        1687           1323          1512

  a
    Percentage = number of stall-day status observations/total number of stall-day observations by stall base by side.
  b
    CMATR = Cow Comfort Corkmat (Amorim Industrial Solutions, Trevor, WI), FMATR1 = Comfy Cow Mattress with Cow-
  Flex Top Cover (Sikkema’s Equipment, Byron Center, MI), FMATR2 = Foamat™ Bedding System (Foxworthy Supply, Kent
  City, MI), FMATR3 = DeLaval Cow Mattress M100 (DeLaval, Inc., Kansas City, MO), FMATR4 = Comfy Cow Mattress with
  Super-Mat Top Cover (Sikkema’s Equipment), RMATR1= Pasture Mat with foam (Promat Ltd., Seaforth, Ontario, Canada),
  RMATR2 = Pasture Mat with felt (Promat Ltd.), RMATR3 = Ulti-Mat All Rubber Cow Mattress (Zartmann Farms, Ephrata,
  PA), RMAT2 = Supreme Comfort Pad (Humane Manufacturing, Baraboo, WI), RMAT3 = Kraiburg Softbed II (Agromatic,
  Fond du Lac, WI), RMAT4 = Heavy Cush (J & D Manufacturing, Eau Claire, WI), RMAT5 = Simplex Dairy Pad (Berg
  Equipment Corporation, Marshfield, WI), and WATR = Cow Waterbed (Relative Products, LLC, Reedsburg, WI).
  c
    Data collected from June 19, 2002 to December 17, 2002.
  d
    Mean percentages are total stall-day status observations/total number stall-day observations by side.
  e
    Pooled standard error for each status by side.
  f
    South side of the freestall barn with 101% stocking density.
  g
    Occupied is defined as lying + standing + half-in, half-out.
  h
    Total number of stall-day observations by stall base by side.
  i
    North side of the freestall barn with low stocking density.
  v,w,x,y,z
            Percentages within rows with different letters are significantly different (P<0.05).



53.9% in October and 18.2% in June             The FMATR2 and RMATR1 were                 month. The RMATR1 was a pre-
to 12% in December, respectively)            the most preferred bases in the              ferred base for 5 of 6 mo. Preference
on the 66% SD side of the barn. The          101% SD pen for each month. The              for the waterbed was intermediate
highest lying percentages for each           RMATR3 was among the most pre-               and remained relatively constant
mattress type on the 101% SD side            ferred bases for 5 of 6 mo; FMATR1           over the course of this study. Prefer-
were observed the last 4 mo of the           and RMATR2 were among the most               ence for CMATR, FMATR4,
study. With the exception of                 preferred bases 4 of 6 mo. There             RMATR2, and RMAT5 declined over
RMATR1 and waterbed on the 66%               were only four incidences when a             time. This was likely more pro-
SD side, all stall bases had their high-     RMAT was among the most pre-                 nounced because of declining SD
est lying percentages in the first 3          ferred bases. Stocking density de-           (72% in June to 42% in December).
mo of the study and the lowest ly-           creased slightly over the course of            Table 7 shows the percentage of
ing percentages in the last 2 mo.            the study with a high of 105% in             stalls occupied for each stall base by
This result was due to the SD den-           July and a low of 92% in October.            month for each side. Trends were
sity in the 66% SD pen steadily de-            On the 66% SD side, the FMATR3             similar to those for lying percent-
clining over the course of the study.        was the most preferred base each             ages. Rubber mat base stalls consis-
478                                                         Fulwider and Palmer




