THE IMPACT OF STALL AND FEED RAIL STOCKING RATE ON COW WELFARE AND
BEHAVIOR IN AUTOMATIC MILKING SYSTEMS
By
Ali Adil Mutlak Witaifi

A THESIS
Submitted to
Michigan State University
in partial fulfillment of the requirements
for the degree of
Animal Science – Master of Science
2014

ABSTRACT
THE IMPACT OF STALL AND FEED RAIL STOCKING RATE ON COW WELFARE AND
BEHAVIOR IN AUTOMATIC MILKING SYSTEMS
By
Ali Adil Mutlak Witaifi
Stocking rate at stalls and the feed rail are critical determinants of dairy cows’ comfort and
welfare. Overstocking in parlor-milked dairies leads to increased displacement or reduced time
lying or feeding. However no studies have examined impacts of stocking rates for cows milked
with an automatic milking system (AMS). Therefore, the aim of this thesis was to examine the
impact of stocking rate in the stalls and at the feed rail on body condition score, time budgets,
lying behavior, lameness (gait scores), and cleanliness of lactating dairy cows through two
projects. The second chapter describes a survey of 39 AMS dairy farms, which were assessed
with respect to the impact of stocking rate at the feed rail, stalls, and AMS on cow behavior and
welfare. Stocking rates were generally not related to cow body condition, measures of lameness
or cleanliness or lying behavior. However, stall stocking rate was related to leg cleanliness with
cleaner cows found at lower stocking rates (P < 0.01). The third chapter examines the impact of
experimentally reducing the number of stalls or feed rail headlocks per cow at an AMS dairy.
Increasing stocking rates to 1:1.2 or 1:1.5 cows per stall or headlock did not affect stall use,
feeding behavior, activity, rumination or displacement with the exception that more displacement
occurred following fresh feed delivery at higher stocking rates (P = 0.01). Cows in AMS dairies
were milk asynchronously rather than in groups as in parlor-based systems, which may result in
less synchronous feeding and lying behavior. In contrast to findings from parlor-milked dairies,
our results show a less clear link between stocking rates in AMS dairies and cow behavior and
welfare.

Copyright by
ALI ADIL MUTLAK WITAIFI
2014

ACKNOWLEDGEMENTS

I would like to thank the following people for supporting my education and passion for animals:
Adil Witaifi & Family
David Mikho & Family
Simon Mikho & Family
My fiancé, Lynae Somers
Salam Witaifi & Family
My advisor, Dr. Janice Siegford
Dr. Janice Swanson, Dr. David Beede, Dr. Daniel Grooms & Dr. Steve Bursian
My sponsor, The Higher Committee for Education Development in Iraq
MSU, Agricultural College, Department of Animal Science
My loving pets, Lennon, Blackie & Zeus
Graduate Students:
Courtney Daigle, Marisa Erasmus, Dana Campbell, Kaitlyn Wurtz, Carly O’Malley, Maher
Alsahlany, Joe Leszcz, and Prafulla Regmi

iv

TABLE OF CONTENTS

LIST OF TABLES ...................................................................................................................... vii
LIST OF FIGURES ................................................................................................................... viii
KEY TO SYMBOLS OR ABBREVIATIONS ......................................................................... xii
INTRODUCTION..........................................................................................................................1
CHAPTER 1
LITERATURE REVIEW……………………………………………………………………….4
INTRODUCTION……………………………………………………………………………..4
Interactive Impacts of Stocking Rate……………….…………………………………….…....5
Stocking Rates and Lying Time...................................................................................................6
Feed Rail Stocking Rates……………………………………………………………………….7
Stocking Rate and Automatic Milking Systems………………………………………………..8
REFERENCES………………………………………………………………………………..11
CHAPTER 2
THE EFFECT OF STOCKING RATES ON LYING BEHAVIOR AND WELFARE OF
COWS MILKED WITH AUTOMATIC MILKING SYSTEMS……………………………16
ABSTRACT…………………………………………………………………………………..18
INTRODUCTION…………………………………………………………………………….20
Stall Stocking Rate……………………………………………………………………….20
Feed Space Stocking Rate………………………………………………………………..22
Stocking Rate in Automatic Milking Systems…………………………………………...22
Objectives………………………………………………………………………………..21
Hypotheses………………………………………………………………………….……21
MATERIALS AND METHODS……………………………………………………………...22
The Farms………………………………………………………………………………..23
Animals………………………………………………………………………………….23
Barn Measurements……………………………………………………………………...25
Analysis………………………………………………………………………………….25
RESULTS………………………………………………………………………...…………..26
DISCUSSION………………………………………………………………………………...29
CONCLUSIONS……………………………………………………………………………...32
ACKNOWLEDGEMENTS…………………………………………………………………..33
APPENDIX……………………………………………………………………………...……34
REFERENCES…………………………….…………………………………………...……..55

v

CHAPTER 3
IMPACT OF STALL AND FEED RAIL STOCKING RATE ON COW BEHAVIOR AND
PRODUCTION IN AUTOMATIC MILKING SYSTEMS……………………………….…60
ABSTRACT…………………………………………………..………………………………61
INTRODUCTION…………………………………………………………………………….64
MATERIALS AND METHODS……………………………………………………………..67
Animals and Facility……………………………………………………………………….67
Treatments…………………………………………………………………………………68
Data Recording…………………………………………………………………………….69
Analysis…………………………………………………………………………………….71
RESULTS……………………………………………………………………………………..72
Stall Stocking Rate……………………………………………………………………….72
Feed Rail Stocking Rate...…………………………………………………………….….72
DISCUSSION...………………………………………………………………………………73
Stall Stocking Rate……………………………………………………………………….73
Feed Rail Stocking Rate…………………………………………………………….……74
CONCLUSIONS……………………………………………………………………………...75
ACKNOWLEDGEMENTS…………………………………………………………………..76
APPENDIX……………………………………………………………………….……..……77
REFERENCES………………………………………………………………………………..95

CHAPTER 4
SUMMARY AND CONCLUSIONS…………………………………...…………………….101
SUMMARY…...……………………………………………………...…………….…………..101
Survey Stocking Rates at AMS Dairies………………………………………………………..101
Experimental Overstocking at AMS Dairies…………………………………………………...104
Stall Stocking Rate……………………………………………………………………………..105
Feed Rail Stocking Rate………………………………………………………………………..106
CONCLUSIONS...……………………………………………………………………………..107

vi

LIST OF TABLES

Table 2.1: Characteristics of AMS farms in the survey .................................................................35
Table 2.2: Descriptive data from the 39 farms surveyed showing stall stocking rate (cow/stall),
feed rail stocking rate (cow/60 cm by 24 headlocks, 19 post-rail, and 1 swing rail), and
stocking rate at the automatic milk system (cow/AMS). ...............................................36

vii

LIST OF FIGURES

Figure 2.1: Correlational analysis showing no relationship between stall stocking rate
(SRS) and feed rail stocking rate (SRF; P = 0.25)…..……………………………….37
Figure 2.2: Correlational analysis showing no relationship between AMS stocking rate
(SRA) and stall stocking rate (SRS; P = 0.08)……………………………………….38
Figure 2.3: Correlational analysis showing no relationship between AMS stocking rate
(SRA) and feed rail stocking rate (SRF; P = 0.11)…………………………………...39
Figure 2.4: Descriptive data showing average total lying time for each farm over 24 h (created
using an average from 4 days of data per cow then averaging all 40 focal cows on the
farm). Data are presented according to number of AMS in each pen: orange = 3
AMS/pen, gray = 2 AMS/pen, and blue = 1 AMS/pen................................................40
Figure 2.5: Descriptive data showing the average lying bout duration of each farm (created using
an average from 4 days of data per cow then averaging all 40 focal cows on the farm).
Data are presented according to number of AMS in each pen: orange = 3 AMS/pen,
gray = 2 AMS/pen, and blue = 1 AMS/pen .................................................................41
Figure 2.6: Descriptive data showing the average number of lying bouts for each farm
(created using an average from 4 days of data per cow, then averaging all 40 focal
cows on the farm). Data are presented according to the number of AMS in each pen:
orange = 3 AMS/pen, gray = 2 AMS/pen, and blue = 1 AMS/pen………………….42
Figure 2.7: Descriptive data showing presence of a limp (which was scored as either absent (0)
or present (1)) for 40 focal cows at each farm. Data are presented according to
number of AMS in each pen: orange = 3 AMS/pen, gray = 2 AMS/pen, and blue = 1
AMS/pen. .....................................................................................................................43
Figure 2.8: Descriptive data showing presence of a limp (which was scored as either
absent (0) or present (1)) for 40 focal cows at each farm. Data are presented according
to number of AMS in each pen: orange = 3 AMS/pen, gray = 2 AMS/pen, and blue =
1 AMS/pen ....................................................................................................................44
Figure 2.9: Descriptive data showing step asymmetry (which was scored as either absent (0) or
present (1)) associated with lameness for 40 focal cows at each farm. Data are
presented according to number of AMS in each pen: orange = 3 AMS/pen, gray = 2
AMS/pen, and blue = 1 AMS/pen ...............................................................................45
Figure 2.10: Descriptive data showing body condition (which was scored on 1-5 scale in 0.25
increments with 1 = a very thin cow and 5 = an obese cow) of 40 focal cows at each
farm. Data are presented according to number of AMS in each pen: orange = 3
AMS/pen, silver = 2 AMS/pen, and blue = 1 AMS/pen ..............................................46
viii

Figure 2.11: Correlational analysis showing no relationship between feed rail stocking rate and
frequency of feed delivery (P = 0.57). .........................................................................47
Figure 2.12: Correlational analysis showing no relationship between frequency of feed push up
and feed rail stocking rate (P = 0.47)...........................................................................48
Figure 2.13: Correlational analysis showing no relationship between body conditions score and
frequency of feed delivery (P = 0.16). .........................................................................49
Figure 2.14: Correlational analysis showing no relationship between BCS and frequency of feed
push up (P = 0.9)…………………………….…………………………………….…50
Figure 2.15: Correlational analysis showing no a relationship between push up and feed delivery
(P = 0.14). ....................................................................................................................51
Figure 2.16: Descriptive data showing leg cleanliness (which was scored on a 0-3 scale, with 0
representing clean and 3 representing very dirty) for 40 focal cows on each farm.
Data are presented according to the number of AMS in each pen: orange = 3
AMS/pen, gray = 2 AMS/pen, and blue 1 = AMS/pen................................................52
Figure 2.17: Descriptive data showing udder cleanliness (which was scored on a 0-3 scale, with 0
representing clean and 3 representing very dirty) for 40 focal cows on each farm.
Data are presented according to the number of AMS in each pen: orange = 3
AMS/pen, gray = 2 AMS/pen, and blue 1 = AMS/pen................................................53
Figure 2.18: Descriptive data showing flank cleanliness (which was scored on a 0-3 scale, with 0
representing clean and 3 representing very dirty) for 40 focal cows on each farm.
Data are presented according to the number of AMS in each pen: orange = 3
AMS/pen, gray = 2 AMS/pen, and blue = 1 AMS/pen................................................54
Figure 3.1: Average counts of stall displacement (± SEM) by time period during the final 2 d of
each treatment for the three stocking rate treatments ..................................................78
Figure 3.2: Average number of cows lying down in stalls (± SEM) presented by time period
during the final 2 d of each treatment for the three stocking rate treatments ..............79
Figure 3.3: Average number of cows standing in stalls (± SEM) presented by time period during
the final 2 d of each treatment for the three stocking rate treatments ..........................80
Figure 3.4: Average number of cows perching in stalls (± SEM) presented by time period during
the final 2 d of each treatment for the three stocking rate treatments ...……..81
Figure 3.5: Average numbers of cows standing in alley (± SEM) presented by time period during
the final 2 d of each treatment for the three stocking rate treatments………..82

ix

Figure 3.6: Average number of cows walking in the alley (± SEM) presented by time period
during the final 2 d of each treatment for the three stocking rate treatments ..............83
Figure 3.7: Cow activity in response to stocking rate during the final 2 d of each stocking rate
treatment ......................................................................................................................84
Figure 3.8: Descriptive data showing the time cows spent ruminating during the final 2 d of
stocking rate treatments at stalls and feed rail .............................................................85
Figure 3.9: Average number of displacements (± SEM) was occurring during feed delivery and
push up at the feed rail during the final 2 d of each stocking rate treatment ...............86
Figure 3.10A: Descriptive data showing number of cows eating at the feed rail for each treatment
after delivery of fresh feed. Numbers are presented as total number of cows eating
along the entire length of the feed rail (averaged between the 2 pens)........................87
Figure 3.10B: Descriptive data showing number of cow with heads in headlocks at the feed rail
for each treatment after delivery of fresh feed. Numbers are presented as total number
of cows eating along the entire length of the feed rail (averaged between the 2
pens..............................................................................................................................88
Figure 3.10C: Descriptive data showing number of cow with heads out of headlocks but standing
at the feed rail for each treatment after delivery of fresh feed. Numbers are presented
as total number of cows eating along the entire length of the feed rail (averaged
between the 2 pens) ......................................................................................................89
Figure 3.10D: Descriptive data showing number of cow standing behind the feed rail for
each treatment after delivery of fresh feed. Numbers are presented as total number
of cows eating along the entire length of the feed rail (averaged between the 2
pens) ..........................................................................................................................90
Figure 3.11A: Descriptive data showing number of cows eating at the feed rail for each treatment
after push up of feed. Numbers are presented as total number of cows eating along the
entire length of the feed rail (averaged between the 2 pens) .......................................91
Figure 3.11B: Descriptive data showing number of cow with heads in headlocks at the feed rail
for each treatment after push up of feed. Numbers are presented as total number of
cows with head in headlocks but not eating along the entire length of the feed rail
(averaged between the 2 pens) .....................................................................................92
Figure 3.11C: Descriptive data showing number of cow with heads out of headlocks but
standing at the feed rail for each treatment after push up of feed. Numbers are
presented as total number of cows with heads out of headlocks but standing at the
feed rail along the entire feed rail (averaged between the 2 pens)...............................93
Figure 3.11D: Descriptive data showing number of cow standing behind the feed rail for
x

each treatment after push up of feed. Numbers are presented as total number of cows
standing behind the feed rail along the entire feed rail (averaged between the 2
pens)……………………………………………………………………...…………..94

xi

KEY TO SYMBOLS OR ABBREVIATIONS

AMS – Automatic milking system
ASY – Step asymmetry
ATV – Activity unit
AVLBD – Average lying bout duration
BCS – Body condition score
FC – Flank cleanliness
FED – Feed delivery
HB – Head bob
LC – Leg cleanliness
LIMP – Presence of a limp
NLB – Number of lying bouts
PU – Push up
RU - Rumination
SRA – Automatic milking system stocking rate
SRF – Feed rail stocking rate
SRS – Stall stocking rate
TL – Total lying time
UC – Udder cleanliness

xii

INTRODUCTION
Stocking rate at the stalls and feed rail are critical determinants of dairy cows’ comfort and
welfare when they are housed in free-stall barns and milked in parlors. Overstocking may have
impacts on animal welfare if there are not enough comfortable lying spaces and enough space at
the feed rail so that all cows may lie or feed at the same time or for the amount of time needed.
When stocking rate has been assessed at farms milking with traditional parlors, overstocking has
been found to have a significant impact on cows’ behavior and welfare through increased
displacement at stalls and feed rail or reduced time spent lying or feeding, which can lead to
increases in lameness and decreases in body condition of cows, thus affecting milk yield.
However no studies have been conducted examining stocking rates for cows milked with an
automatic milking system (AMS). Though cow behavior is no longer synchronized by a fixed
milking routine in AMS, cows may still wish to sleep or feed synchronously or may follow a
circadian pattern when performing these behaviors. Therefore, it is important to examine the
impact of stall and feed rail stocking rates on dairy cows in AMS to determine the effects on
production and behavior. Therefore, the aim of this thesis was to examine the impact of stocking
rate in the stalls and at the feed rail on body condition score, lying behavior, lameness (gait
scores), and cleanliness of lactating dairy cows through two projects.