  TABLE 6. Percentagea of stalls with cows lying for each baseb and mean number of cows by month for
  each sidec.
  Item                  Cows (no.)d RMATR1 RMATR2 FMATR1 FMATR2 RMATR3 RMAT2 RMAT3 RMAT4                                        Xe    SEf
  101% SD (South)g
   June                 51.5    ±   0.4     54.2w     41.7wx       58.3w    52.4w    56.0w     47.9wx      42.9wx     34.5x    48.7   0.06
   July                 52.7    ±   0.1    51.4wx     46.7wxy     51.4wx    56.7w    55.5w     43.9xy      43.3xy     38.0y    48.2   0.04
   August               52.6    ±   0.1    57.6wxy    55.6xy       49.8y    65.8w    61.5wx    54.6xy       36.4z     39.4z    52.1   0.03
   September            50.9    ±   0.5     61.3w      62.4w      55.3wx    61.8w    57.1w      56.5w      45.2xy     41.0y    54.1   0.04
   October              46.0    ±   0.2     65.3w     51.9xy       48.4y    67.5w    60.7wx     46.2y      44.4y      46.8y    53.2   0.03
   November             49.3    ±   0.4    60.0wxy     52wxy      51.4xy    62.3wx   65.1w     54.5wxy     47.4y      48.6xy   55.1   0.05
   December             49.1    ±   0.1     72.9w     55.6xyz     58.3wxy   66.7wx   51.2xyz   51.0xyz     42.9yz     41.7z    53.8   0.05

                        Cows (no.)d         RMATR1      RMATR2       CMATR     FMATR3    FMATR4      RMAT5          WATR        X     SE

      66% SD (North)b
       June              38.0   ±    0.3     45.5w        18.2y      23.9wxy    40.8wx    34.0wxy        19.5y      30.7wxy    30.5   0.06
       July              36.9   ±    0.1     41.4w        19.4x       20.7x     48.0w      41.6w         21.2x       23.2x     31.2   0.03
       August            37.2   ±    0.2     45.7x        28.4y       21.8yz    55.8w      40.5x         13.4z       29.4y     33.5   0.03
       September         34.2   ±    0.4     45.8wx       21.9z       18.4z     50.2w     40.2xy         16.5z       33.6y     32.7   0.03
       October           30.1   ±    0.2     53.9w        24.6y       21.5yz    50.4w      35.6x         16.1z       40.4x     34.5   0.03
       November          26.4   ±    0.6     49.4w         9.2z       10.6z     47.3w      25.2y          9.9z       33.2x     26.4   0.03
       December          22.1   ±    0.1     42.2w        12.0z        6.7z     38.1wx     17.8yz        11.4z      28.3xy     22.0   0.04

  a
    Percentage = number of stall-day lying observations/total number of stall-day observations by stall base for each month by
  side.
  b
    CMATR = Cow Comfort Corkmat (Amorim Industrial Solutions, Trevor, WI), FMATR1 = Comfy Cow Mattress with Cow-
  Flex Top Cover (Sikkema’s Equipment, Byron Center, MI), FMATR2 = Foamat™ Bedding System (Foxworthy Supply, Kent
  City, MI), FMATR3 = DeLaval Cow Mattress M100 (DeLaval, Inc., Kansas City, MO), FMATR4 = Comfy Cow Mattress with
  Super-Mat Top Cover (Sikkema’s Equipment), RMATR1= Pasture Mat with foam (Promat Ltd., Seaforth, Ontario, Canada),
  RMATR2 = Pasture Mat with felt (Promat Ltd.), RMATR3 = Ulti-Mat All Rubber Cow Mattress (Zartmann Farms, Ephrata,
  PA), RMAT2 = Supreme Comfort Pad (Humane Manufacturing, Baraboo, WI), RMAT3 = Kraiburg Softbed II (Agromatic,
  Fond du Lac, WI), RMAT4 = Heavy Cush (J & D Manufacturing, Eau Claire, WI), RMAT5 = Simplex Dairy Pad (Berg
  Equipment. Corporation, Marshfield, WI), and WATR = Cow Waterbed (Relative Products, LLC, Reedsburg, WI).
  c
    Data collected from June 19, 2002 to December 17, 2002.
  d
    Mean (±SE) number of cows during each month by side.
  e
    Total stall-day lying observations/total number stall-day observations for each month by side.
  f
    Pooled standard error for each month by side.
  g
    South side of the barn with 100% stocking density.
  h
    North side of the freestall barn with low stocking density.
  w,x,y,z
          Percentages within rows with different letters are significantly different (P<0.05).