The first chapter of this thesis describes a survey of thirty-nine dairy farms that milked their
cows with AMS, which were assessed with respect to the impact of stocking rate at the feed rail,
stalls, and AMS on cow behavior and welfare. Data were collected from 40 focal lactating cows
over two visits to each farm. To measure stocking rate in the stalls, the number of cows in a pen
was divided by the number of stalls in the same pen. To measure stocking rate at the feed rail,

1

the length of feed rail in the pen was divided by 60 cm and compared to the number of cows in
the same pen. Also, to measure stocking rate at AMS, the number of cows divided by the number
of robots in the same pen. Stocking rates were related to measures of lameness and cleanliness as
well as to body condition scores and lying behavior. An assessment was made to determine if the
various stocking rates were correlated with one another (and thus interacted or acted
independently). The impact on delivery of fresh feed and of pushing feed back to the rail where
cows could reach it (feed push up) was also considered with respect to feed rail stocking rate and
cow body condition score (BCS).

The second chapter of this thesis examines the impact of experimentally altering the number of
stalls and amount of feed rail space provided per cow (stocking rate) at an AMS dairy.
Specifically, this chapter examines the effect of reducing the number of stalls available or the
amount of feed space per cow on displacement behavior. At the AMS study dairy, each of two
the pens contained 60 cows/pen and 58 free-stalls and 60 headlocks. The impact of stocking rate
was examined using 3 treatments: a 1:1 treatment with 58 stalls or 60 headlocks available for 60
cows; a 1:1.2 treatment with 50 stalls or 50 headlocks available for 60 cows; and a 1:1.5
treatment with 40 stalls or 40 headlocks available for 60 cows. Each stall or feed rail treatment
was applied for one week in a randomized order that was different for each pen, with a one-week
washout period between treatments. Observations of displacement behavior were made from
video recordings on the last two days of each treatment week to examine whether treatment
affected how often cows were displaced from stalls. In addition, data were recorded at the feed
rail the by live observation for the 45 min before fresh feed delivery and for the 45 min before
feed push up.

2

Cows in AMS have free movement to and from the AMS and may milk asynchronously as
opposed to cows milking as a single group in parlor-based systems; this freedom may result in
less synchrony of feeding and lying behavior. Additionally, the stocking rate in AMS systems
may also be influenced by the design of barn and traffic system (forced, free, or guided), and
producers’ management decisions such as how often to fetch. Further, when AMS are used,
farms typically alter management routines in their entirety as producers now spend less time
milking cows and have more time to devote to cow health, feeding or breeding routines. Results
from these studies may help producers determine optimal stocking rates for AMS dairies,
allowing them to optimize facility design and management in ways that maintain good welfare
and production without wasting space or resources.

3

CHAPTER 1
LITERATURE REVIEW

INTRODUCTION
Stocking rate in animal production systems refers to the number of animals with respect to a
particular resource (e.g., cows per water trough), and stocking density refers to the number of
animals for a given amount of space (e.g., cows per square meter of pen space). For simplicity,
from this point forward, the term stocking rate will be used to refer to both situations. On dairy
farms the two most commonly considered stocking rates are the number of cows per lying stall
and the number of headlocks (or amount of space) per cow at the feed rail. In automatic milking
systems (AMS), stocking rate can also be considered with respect to the number of cows per
robot. Stocking rates can have important consequences to animal welfare as they may influence
whether animals have adequate access to a resource in order to meet their needs. Increasing the
stocking rate at the feed rail and stalls can lead to changes in the behavior and biology of cows
(Huzzey et al., 2013). Overstocking of dairy free-stall barns (i.e., barns that allow cows to move
between feeding areas and lying stalls and to choose where they feed and eat, respectively)
reduces lying time and increases aggressive interactions (Friend et al., 1977; Weirenga, 1983;
Weirenga and Hopster, 1990; Fregonesi and Leaver, 2002). However, overstocking is attractive
to producers because it allows them to improve their return on facility investment (Bewley et al.,
2001).

The European Food Safety Authority has stated that overstocking is one of the most important
risks to the welfare of cattle in free-stall housing (EFSA, 2009a). Yet though dairy producers are

4

often advised to keep stocking density below 120% (e.g., NFACC, 2009), overstocking is
commonly found on commercial dairy farms in North America (USDA, 2007; von Keyserlingk
et al., 2012). Further, many countries, including the U.S., do not have national guidelines for cow
welfare. Canada does have national science-based guidelines, which specifically recommend that
cows should be stocked at a rate of less than 1.2 cows per stall, and have 60 cm of space per cow
at the feed bunk (NFACC, 2009). Although most dairy farm animal welfare standards are still
reliant on resource-based criteria, such as recommended stall stocking rates, there is increased
interest in assessing the outcomes or the consequences of these recommendations on the animals
themselves (Rushen et al., 2011). With regard to stall stocking rate, the desired outcome is that
cows are able to lie down as much as they need to; and the current Code of Practice for Care and
Handling of Dairy Cattle now has best practice recommendations aimed at cows lying for 12 h
per day (NFACC, 2009). However, many recommendations from producer groups remain vague,
simply advising that stall stocking rate should allow for adequate rest, exercise and feed and
water consumption (NMPF, 2010).

Interactive Impacts of Stocking Rate
Stocking rate of the milking herd is important for dairy cows in general because it can have a
direct effect on cows’ behavior, welfare, health and production as will be discussed below. In
addition to considering the stocking rate for stable groups of animals, the impact of mixing
animals into new groups or of group size changing for management reasons must also be
considered (Talebi et al., 2014). For dairy cows, the transition period, when cows move from
pregnancy to lactation, results in the entry of a cow into a new group. This is a risk factor for that
individual cow that, for example, can make her less likely to feed when space at the feed rail is

5

limited though a similar stocking rate may not affect her once she is an established member of
the group. At the group level, entry and exit of cows to and from the milking herd during the
start and end of lactation, respectively, can change stocking rate and impact behavior (Talebi et
al., 2014).

Stocking Rates and Lying Time
In general, the current recommendation given in most dairy cow welfare guidelines is that cows
in comfortable free stalls should spend an average of about 12 h/d lying down (EFSA, 2009b).
However, in North America there are large differences between farms in herd lying time, with
most farms having a herd mean lying time less than 12 h/d (Ito et al., 2009; USDA, 2007; von
Keyserlingk et al., 2012). The stocking rate of free stalls in parlor-milked dairy herds is a
significant element of cow comfort as it can impact lying time (von Keyserlingk et al., 2009) as
well as individual behavioral variability in lying behavior (Ito et al., 2014). Reducing time spent
lying down by cows leads to both behavioral and physiological stress (Munksgaard and
Simonsen, 1996). For example, overstocking can lead to lower milk production on farms with
fewer stalls per cow (Bach et al., 2008). According to Fregonosi and colleagues (2007), lying
time decreased 1.7 hour when stall stocking rate was 150% compared with a 100% stocking rate.
In particular, lying time decreased mid day and overnight as a result of overstocking (Fregonosi
et al., 2007); two times when the circadian rhythm of cows motivates them to lie down to rest
and synchronized lying behavior is commonly seen in herds (DeVries and von Keyserlingk,
2005; Overton et al., 2005; Tucker et al., 2008). One consequence of short lying times may be
lameness (Chapinal et al., 2009; Proudfoot et al., 2010, Bell et al., 2009; Cook et al., 2004).
However, overstocking has not yet been definitively proven to be a cause of lameness, but

6

effectively, there is a positive relationship between lying time and lameness, where the healthy
cows lie less than the lame cows (Chapinal et al., 2009; Chapinal et al., 2010a,b).

In addition to stall stocking density, the time away from the home pen for milking also likely
affects the time cows spend lying down and may be responsible for some of the variation
between farms in mean herd lying time. For example, several studies have shown that the time
cows lie down in stalls is reduced by high stocking rates (Fregonesi et al., 2007; Krawczel et al.,
2012; Telezhenko et al., 2012), while others have demonstrated an effect of time spent away
from the home pen for milking on lying time as well (Botheras, 2006; Gomez and Cook, 2010).
Further, stocking rate and time out of the pen for milking could interact to affect lying time, as
competition for lying stalls would be higher if fewer stalls were available and if less time were
available to lie down; but this interaction has not been studied explicitly.