tently had the lowest rankings for               intermediate in preference, but was           have higher lying and occupied per-
occupied percentages for month to                equally preferred with FMATR3 in              centages on the EXTR side at 0400
month. The FMATR2 and RMATR1                     June and December. The CMATR                  h and most of the time at 2000 h.
were consistently the highest for                and RMAT5 consistently ranked the             Occupancy for the INTR is only
percentage occupied in the 101%                  lowest for occupied percentages.              higher for the occupied percentage
SD pen on a month-to-month basis.                  Table 8 and Figure 2 show the ef-           at the 1400-h observation on the
Preference for FMATR1, RMATR2,                   fect of row location, INTR vs EXTR,           north side of the barn. These results
and RMATR3 started out high and                  on stall usage. Figure 2 shows that,          differ from (Wagner-Storch et al.,
declined over time. Results were sim-            overall, cows prefer the external             2003), as cows in that study pre-
ilar in the 66% SD pen, with                     row of stalls for both lying and oc-          ferred the EXTR row for lying in
FMATR3 having the highest percent-               cupied percentages for both sides of          both pens at all times, with the ex-
age occupied every month. The                    the barn. Table 8 shows that this             ception of the 14000-h observation
RMATR1 was the next most pre-                    preference is different for different         in the 101% SD pen, when there
ferred. The waterbed tended to be                times of the day. It shows that cows          was no preference difference. Cows
                                       Stall Usage Differences of Thirteen Different Stall Base Types                                   479



  TABLE 7. Percentagea of stalls occupiedb for each basec and mean number of cows by month for each
  sided.
  Item                Cows (no.) FMATR1 FMATR2 RMATR1 RMATR2 RMATR3 RMAT2 RMAT3 RMAT4                                           Xe    SEf
  101% SD (South)g
   June               51.5   ±   0.4     84.5w       79.8wx    85.4w      80.6wx       83.3w    75.0wxy   65.5xy     60.7y     76.0   0.06
   July               52.7   ±   0.1     81.2w       86.5w     79.3w      78.0w        85.7w     68.6x    67.4x      64.1x     75.9   0.03
   August             52.6   ±   0.1    82.3xy       93.9w     91.7w      91.9w        87.9wx    80.7y    62.8z      61.9z     80.2   0.02
   September          50.9   ±   0.5    86.2xy       94.0w     96.0w      86.0xy       84.8xy    80.7y    65.9z      68.7z     81.7   0.03
   October            46.0   ±   0.2     74.2y       90.5w     89.6w      74.1xy       80.6xy    66.7z    61.1z      65.5z     74.3   0.03
   November           49.3   ±   0.4    81.1wxy      92.0w     83.0w      74.7xyz      87.4wx    72.0yz   61.1z      68.6yz    77.3   0.04
   December           49.1   ±   0.1    84.5wx       90.5w     89.6w      75.0xy       75.0xy    66.7y    66.7y      63.1y     75.5   0.05

                      Cows (no.)         CMATR       FMATR3     FMATR4       RMATR1       RMATR2        RMAT5      WATR        X      SE

  66% SD (North)b
   June               38.0   ±   0.3     31.8x        54.0w      49.5wx       59.1w         29.1z       31.2x      37.5wx     41.9    0.07
   July               36.9   ±   0.1     25.4y        62.6w      59.4w        55.7w         30.3x       28.6y      33.6y      42.6    0.03
   August             37.2   ±   0.2     31.5yz       72.4w      65.2w        64.0w         36.8x       25.4z      39.5y      48.1    0.04
   September          34.2   ±   0.4     25.8z        63.0w      54.9x         53.7x        27.5x       22.8z      40.6y      41.8    0.03
   October            30.1   ±   0.2     26.9y        62.5w      46.8x        67.1w         30.3y       22.7y      48.7x      43.5    0.03
   November           26.4   ±   0.6     14.9z        58.8w      32.9y        62.2w         20.0y       15.9z      40.9xy     34.8    0.03
   December           22.1   ±   0.1     10.8y        43.8w      23.7x        54.4w         21.3yz      16.2xy     35.0w      28.7    0.04