Feed Rail Stocking Rates
Stocking rate at the feed rail is a key element that must be considered to ensure adequate feed
intake at dairy farms as well as cow comfort. For example, increased or decreased availability of
places per cow at the feed rail, depending on the level of competition, leads to a change in
behaviors such as time spent eating (Nielsen, 1999; Olofsson, 1999). It is commonly
recommended that each cow have approximately 60 cm of space at a linear feed rail (Grant and
Albright, 2001; NFACC, 2009), though one study at least suggests that this recommendation is
overly generous, as cows can have similar feed intake when given less feed space (Menzi and
Chase, 1994). Generally, however, feeding space available for each cow has been related to the
level of competition at the feed rail, and displacement at the feed rail increases when stocking

7

rate is increased at the feed rail (DeVries et al., 2004; Huzzey el at., 2006; Proudfoot et al.,
2009). However, reducing feed bunk space has not been directly linked with lower milk yield
(Bach et al., 2008; Proudfoot et al., 2009).

Another important factor to consider when evaluating the impact of space at the feed rail is that
cows prefer to feed at the same time, as they are social animals that naturally show synchronized
behavior, which is (von Keyserling and Weary, 2010). For example, in the 90 min after fresh
feed delivery, feeding activity was highest (DeVries and von Keyserlingk, 2005), and
overstocking at the feed rail reduced the number of cows that feed at this time (Huzzey et al.,
2006). This means all cows are motivated to feed at the same time, and so ensuring that all
animals have the space to feed simultaneously will allow cows with less competitive ability (e.g.,
fresh or primiparous cows) to feed with the herd. Thus, competition at the feed rail affects
feeding behavior and welfare as increased competition at high feed rail stocking density could
limit the ability of some cows to access to the feed rail at time of fresh feed delivery translating
into altered feeding behavior, reduced feed intake and lying time and increased time spent
standing (Collings et al., 2011; DeVries et al., 2004; Grant and Albright, 2001; Huzzey et al.,
2006; Krawczel et al., 2012).

Stocking Rate and Automatic Milking Systems
At present little is known about how stocking rate at stalls or the feed rail affects behavior or
welfare in herds milked with automatic milking systems (AMS). Cows in AMS have free
movement to and from the AMS and may milk asynchronously as opposed to cows milking
together as groups in parlor-based systems. On the other hand, stocking rate in AMS dairies may

8

also be influenced by other factors, such as the design of the barn and traffic system (forced, free,
or guided), the number of cows per AMS and whether cows are fed in the AMS. For example,
according to Uetake and colleagues (1997), the design of AMS milking and holding area impacts
synchronization of cows’ eating and resting and can reduce time spent eating, with a more
pronounced effect seen for the AMS holding area. However, as return from the milking parlor
appears to be one of the factors that synchronizes feeding in parlor-milked cows (DeVries et al.,
2003), it is possible that cows who are milking individually in AMS may also eat and sleep more
asynchronously compared to their parlor-milked counterparts. Regardless, cows may still wish to
lie down synchronously (Rushen et al., 2001). A limited number of studies examining lying time
for cows in AMS showed that cows spent between 10.8 -12.8 hours a day lying when stalls were
stocked at ≤ 100% (Lexer et al., 2009; DeVries et al., 2011; Munksgaard et al., 2011; Deming et
al., 2013). However, lying time in AMS has the potential to be greater than in parlor systems
because cows might spend less time on their feet waiting to be milked.

Generally, according to all studies been mentioned previously increasing stocking rate at stall
and feed rail have impact on cow behavior and welfare at the parlor-milk dairies. on other hand,
AMS dairies the impact on the stocking rate may be not strong as found at the parlor farms.
Cows in AMS have free movement to and from the robot and may milk asynchronously as
opposed to cows milking as a single group in parlor-based systems. However, cows may still
wish to lie down synchronously. Productively, AMS farms have more milk product from parlor
farms. Also, the AMS robot reduce the laborers in farm, thus leads to reduce the amount of cost;
unlike in parlor farms. Additionally, AMS farms alter management routines in their entirety as

9

producers now spend less time milking cows and have more time to devote to cow health,
feeding or breeding routines.

10

REFERENCES

11

REFERENCES

Bach, A., N. Valls, A. Solans, and T. Torrent. 2008. Association between nondietary factors and
dairy herd performance. J. Dairy Sci. 91:3259-3267.
Bell, N. J., M. J. Bell, T. G. Knowles, H. R. Whay, D. J. Main, and A. J. F. Webster. 2009. The
development, implementation and testing of a lameness control programme based on HACCP
principles and designed for heifers on dairy farms. Vet. J. 180:178–188.
Bewley, J., R. W. Palmer, and D. B. Jackson-Smith. 2001. A comparison of free-stall barns used
by modernized Wisconsin dairies. J. Dairy Sci. 84:528–541.
Botheras, N. A. 2006. The behaviour and welfare of grazing dairy cows (Bos taurus): Effects of
time away from pasture and position in the milking order. PhD Thesis. University of Melbourne,
Melbourne Australia.
Chapinal, N., A. M. de Passillé, D. M. Weary, M. A. G. von Keyserlingk, and J. Rushen. 2009.
Using gait score, walking speed, and lying behavior to detect hoof lesions in dairy cows. J. Dairy
Sci. 92:4365-4374.
Chapinal, N., A. M. de Passillé, and J. Rushen. 2010a. Correlated changes in behavioral
indicators of lameness in dairy cows following hoof trimming. J. Dairy Sci. 93: 5758-5763.
Chapinal, N., A. M. de Passillé, J. Rushen, and S. Wagner. 2010b. Effect of analgesia during
hoof trimming on gait, weight distribution and activity of dairy cattle. J. Dairy Sci. 93:3039–
3046.
Collings, L. K. M., D. M. Weary, N. Chapinal, and M. A. G. von Keyserlingk, 2011. Temporal
feed restriction and overstocking increase competition for feed by dairy cattle. J. Dairy Sci.
2011; 94: 5480–5486.
Cook, N. B., T. B. Bennett, and K. V. Nordlund. 2004. Effect of free stall surface on daily
activity patterns in dairy cows with relevance to lameness prevalence. J. Dairy Sci. 87:2912–
2922.
DeVries, T. J., M. A. G. von Keyserlingk. 2005. Time of feed delivery affects the feeding and
lying patterns of dairy cows. J. Dairy Sci. 88:625-631.
DeVries, T. J., M. A. G. von Keyserlingk, and K. A. Beuchemin. 2003. Short communication:
diurnal feeding pattern of lactating dairy cows. J. Dairy Sci. 86:4079-4082.
DeVries, T. J., M. A. G. von Keyserlingk, and D. M. Weary. 2004. Effect of feeding space on the
inter-cow distance, aggression, and feeding behavior of free-stall housed lactating dairy cows. J.
Dairy Sci. 87:1432–1438.
12

European Food Safety Authority (EFSA). 2009a. Scientific opinion on welfare of dairy cows in
relation to behaviour, fear and pain based on a risk assessment with special reference to the
impact of housing, feeding, management and genetic selection. EFSA J. 1139:1–66. EFSA,
Parma, Italy.
European Food Safety Authority (EFSA ). 2009b. Scientific Opinion of the Panel on Animal
Health and Welfare on a request from the Commission on the risk assessment of the impact of
housing, nutrition and feeding, management and genetic selection on metabolic and reproductive
problemsin dairy cows. The EFSA J. 1140:1–75. EFSA, Parma, Italy.
Fregonesi, J. A., and J. D. Leaver. 2002. Influence of space allowance and milk yield level on
behaviour, performance and health of dairy cows housed in strawyard and cubicle systems.
Livest. Prod. Sci. 78:245–257.
Fregonesi, J. A., C. B. Tucker, and D. M. Weary. 2007. Overstocking reduces lying time in dairy
cows. J. Dairy Sci. 90:3349-3354.
Friend, T. H., C. E. Polan, and M. L. McGilliard. 1977. Free stall and feed bunk requirements
relative to behavior, production and individual feed intake in dairy cows. J. Dairy Sci. 60:108–
116.
Gomez, A. and N. B. Cook. 2010. Time budgets of lactating dairy cattle in commercial freestall
herds. J. Dairy Sci. 93:5772-5781.
Grant, R. J., and J. L. Albright. 2001. Effect of animal grouping on feeding behavior and intake
of dairy cattle. J. Dairy Sci. 84(E. Suppl.):E156-E163.
Huzzey, J. M., T. J. DeVries, P. Valois, and M. A. G. von Keyserlingk. 2006. Stocking density
and feed barrier design affect the feeding and social behavior of dairy cattle. J. Dairy Sci.
89:126–133.
Huzzey, J. M., T. R. Overton, M. A. G. von Keyserlingk. 2013. Effect of stocking density on
feeding strategies and health. Pages 51-62 in Proceedings of the Tri-State Dairy Nutrition
Conference. Fort Wayne, IN.
Ito, K., D. M. Weary, and M. A. G. von Keyserlingk. 2009. Lying behavior: Assessing withinand between-herd variation in free-stall housed dairy cows. J. Dairy Sci. 92:4412–4420.
Ito, K., N. Chapinal, D. M. Weary, and M. A. G. von Keyserlingk. 2014. Associations between
herd-level factors and lying behavior of freestall-housed dairy cows. J. Dairy Sci. 97:2081-2089.
Krawczel, P. D., L. B. Klaiber, R. E. Butzler, L. M. Klaiber, H. M. Dann, C. S. Mooney, and R.
J. Grant. 2012. Short-term increases in stocking density affect the lying and social behavior, but
not the productivity, of lactating Holstein dairy cows. J. Dairy Sci.95:4298–4308.

13

Menzi, W., Jr., and L. E. Chase. 1994. Feeding behavior of cows housed in free stall barns.
Pages 829–831 in Dairy Systems for the 21st Century. Am. Soc. Agric. Eng., St. Joseph, MI.
Munksgaard, L., and H. B. Simonsen. 1996. Behavioral and pituitary adrenal-axis responses of
dairy cows to social isolation and deprivation of lying down. J. Anim. Sci. 74:769–778.
National Farm Animal Care Council (NFACC). 2009. Code of Practice for Care and Handling of
Dairy Cattle. Dairy Farmers of Canada, Ottawa, ON, Canada.
National Milk Produces Federation (NMPF). 2010. National Dairy Animal Care Manual
Available at: http://www.nationaldairyfarm.com/animal-care-resources.html.
Nielsen, B. L. 1999. On the interpretation of feeding behaviour measures and the use of feeding
rate as an indicator of social constraint. Appl. Anim. Behav. Sci. 63:79–91.
Olofsson, J. 1999. Competition for total mixed diets fed for ad libitum intake using one or four
cows per feeding station. J. Dairy Sci. 82:69-79.
Overton, M.W., W. M. Sischo, G. D. Temple, and D. A. Moore. 2002. Using time-lapse video
photography to assess dairy cattle lying behavior in a free-stall barn. J. Dairy Sci. 285: 24072413.
Proudfoot, K. L., D. M. Weary, and M. A. G. von Keyserlingk. 2009. Competition at the feed
bunk changes the feeding, Standing, and social behavior of transition dairy cows. J. Dairy Sci.
92:3116-3123.
Proudfoot, K. L., D. M. Weary, and M. A. G. von Keyserlingk. 2010. Behavior during transition
differs for cows diagnosed with claw horn lesions in mid lactation. J. Dairy Sci. 93:3970–3978.
Rushen J., L. Munksgaard, P.G. Marnet, and A. M. de Pasillé. 2001. Human contact and the
effects of acute stress cows at milking. Appl. Anim. Behav. Sci. 73:1–14.
Rushen, J., A. Butterworth, and J. C. Swanson. 2011. Animal Behavior and Well-Being
Symposium: Farm animal welfare assurance: Science and application. J. Anim. Sci. 89:1219–
1228.
Talebi, A., M. A. G. von Keyserlingk, E. Telezhenko, and D. M. Weary. 2014. Reduced stocking
density mitigates the negative effects of regrouping in dairy cattle. J. Dairy Sci. 97:1358-1363.
Telezhenko, E., M. A. G. von Keyserlingk, A. Talebi, and D. M. Weary. 2012. Effect of pen
size, group size, and stocking density on activity in freestall-housed dairy cows. J. Dairy Sci.
95:3064–3069.
Tucker, C. B. 2009. Behaviour of cattle. Pages 151-160 in The Ethology of Domestic Animals.
P. Jensen, ed. CABI, Oxfordshire, UK.

14

Tucker, C. B., A. R. Rogers, and K. E. Schutz. 2008. Effect of solar radiation on dairy cattle
behaviour, use of shade and body temperature in a pasture-based system. Appl. Anim. Behav.
Sci. 109:141-154.
USDA. 2007. Facility characteristics and cow comfort on U.S. dairy operations. No. 524.1210.
USDA-Animal and Plant Health Inspection Service (APHIS)-Veterinary Services (VS), Centers
for Epidemiology and Animal Health (CEAH), Fort Collins, CO.
Uetake, K., J. F. Hurnik, and L. Johnson. 1997. Behavioral pattern of dairy cows milked in a
two-stall automatic milking system with a holding area. J. Anim. Sci.75:954-958.
von Keyserlingk, M. A. G, and D. M. Weary. 2010. Review: feeding behavior of dairy cattle:
Measures and applications. Can. J. Dairy Sci. 90:303-309.
von Keyserlingk, M. A. G., J. Rushen, A. M. de Passillé, and D. M. Weary. 2009. Invited
review: The welfare of dairy cattle— Key concepts and the role of science. J. Dairy Sci.
92:4101–4111.
von Keyserlingk, M. A. G., A. Barrientos, K. Ito, E. Galo, and D. M. Weary. 2012.
Benchmarking cow comfort on North American freestall dairies: Lameness, leg injuries, lying
time, facility design, and management for high-producing Holstein dairy cows. J. Dairy Sci.
95:7399–7408.
Wierenga, H. K. 1983. The influence of the space for walking and lying in a cubicle system on
the behaviour of dairy cattle. Pages 171–180 in Farm Animal Housing and Welfare. S. H. Baxter,
J. A. C. McLornack, and M. R. Baxter, ed. Martinus Nijhoff, the Hague, the Netherlands.
Wierenga, H. K., and H. Hopster. 1990. The significance of cubicles for the behaviour of dairy
cows. Appl. Anim. Behav. Sci. 26:309–33.

15

CHAPTER 2
THE EFFECT OF STOCKING RATES ON LYING BEHAVIOR AND WELFARE OF
COWS MILKED WITH AUTOMATIC MILKING SYSTEMS

Interpretive summary: Stall and feed rail stocking rate in AMS. Witaifi
Canadian and U.S. farms using automatic milking systems (AMS) were surveyed to understand
how stocking rates of stalls, the feed rail, or the AMS may affect the behavior or welfare of cows
milked with AMS. Stocking rates were related to gait scores, cleanliness, lying behavior, and
body condition; however only decreased leg cleanliness was related to increased stall stocking
rate. Other management factors may minimize the impact of stocking rate in AMS systems. For
exampled increased stocking rate at the feed rail may have been compensated for by the frequent
feed delivery and push up observed in this study.