  a
    Percentage = number of stall-day lying observations/total number of stall-day observations by stall base for each month by
  side.
  b
    Occupied is defined as lying + standing + half-in, half-out.
  c
    CMATR = Cow Comfort Corkmat (Amorim Industrial Solutions, Trevor, WI), FMATR1= Comfy Cow Mattress with Cow-Flex
  Top Cover (Sikkema’s Equipment, Byron Center, MI), FMATR2 = Foamat™ Bedding System (Foxworthy Supply, Kent City,
  MI), FMATR3 = DeLaval Cow Mattress M100 (DeLaval, Inc., Kansas City, MO), FMATR4 = Comfy Cow Mattress with
  Super-Mat Top Cover (Sikkema’s Equipment), RMATR1= Pasture Mat with foam (Promat Ltd., Seaforth, Ontario, Canada),
  RMATR2 = Pasture Mat with felt (Promat Ltd.), RMATR3 = Ulti-Mat All Rubber Cow Mattress (Zartmann Farms, Ephrata,
  PA), RMAT2 = Supreme Comfort Pad (Humane Manufacturing, Baraboo, WI), RMAT3 = Kraiburg Softbed II (Agromatic,
  Fond du Lac, WI), RMAT4 = Heavy Cush (J & D Manufacturing, Eau Claire, WI), RMAT5 = Simplex Dairy Pad (Berg
  Equipment. Corporation, Marshfield, WI), and WATR = Cow Waterbed (Relative Products, LLC, Reedsburg, WI).
  d
    Data collected from June 19, 2002 to December 17, 2002.
  e
    Mean percentage occupied equals total stall-day occupied observations/total number stall-day observations for each month
  by side.
  f
    Pooled standard error for each month by side.
  g
    South side of the barn with 101% stocking density.
  h
    North side of the freestall barn with low stocking density.
  w,x,y,z
          Percentages within rows with different letters are significantly different (P<0.05).




in the Wagner-Storch et al. study                 TEND stall location by side of barn.            Figure 4 shows the percentage of
(2003) preferred to occupy the                    The NOTEND stalls were preferred              stalls with cows lying on the 101%
EXTR row at all times. Table 4                    to END stalls for lying and occupied          SD pen for each stall base type by
shows a strong RLOC × TEMP inter-                 percentages in the 66% SD pen;                temperature (°F) interval. The tem-
action. Combining this information                there was no difference between               perature effect influenced lying and
suggests that cows prefer the out-                END and NOTEND stalls in the                  occupied percentages for each stall
side rows of stalls during the eve-               101% SD pen. This was probably                base. Stall use decreased as tempera-
ning and early morning hours of                   the result of lack of choice because          ture increased in the 101% SD pen.
the day and inside rows of stalls dur-            of the higher SD. The interaction be-         This was supported by the signifi-
ing the hotter times of the day.                  tween stall base and STLLOC sug-              cant (P<0.05) stall base × tempera-
  Figure 3 shows the percentage of                gests that something about END                ture interaction, which indicates
stalls with cows standing, lying, or              stalls affects cow preference for spe-        cow preference for different stall
stalls occupied for END and NO-                   cific stall base types.                        base types at different temperature
480                                                          Fulwider and Palmer




   TABLE 8. Percentagea of stalls with cows lying and stalls occupiedb
   for INTR (interior) and EXTR (exterior) row by time of day for each
   sidec.
                                     Lying                           Occupied
   Item                  INTR     EXTR         Xd      SEe   INTR    EXTR     Xf    SEg
   101% SD (South)h
    1400 h               47.3x    41.9x    44.4       0.03   73.6x   69.9x   64.0   0.02
    2000 h               44.3x    48.2x    46.4       0.02   62.6y   79.9x   71.9   0.02
    0400 h               54.4y    59.7x    57.3       0.02   77.2y   88.7x   83.4   0.02
    0900 h               58.7x    59.6x    59.2       0.02   77.5y   85.7x   81.9   0.02   Figure 3. Percentage of stalls with cows
   66% SD (North)i                                                                         standing, cows lying, or stalls occupied for
    1400 h               25.3x    22.5x    23.8       0.02   38.8x   29.0y   33.5   0.02   END and NOTEND stall locations by side
    2000 h               24.7y    30.0x    27.6       0.01   33.4y   43.7x   39.3   0.01   of barn. Occupied = lying + standing +
    0400 h               39.5y    47.0x    43.6       0.02   51.2y   57.4x   54.6   0.02   half-in, half-out. Percentages = number of
    0900 h               25.8x    29.3x    27.7       0.02   35.1x   37.1x   36.2   0.02   stall-day preference measure observations/
                                                                                           total number of stall-day observations.
   a
     Percentage = number of stall-day lying (occupied) observations/total number           END = end stall of a stall section and
   stall-day observations by row for each time by side.                                    NOTEND = not end of a stall section.
   b
     Occupied is defined as lying + standing + half-in, half-out.                           SD = stocking density. Percentages within
   c
     Data collected from June 19, 2002 to December 17, 2002.                               preference measure, by side, with different
   d                                                                                       letters (a, b) differ (P<0.05). SE =
     Mean percentage lying equals total stall-day lying observations/total number          preference measures’ pooled SE.
   stall-day observations for each time by side.
   e
     Pooled standard error for lying by time.
   f
     Mean percentage occupied equals total stall-day occupied observations/total
   number stall-day observations for each time by side.
   g
     Pooled standard error for occupied by time.
   h
     South side of the freestall barn with 101% stocking density.
   i
     North side of the freestall barn with 66% stocking density.
   x,y
       Percentages within rows, lying and occupied analyzed separately, with
   different letters are significantly different (P<0.05).



                                                    intervals. Cows preferred to lie on
                                                    RMATR1, FMATR1, and RMATR2, re-
                                                                                           Figure 4. Percentage of stalls with cows
                                                    spectively, for the lowest tempera-    lying on the 101% SD (stocking density)
                                                    ture range. The RMATR1, FMATR2,        side for each stall base type by
                                                    and RMATR3 were preferred at me-       temperature interval. Percentage lying =
                                                    dian temperature ranges. At the        number of stall-day lying observation/total
                                                    highest temperature range,             stall-day observations by stall base.
                                                    FMATR2, RMATR3, and RMATR1,            FMATR1 = Comfy Cow Mattress with
                                                    had the highest lying percentages.     Cow-Flex Top Cover (Sikkema’s
                                                                                           Equipment, Byron Center, MI), FMATR2 =
                                                    Rubber mats consistently had the
                                                                                           Foamat™ Bedding System (Foxworthy
                                                    lowest percentage lying across all     Supply, Kent City, MI), RMATR1 =
                                                    temperatures.                          Pasture Mat with foam (Promat Ltd.,
Figure 2. Percentage of stalls with cows              Figure 5 shows the percentage of     Seaforth, Ontario, Canada), RMATR2 =
standing, cows lying, or stalls occupied for        stalls with cows lying in the 66%      Pasture Mat with felt (Promat Ltd.),
interior and exterior rows by side of barn.         SD pen, with highest lying percent-    RMATR3 = Ulti-Mat All Rubber Cow
Occupied = lying + standing + half-in,              ages at mid-range temperatures. At     Mattress (Zartmann Farms, Ephrata,
half-out. Percentages = number of stall-            the temperature range 1 to 20°F,       PA), RMAT2 = Supreme Comfort Pad
day preference measure observations/total                                                  (Humane Manufacturing, Baraboo, WI),
                                                    the waterbed was most preferred for
number of stall-day observations. SD =                                                     RMAT3 = Kraiburg Softbed II
stocking density. Percentages within                lying at 40%, followed by RMATR1       (Agromatic, Fond du Lac, WI), and
preference measure, by side, with different         at 33.3%. Cows preferred to lie on     RMAT4 = Heavy Cush (J & D
letters (a,b) differ (P<0.05). SE =                 RMATR1, FMATR3, waterbed, and          Manufacturing, Eau Claire, WI). SE =
preference measures’ pooled SE.                     FMATR4, respectively, when temper-     each temperature interval’s pooled SE.
                                     Stall Usage Differences of Thirteen Different Stall Base Types                               481