OUR INDUSTRY TODAY
The Effect of Stocking Rates on Lying Behavior and Welfare of Cows Milked with
Automatic Milking Systems
A. Witaifi,* G. Charlton,† J. Gibbons,§ E. Vasseur,‡ E. Pajor,^ J. Rushen,** A. M. de
Passille,** D. Haley,†† D. Pellerin,§§ T. DeVries,‡‡ J. M. Siegford*1
*

Department of Animal Science, Michigan State University, East Lansing, MI, US, 48824

†

Department of Animal Production, Welfare and Veterinary Sciences, Harper Adams University,

Shropshire, TF10 8NB, UK
§

DairyCo, Agriculture and Horticulture Development Board, Stoneleigh Park, Kenilworth,

Warwickshire, England, UK, CV8 2TL

16

‡

Alfred Organic Dairy Research Center, University of Guelph, 10 St-Paul Street, Alfred, ON,

Canada, K0B 1J0
^

University of Calgary Veterinary Medicine, 2500 University Dr. NW,
Calgary, AB,

Canada,
T2N 1N4
**

Pacific Agri-Food Research Centre, Agriculture and Agri-Food Canada, PO Box 1000, 6947

Highway 7, Agassiz, BC, Canada, V0M 1A0
††

§§
‡‡

Department of Population Medicine, University of Guelph, Guelph, ON, Canada, N1G 2W1
Department of Animal Science, Laval University, Quebec, QC, Canada, G1V 0A6
Department of Animal and Poultry Science, University of Guelph, Kemptville Campus,

Kemptville, ON, Canada, K0G 1J0

1

Corresponding author: Janice M. Siegford

Department of Animal Science, Michigan State University, 1290 Anthony Hall, East Lansing,
MI 48824, Phone: 517-432-1388, Fax: 517-353-1699, Email: siegford@msu.edu

17

ABSTRACT

Thirty-nine dairy farms that milked their cows with automatic milking systems (AMS) were
assessed with respect to the relationship of stocking rate at the feed rail, stalls, and AMS on cow
behavior and welfare. Stocking rate on farms milking with traditional parlors has a significant
impact on cows’ behavior and welfare through increased displacement at stalls and feed rail,
reduced lying time or reduced amount of time spent at the feed rail. These changes can lead to
increases in lameness and decreases in body condition of cows and affect milk yield. However,
no studies have examined stocking rate of farms milking with AMS. This portion of the study
focused on AMS-milked cows and highlights differences relating to stocking rates in these
systems versus traditional parlor-milked dairies. Data from each farm were collected from 40
focal lactating cows over two visits. On average, stalls were slightly understocked in this study,
with 0.94 ± 0.16 cows per stall. Average stocking rate at the feed rail was 1.07 ± 0.33 cows per
60 cm, and there was an average of 55 ± 8.6cows per AMS. Leg cleanliness was related to stall
stocking rate, with cleaner cows at lower stocking rates (P < 0.01). However, neither udder nor
flank cleanliness were related to stall stocking rate (P > 0.05 for both), and feed rail stocking rate
was not related to any cleanliness score (P > 0.05 for all). Cows spent an average of 11.2 ± 2.32
h lying down per day, but lying time was not related to stall, feed rail or AMS stocking rates (P <
0.05 for all). Stall, feed rail and AMS stocking rates were also not related to average lying bout
duration or number of lying bouts (P < 0.05 for both). Stocking rate at the feed rail was not
related to body condition score (P = 0.74). The presence of a limp, head bob, or gait asymmetry
was not significantly related to stocking rate of stalls or feed rail (P > 0.05 for all). Results from
this study indicate a less clear link between stocking rates in AMS dairies and cow behavior and

18

welfare than has been found in studies of parlor-milked dairies. Cows may milk asynchronously
as opposed to cows milking in groups in parlor-based systems; this freedom may result in less
synchrony of feeding and lying behavior. Additionally, impacts of stocking rate in AMS systems
may be influenced by barn design and traffic system (forced, free, or guided). Further, when
AMS are used, producers typically alter management routines in their entirety, spending less
time milking cows and more time on cow health, feeding or breeding routines. To fully explore
stocking rate implications in AMS, additional comparisons must be made between stocking rates
of stalls and feed rails (and AMS) at AMS- and parlor-milked dairies.

Key Words: robotic milking, stocking density, lying, gait score, body condition, cleanliness

19

INTRODUCTION

Stocking rates are important for lactating dairy cows in general (regardless of milking system)
because they can have direct effects on cows’ behavior, welfare, health and production.
Increasing the stocking rate at the feed rail and stalls can lead to changes in the behavior and
biology of cows (for a review see Huzzey et al., 2013). For example, increased or decreased
availability of places per cow at the feed rail, depending on the level of competition, can change
the time spent eating (Nielsen, 1999; Olofsson, 1999). For an individual cow, the entry into the
milking herd at the start of lactation (i.e., a new social group) is a risk factor that can make her
more likely to feed less when space is limited or competition is increased. At the group level,
entry and exit of cows to and from the milking herd during freshening and dry off respectively
can change stocking rate and affect social behavior (Talebi et al., 2014).

Stall Stocking Rate
While the U.S. does not have national guidelines for cow welfare, some recommendations advise
that the stocking rate should allow for adequate rest, exercise and feed and water consumption
(NMPF, 2010). Canada has national guidelines, which specifically recommend that cows should
have about 12 h of lying time per day and have 60 cm per cow at a linear feed bunk (NFACC,
2009). According to Fregonosi and colleagues (2007), lying time decreased 1.7 h/day when stalls
were stocked at 150% compared with a 100% stocking rate. Most noticeably, lying time
decreased in the overnight hours and during the mid-day period as a result of such overstocking
(Fregonosi et al., 2007). Bach and colleagues (2008) also found a correlation between high stall
stocking rate and lower milk production.

20

Feed Space Stocking Rate
Displacement at the feed rail has been found to increase when the feed rail stocking rate is
increased (DeVries et al., 2004; Huzzey el at., 2006; Proudfoot et al., 2009). However, reducing
feed bunk space has not been directly linked with lower milk yield (Bach et al., 2008; Proudfoot
et al., 2009). Another important factor to consider when evaluating the impact of space at the
feed rail is that cows are social animals that naturally show synchronized behavior and, thus,
prefer to feed at the same time, (von Keyserlingk and Weary, 2010). For example, in one study,
the feeding activity was highest in the 90 min after fresh feed delivery (DeVries and von
Keyserlingk, 2005).

Stocking Rate in Automatic Milking Systems
The stocking rate at stalls and feed rails in automatic milking systems (AMS) may also be
influenced by the design of barn and traffic system (forced, free, or guided), the number of cows
per AMS, and whether cows are fed in the AMS. Because AMS cows have relatively free
movement to and from the AMS, they may milk asynchronously as opposed to cows milking in
groups in parlor-based systems. Therefore, as return from the milking parlor appears to be one of
the factors that synchronizes feeding (DeVries et al., 2003), it is possible that cows in AMS who
are milking individually may also eat or lie down more asynchronously compared to their parlormilked counterparts.

A limited number of studies examining lying time for cows in AMS showed that cows spent
between 10.8 -12.8 hours a day lying down even when stalls were stocked at ≤ 100% (Lexer et

21

al., 2009; DeVries et al., 2011; Munksgaard et al., 2011; Deming et al., 2013). However, lying
time in AMS has the potential to be greater than in parlor systems because cows might spend less
time on their feet waiting to be milked.

Objectives
The stocking rates of stalls, feed rails, and AMS were assessed in herds of lactating dairy cows
that were milked with AMS to examine the impact of stocking rate on the cows’ behavior and
welfare. Stall stocking rate at each farm was related to gait scores, cow cleanliness, and lying
behavior while available space per cow at the feed rail was related to body condition and gait
scores. The AMS stocking rate at each farm was related to gait scores and lying behavior.

Hypotheses
We hypothesized that increased stall stocking rate would be related to higher (worse) gait and
cleanliness scores and more numerous lying bouts. Increased stall stocking rate would be related
to less total lying time and shorter lying bouts. Increased feed rail stocking rates would be related
to decreased BCS and shorter total lying time and lying bout durations. Increased feed rail
stocking rate would also be related to higher (worse) gait scores. Increased AMS stocking rate
would be related to higher (worse) gait scores and more numerous lying bouts. Finally, increased
AMS stocking rate would be related to less total lying time and shorter lying bout durations.

22

MATERIALS AND METHODS

The Farms
Stocking rates in the free stalls, at the feed rail and at the automatic milking systems (AMS) were
assessed in 39 lactating herds of dairy cows (30 Canadian and 9 Michigan farms).
Characteristics of these farms are described in Table 2.1 This study is complementary to an onfarm assessment study examining cow comfort on AMS farms. Two trained technicians recorded
data at each farm.

Animals
Prior to the start of the study, all animal-based protocols were approved by the Michigan State
University Institutional Animal Care and Use Committee.

At each farm, 40 focal cows were selected for detailed measurements using a list of cow
identification numbers obtained from the producer prior to the first farm visit. Focal cows at each
farm were ideally < 120 days in lactation and from a range of parities. For this stocking rate
study, body condition score (BCS), gait score parameters, cow width, and cow cleanliness were
measures of interest. However, before the survey start; there was a training and reliability testing
for observers. The width of cows was measured between the points of the hook bones. Gait was
assessed using a 0-1 scale for limping, head bob, and step asymmetry. Specifically, for limping
the following scores used: (0) no limp when all legs bear the weight of the cow equally, and (1)
limp when the cow walks with uneven and uneasy steps. Head bob was scored as: (0) no head
bob, just a natural up and down movement of the head when walking, and (1) an unnatural or

23

exaggerated head bob is observed when the cow walks. Step asymmetry was scored as: (0) the
feet hit the ground symmetrically with the rhythm of foot placement an even four beat gait, and
(1) asymmetric steps are noted because the rhythm of foot placement is not even, for example 1,
2, ... 3,4. Cow cleanliness was scored for the 40 focal cows using a 0-3 scale for the lower half of
the udder, the lower leg (the area from the top of the claw to the middle of the hock), and the
flank (the area from middle tarsal joint to a virtual line between pin and hook bones). All regions
were scored as follows: (0) the cow has fresh splashes of manure covering < 50% of the area; (1)
the cow has fresh splashes of manure covering > 50% of the area; (2) dried, caked manure and
fresh splashes of manure cover > 50% of the area; and (3) the entire area is covered with dried,
caked manure. Body condition scoring of the 40 focal cows at each farm was performed by using
the 1-5 scale and pictorial scorecard developed by Elanco Animal Health (2009) based on the
work of Ferguson and colleagues (1994). The BCS was recorded in increments of 0.25 points
with a lower limit of < 2.0 and an upper limit of 5.0 points. All focal cows scoring were
performed on the first visit to the farm.
Cows at the 30 Canadian farms were fitted with HOBO data loggers (HOBO Pendant® G Data
Logger - UA-004-64). Cows at the 8 Michigan farms were fitted with IceQube pedometers
(IceRobotics, Edinburgh, Scotland). In both cases, the devices were attached to a rear leg of the
cow between the fetlock and the pastern on the first visit to the farm. Loggers were removed
during a second visit to the farm that occurred after a minimum of 4 days. Average daily lying
time (hours/day) was then calculated for each cow from 4 complete days of data. Average lying
time per farm was calculated by adding together the average lying time of each cow then
dividing by the total number of cows that data were collected from at each farm. For further

24

information on the development of the animal-based measures used in this study, see Gibbons
and colleagues. (2012).

Barn Measurements
From the in-barn measurements that were taken as part of the on farm survey (Charlton et al.,
2014), the following measures were used in this portion of the study: number of cows per pen,
number of functional stalls per pen, amount of feeding space per pen (measured as length of feed
rail(s)), and number of AMS per pen.

Analysis
Stocking rate at the free stalls was calculated by dividing the number of cows in a pen by the
number of usable stalls in the same pen. The stocking rate at the feed rail was measured by
dividing the length of feed rail by 60 cm then dividing this number by the number of cows in the
same pen to create a cows/60 cm stocking rate (using the premise that each cow was
approximately 60 cm wide, and, in fact, average width of cows in this study was 57 cm).
Stocking rate at the AMS was measured by dividing the number of cows in the pen by the
number of AMS in the pen. Data analyzed per pen totaled from 44 observations when all pens at
each farm with focal were used. All analyses were completed using SAS (version 9.4, SAS
Institute Inc., Cary, NC).

Analyses of gait and cleanliness scores were conducted using nonparametric statistics (PROC
NPAR1WAY) because the data were not normal. The model was as follows:
Yij1 = μ + Ki + Cj + eijk

25

Where Ki is the fixed effect of stocking rate, Cj is the covariate of (lying behavior, gait score,
BCS, and Cleanliness score), and eijk is the residual.

, Data for gait score, cleanliness score, and body condition score were analyzed using from 39
observations (the total number of farms surveyed) including the pasture-based farms. To
examine the relationship between stocking rate and body condition score a mixed model was
used (PROC MIXED). The model was as follows:
Yijkl = μ + FDi + CNj + Cm + Cm×GCBL+ eijk
Where FDi is random effect of farm, CNj is the first fixed effect of cow, Ck is the second fixed
effect of stocking rate, GCBL is the covariate of (lying behavior, gait score, BCS, and
cleanliness score), Ck×GCBL is the interaction between stocking rate and covariate, and eijk is the
residuals, while the relationship between feed delivery and push up were analyzed using PROC
NPAR1WAY (for non-normal data). Data on feed delivery and feed push up were analyzed
using 33 observations because 6 farms had a missing data. Total lying time, number of lying
bouts, and lying bout duration were analyzed using a linear mixed model (PROC MIXED).Data
were analyzed using 37 observations as pasture-based farms were excluded.

RESULTS

The average stocking rate in free stalls (SRS) was 0.94 ± 0.16 (i.e., 0.94 cows kept per 1 stall)
with a minimum stocking rate of 0.56 cows per stall and a maximum of 1.26 cows per stall
(Table 2.2). The average stocking rate at the feed rail was 1.07 ± 0.33 cows per 60 cm of rail
26

space (i.e., 56.1 cm per cow), with a minimum feed rail stocking rate of 0.4 cows per 60 cm (i.e.,
150 cm per cow) and a maximum feed rail stocking rate of 1.92 (i.e., 31 cm per cow; Table 2.2).
In addition, the average stocking rate at the AMS was 55.1 ± 8.6 cows per AMS with a
maximum stocking rate of 72 cows per AMS and a minimum of 36 cows per AMS (Table 2.2).
There was no relationship between stall, feed rail, and AMS stocking rates (Figures 2.1, 2.2 and
2.3).

Descriptive data for total lying time (TL), number of lying bouts (NLB) and duration of lying
bouts (AVLBD) are presented in Figures 2.4, 2.5 and 2.6. The minimum TL was 9.78 ± 1.96 h,
the minimum NLB was 0.02 ± 0.14 bouts/60 min, and the minimum AVLBD was 0.89 ± 0.3 h
for cows at all farms. Further, the maximum TL was 13.27 ± 1.66 h, maximum NLB was 0.22 ±
0.02 bouts/60 min, and maximum AVLBD was 1.58 ± 0.57 h. Stall stocking rate was not related
to any measure of lying behavior (P > 0.05 in all cases). Stocking rate of the AMS was also not
correlated to TL (F4 = 1.77, P < 0.16), NLB (F4 = 0.45, P < 0.77) or AVLBD (F4 = 0.37, P <
0.83). Stall stocking rate was also not correlated to TL (F2 = 1.08, P < 0.35), NLB (F2 = 0.99, P <
0.38) or AVLBD (F2 = 0.90, P < 0.42). Feed rail stocking rate was also not significantly
correlated to TL (F3 = 1.82, P < 0.49), NLB (F3 = 0.76, P < 0.53) or AVLBD (F4 = 1.39, P<0.26).

Figures 2.7, 2.8, and 2.9 display the descriptive data for gait score variables on a per farm basis
(presence of a limp (LIMP), head bob (HB), and step asymmetry (ASY)). On average, cows
were relatively sound as shown by average scores of less than 1, when 0 indicates no problem
and 1 indicates the presence of a gait-related problem (Mean ± SD; LIMP = 0.16 ± 0.1; HB =
0.07 ± 0.1; and ASY = 0.55 ± 0.22). Stall stocking rate was not related to gait score parameters

27

(LIMP: F2 = 0.46, P < 0.63; HB: F2 = 1.35, P < 0.26; and ASY: F2 = 0.08, P < 0.93). In addition,
feed rail stocking rate was also not related to cow gait score parameters (LIMP: F3 = 0.75, P <
0.05; HB: F3 = 1.81, P < 0.14; and ASY: F3 = 0.92, P < 0.43). Nor was stocking rate at the AMS
related to LIMP (F4 = 0.47, P < 0.75) HB (F4 = 1.02, P < 0.40), or ASY (F4 = 0.76, P < 0.55).