Figure 5. Percentage of stalls with cows
lying on 66% SD (stocking density) side        Figure 6. Percentage of stalls occupied on     Figure 7. Percentage of stalls occupied on
for each stall base type by temperature        101% SD (stocking density) side for each       66% SD (stocking density) side for each
interval. Percentage lying = number of         stall base type by temperature interval.       stall base type by temperature interval.
stall-day lying observations/total stall-day   Occupied = lying + standing + half-in,         Percentage occupied = number of stall-day
observations by stall base. CMATR = Cow        half-out. Percentage occupied = number of      occupied observations/total stall-day
Comfort Cork Mattress (Amorim                  stall-day occupied observations/total stall-   observations by stall base. CMATR = Cow
Industrial Solutions, Trevor, WI),             day observations by stall base. FMATR1 =       Comfort Cork Mattress (Amorim
FMATR3 = DeLaval Cow Mattress M100             Comfy Cow Mattress with Cow-Flex Top           Industrial Solutions, Trevor, WI),
(DeLaval, Inc., Kansas City, MO),              Cover (Sikkema’s Equipment, Byron              FMATR3 = DeLaval Cow Mattress M100
FMATR4 = Comfy Cow Mattress with               Center, MI), FMATR2 = Foamat™                  (DeLaval, Inc., Kansas City, MO),
Super Mat Top Cover (Sikkema’s                 Bedding System (Foxworthy Supply, Kent         FMATR4 = Comfy Cow Mattress with
Equipment, Byron Center, MI), RMATR1 =         City, MI), RMATR1 = Pasture Mat with           Super Mat Top Cover (Sikkema’s
Pasture Mat with foam (Promat Ltd.,            foam (Promat Ltd., Seaforth, Ontario,          Equipment, Byron Center, MI), RMATR1 =
Seaforth, Ontario, Canada), RMATR2 =           Canada), RMATR2 = Pasture Mat with             Pasture Mat with foam (Promat Ltd.,
Pasture Mat with felt (Promat Ltd.),           felt (Promat Ltd.), RMATR3 = Ulti-Mat          Seaforth, Ontario, Canada), RMATR2 =
RMAT5 = Simplex Dairy Pad (Berg                All Rubber Cow Mattress (Zartmann              Pasture Mat with felt (Promat Ltd.),
Equipment Corporation, Marshfield, WI),         Farms, Ephrata, PA), RMAT2 = Supreme           RMAT5 = Simplex Dairy Pad (Berg
and WATR = Cow Waterbed (Relative              Comfort Pad (Humane Manufacturing,             Equipment Corporation, Marshfield, WI),
Products, LLC, Reedsburg, WI). SE = each       Baraboo, WI), RMAT3 = Kraiburg Softbed         and WATR = Cow Waterbed (Relative
temperature interval’s pooled SE.              II (Agromatic, Fond du Lac, WI), and           Products, LLC, Reedsburg, WI). SE = each
                                               RMAT4 = Heavy Cush (J & D                      temperature interval’s pooled SE.
                                               Manufacturing, Eau Claire, WI). SE =
atures were at mid ranges. The                 each temperature interval’s pooled SE.
                                                                                              difference was likely due to cows’
CMATR, RMATR2, and RMAT5                                                                      preference for particular stall bases.
ranked low across all temperatures.                                                           Cows preferred stalls closer to and
  Figure 6 shows the percentage of             RMATR1, and waterbed had the                   farthest from water in the 66% SD
stalls occupied in the 101% SD pen             highest occupied percentages, re-              pen. This result was explained by
for each stall base type by tempera-           spectively, for mid-range tempera-             the significant WDIST × stall base in-
ture (°F) interval. Results for stall oc-      tures. The FMATR4, FMATR3, and                 teraction.
cupied percentages in the101% SD               RMATR1 had the highest occupied                  Lying and occupied percentages
pen showed FMATR2, RMATR1, and                 percentages, respectively, at the              for RMAT (with the exception of
RMATR3 having the highest                      highest temperatures. The waterbed             RMAT2, which had a foam interior)
(P<0.05) occupied percentages                  had higher occupied percentages                in both pens, and CMATR and
across all temperatures. The                   than CMATR, RMATR2, and RMAT5                  RMATR2 in the 66% SD pen were
FMATR1 and RMATR2 follow with                  at all temperature ranges. Waterbed            below the mean lying and occupied
the next highest lying percentages.            preference during the coldest tem-             percentages. The RMATR1 and
The RMAT2, RMAT3, and RMAT4                    peratures may be enhanced by the               RMATR2 were the only stall base
consistently showed the lowest oc-             ability of water to hold and save              types available in both pens (101%
cupied percentages.                            heat.                                          and 66% SD). The RMATR1 ranked
  Figure 7 shows the percentage of                Stall use percentages by WDIST              highest in previous research efforts
stalls occupied in the 66% SD pen              showed that cows preferred stalls              (Wagner-Storch et al., 2003) and in
for each stall base type by tempera-           the furthest away from water in the            this trial for percentages standing
ture (°F) interval. The RMATR1 and             101% SD pen. This differs from a               and occupied in both pens.
waterbed had the highest occupied              previous study (Wagner-Storch et                 On the 101% SD side, RMATR1,
percentages for the lowest tempera-            al., 2003), where cows preferred               FMATR2, and RMATR3 had the
ture range. The FMATR3, FMATR4,                stalls near and far from water. The            highest lying percentages, whereas
482                                                Fulwider and Palmer



RMATR1 and FMATR2 had the high-          protecting the cow during times of        thank Peter Crump and Yu-mei
est occupied percentages. The            cold weather. Waterbeds were the          Chang for their assistance in the sta-
FMATR2 and RMATR1 ranked                 stall base of choice in the 66% SD        tistical analysis of the data. This re-
higher for lying and occupied per-       pen, during the coldest weather,          search was sponsored in part by
centages than FMATR1 and                 RMATR and the RMAT. Percentages           UDSA/Hatch Project number
RMATR2. Rubber mats RMAT3 and            lying and occupied increased with         WISO4703.
RMAT4 had the lowest lying and oc-       SD, suggesting that SD needs to be
cupied percentages. For the 66% SD       considered when assessing stall us-
pen, FMATR3 and RMATR1 ranked            age. Percentage of cows lying in or
highest for lying and occupied per-      occupying stalls was highest in
centages. The FMATR4 and wa-             early morning before milking and
terbed were intermediate for lying
and occupied percentages, whereas
                                         late morning after feeding for 101%
                                         SD cows. Cows in the 66% SD pen,
                                                                                                   Literature Cited
CMATR, RMATR2, and RMAT5 were            with robot access, have the luxury        Anderson, N. G. 2002. Cozying up to cow
inferior. Competition for stalls was     of choosing their own schedule.           comfort. In Midwest Dairy Herd Health
less in the 66% SD pen, so cows          These cows exhibited their highest        Conf., Middleton, WI. p 11. Dairy Team,
                                                                                   Univ. of Wisconsin Extension, Dept. of
could more easily occupy the stall       lying percentages at early morning,       Dairy Sci., College of Ag. and Life Sci., Univ.
of their choice and avoid less desir-    with equal percentages for early          of Wisconsin-Madison.
able stall types. The higher occu-       morning and evening. Highest stall        Gaworski, M. A., C. B. Tucker, D. M. Weary,
pied percentages for RMAT in the         occupied percentages were shown           and M. L. Swift. 2003. Effects of stall design
101% SD pen was likely due to the        early morning and evening.                on dairy cattle behavior. In Proc. 5th Int.
                                                                                   Dairy Housing Conf., Fort Worth, TX. p
higher SD in this pen.                                                             139, Am Soc. Ag. Eng., St. Joseph, MI.
   Cows preferred the EXTR row of
stalls during the evening and early      Implications                              House, H. K., J. Rodenburg, and B. R. Lang.
                                                                                   2003. The effect of neck rail and mounting
morning hours and tended to like                                                   rail position on cow behavior. In Proc. 5th
                                            Stall base type was shown to af-       Int. Dairy Housing Conf., Fort Worth, TX. p
the INTR row during the daytime
                                         fect cow preference in this study. Ly-    147, Am Soc. Ag. Eng., St. Joseph, MI.
hours. This result might have been
                                         ing and occupied percentages were         McFarland, D. F. 2003. Freestall design: Cow
due to a cow’s preference for the
                                         higher overall for FMATR and              recommended refinementsIn Proc. 5th Int.
sun or breeze, relative to tempera-                                                Dairy Housing Conf., Fort Worth, TX. p
                                         RMATR. Waterbeds, available on
ture. The NOTEND stalls were pre-                                                  131, Am Soc. Ag. Eng., St. Joseph, MI.
                                         the 66% SD side, were the base of
ferred to END stalls for lying and oc-                                             Overton, M. W., W. M. Sischo, G. D. Tem-
                                         choice during the coldest weather.
cupied percentages in the 66% SD                                                   ple, and D. A. Moore. 2002. Using time-
                                         A number of stall base types were in-     lapse video photography to assess dairy cat-
pen, and there was no difference be-
                                         ferior, regardless of temperature or      tle lying behavior in a free-stall barn. J.
tween END and NOTEND stalls in                                                     Dairy Sci. 85:2407.
                                         XPOSR. With respect to this study,
the 101% SD pen. This was proba-
                                         the following guidelines are sug-         Rodenburg, J., and H. K. House. 2000. The
bly the result of lack of choice be-                                               impact of freestall base and bedding on cow
                                         gested when assessing freestall com-
cause of the higher SD. The interac-                                               comfort. In Dairy Housing and Equipment
                                         fort. 1) Observe stall occupancy in       Systems Managing and Planning for Profit-
tion between stall base and STLLOC
                                         the morning before feeding and            ability, p 214. NRAES-129.
suggests that something about END
                                         milking activities begin. 2) Count        Rushen, J., and A. M. de Passille. 1999. Envi-
stalls affects cow preference for spe-
                                         stalls that have a cow lying or stand-    ronmental design for healthier and more
cific stall base types. The WDIST                                                   profitable cows. [Online.] Available: http://
                                         ing in them. 3) If the barn is
was a significant predictor of stall                                                www.afns.ualberta.ca/Hosted/WCDS/
                                         stocked at 100%, a 55 to 65% lying        Proceedings/1999/chap28.htm. Accessed Sep
use, but the analysis yielded con-
                                         percentage and 80 to 85% occupied         9, 2003.
flicting results regarding cow prefer-
                                         percentage is very good. 4) Observa-
ence for stall WDIST. This confus-                                                 Sonck, B., J. Daelemans, J. Langenakens.
                                         tions taken later in the day may          1999. Preference test for free stall surface
ing result was probably due to ran-
                                         drop by 15 to 20%.                        material for dairy cows. In Emerging Tech-
dom location of cow-preferred stall                                                nologies for the 21st Century. p 1. Am. Soc.
bases.                                                                             Ag. Eng., St. Joseph, MI.
   Cow preference for a particular       Acknowledgments                           Stokes, M. E., C. S. Davis, and G. G. Koch.
base appears to be affected by tem-                                                1995. Using the GENMOD procedure. In
perature. Foam-filled mattresses and                                                Categorical Data Analysis Using the SAS Sys-
                                           The authors thank the employees
                                                                                   tem. p 208. SAS Institute, Inc., Cary, NC.
RMATR were the preferred bases           of the Blaine Dairy, Arlington re-
across all temperature intervals in      search facility for their assistance in   Wagner-Storch, A. M., R. W. Palmer, and D.
                                                                                   W. Kammel. 2003. Factors affecting stall use
the 101% SD pen. Mattresses pro-         switching tapes and monitoring            for different freestall bases. J. Dairy Sci.
vide cushion, as well as insulation,     camera function. The authors also         2003 86:2253.

								
To top