Feed rail stocking rate was not related to body condition score (F3 = 0.39, P < 0.7), and the
average BCS for the study was 2.88 ± 0.32 (with descriptive BCS data from all farms presented
in Figure 2.10). Additionally, as Figures 2.11 and 2.12 show, there was no correlation between
feed rail stocking rate and frequency of fresh feed delivery (FED, F29 = 1.06, P < 0.57) or
between feed rail stocking rate and push up frequency (PU, F29 = 1.33, P < 0.47). Also, there
was no correlation between BCS and frequency of FED (F14 = 1.63, P < 0.16; Figure 2.13), BCS
or BCS and PU (F14 = 0.504, P < 0.90; Figure 2.14). In addition, there was no correlation
between FED and PU (F3 = 1.99, P < 0.14; Figure 2.15), meaning that farmers did not deliver
fresh feed less often as they pushed feed up to the rail more frequently and vice versa.

Stocking rate at stalls was correlated with leg cleanliness (LC, F2 = 4.85, P = 0.008) Average leg
cleanliness scores were 0.45 ± 0.46 (Mean ± SD, on a score of 0 – 3 with a 0 = very clean and 3
= very dirty); and scores at each farm are presented in Figure 2.16. On the other hand, feed rail
stocking rate was not correlated with LC (F3 = 1.76, P < 0.15). Figures 2.17 and 2.18 show udder
(UC) and flank cleanliness (FC) scores by farm. Stocking rate at the feed rail was not correlated
with UC (F3 = 1.10, P < 0.35) or FC (F3 = 1.71, P < 0.16) nor was stall stocking rate correlated
with UC (F2 = 0.11, P < 0.90) or FC (F2 = 0.6, P < 0.55).

28

DISCUSSION

In the present study, few relationships were found between stocking rates (at stalls, the feed rail
or the AMS) and variables of interest including lying behavior or lameness. In fact, only leg
cleanliness was significantly correlated with stall stocking rate. This is in contrast to what has
generally been observed in parlor-milked systems where increases in the stocking rate at stalls
and feed rail to a level greater than 100–113% have been found to lead to changes in behavior of
lactating dairy cows, especially with regard to lying time (Christopher et al., 2008).
Stall stocking rate observed at the surveyed farms ranged from 56 to 126%, with an average of
94% (i.e., less than one cow per stall); while stocking rate at the feed rail ranged from 0.4 to 1.92
cows per 60 cm of feed rail with an average of 1.07 cows/60 cm (i.e., 56.1 cm of feeding space
per cow). Stocking rate at the AMS ranged from 36-72 cows per AMS with an average stocking
rate of 55 ± 8.6 cows per AMS. Thus, in all cases, the average stocking rates were in accordance
with recommendations of various codes of practice and scientific studies (e.g., NFACC, 2009;
Anderson, 2008; Armstrong, 1994; Brouk, 2008; Collier et al., 2007). There were no cases of
extremely high stall stocking rate and only four cases when an AMS was stocked with more than
71 cows. However, 20% of the surveyed farms provided less than 45 cm of feed rail space per
cow, which could be expected to cause reduced feed intake and thus lower body condition
scores. It was surprising then, that our results found no effect of high feed rail stocking rate on
body condition, measures of lamenes, or gait scores (Grant, 2009).

Cows synchronize their behavior as a result of circadian rhythms and group dynamics, with the
result that reducing the number of stalls may lead to less lying and increased standing (Fregonesi
29

et al., 2007), and may also influence to when cows feed or milk. Stocking rate at the stalls, feed
rail, and AMS can each affect cow health and lying time individually, and it is also possible that
effects of the various stocking rates may interact. Thus, we examined the correlation between
stall stocking rate, feed rail stocking rate, and AMS stocking rate to better understand whether
they were related, which would alter how we assessed the effects of these three stocking rates on
our variables of interest such as lying behavior, body condition score, gait, or cleanliness.
However, we found no significant relationships between the stocking rates in our study which
may mean these variables may have stronger individual rather than interactive effects on cow
behavior and welfare.

On average, cows spend between 8-12 hours in a day lying when the stocking rate in free stalls is
1:1 (Jensen, 2009F). Further, lying time decreases particularly in the overnight hours and during
the mid-day period, as a result of stall overstocking (Fregonosi et al., 2007). However, in the
present study, we found no relationship between stocking rate at the feed rail, stalls, or AMS
with lying time, the number of lying bouts or lying bout duration.

According to our analysis, there was also no relationship between gait scores and the stocking
rates at the 39 farms that were examined. Specifically, the data showed no relationship between
presence of a limp, step asymmetry or head bob and the stocking rate at the stalls, feed rails or
AMS at the surveyed farms. According to Galindo and Broom (2000), cows with lower access to
stalls and feed rail had increased standing time, which increased their risk of becoming lame.
However, it is important to note that there is also a positive relationship between lameness and

30

lying time, where lame cows lay more than healthy cows (Chapinal et al., 2009; Chapinal, 2010a,
b).

The average body condition score of cows in this study was 2.88 (on a 5 point scale). According
to the Canadian Code of Practice for dairy cattle, the ideal body condition score for a dairy cow
is 2.50-3.25 for the first third of lactation, the stage targeted for focal cows in our study
(NFACC, 2009). We found that neither stocking rate at the feed rail or AMS had an effect on
BCS. According to Thorne and colleagues (2007), high stocking rates lead to loss in body
condition score and poor herd health. Given that we did not see an impact of high feed rail
stocking rates on BCS in the farms we surveyed, it was possible that managers had increased
frequency of feeding or feed push up frequency to compensate. Therefore, we analyzed the data
to evaluate relationships between stocking rate at the feed rail and body condition score,
frequency of fresh feed delivery, or push up of feed to the rail. We did not find any significant
relationships between these variables; however, producers were typically delivering frsh feed or
pushing up feed more than three times per day. According to Huzzey and colleagues (2006) the
aggressive displacements seemed lessin headlocks feed barrier than at the post-and-rail barrier.
This feeding management resulted in over 90% of cows having feed in front of them during all
of our observations at each farm. Thus, good feeding management may have been compensating
for high feed rail stocking rates at the eight farms with limited feed rail space. Another
possibility is that the cows at farms with high feed rail stocking rates could be feeding
asynchronously or adapting their feeding to times when the rail is less crowded.

31

According to Krawczel there is no relationship between udder and leg hygiene and stocking rate
(Krawczel et al., 2012). Though another study reported that lower ranking cows were more likely
to lie down outside of free stalls when there is overstocking (Gonzalez et al., 2003) and might be
expected to be dirtier. In our study, stocking rate at the stall and feed rail did not correlate with
udder and flank cleanliness, which may mean the producers were keeping the stall bedding clean
and dry. Leg cleanliness, however, showed a significant correlation with stall stocking rate, but
there was no relationship of leg cleanliness with feed rail stocking rate; perhaps indicating that
time spent scraping alleys in the stall area has an effect on cow leg cleanliness.

CONCLUSIONS

In this study, we focused on examining the impact of stocking rates on cow behavior and welfare
parameters rather than on milk production. Our data showed no strong relationships between
stocking rates and lying behavior, body condition, gait scores, or cleanliness scores. However,
cows in AMS have free movement to and from the AMS and may milk asynchronously, as
opposed to cows milking as groups in parlor-based systems. This freedom may result in less
synchrony of feeding and lying behavior. Also, the stocking rate in AMS systems may also be
influenced by the design of barn and traffic system (forced, free, or guided) cow are also fed
concentrate in the AMS; which may mitigate high stocking rates at the feed rail. Typically
producers who adopt AMS alter management routines in their entirety, spend in less time
milking cows and now have more time on cow health, feeding or breeding routines. This may
allow producers to make management changes that can compensate for any issues caused by

32

high stocking rates. For example, delivering or pushing up feed frequently to provide a constant
source of feed for cattle when little feed rail space is available may reduce competition and
mitigate any impacts on body condition score.

ACKNOWLEDGEMENTS

We would like to thank the Michigan State University Department of Animal Science for
supporting this research through an award from the Elwood Kirkpatrick Dairy Research Science
Endowment to Janice Siegford (East Lansing, MI). This study was funded by Agriculture &
Agri-Food Canada and Dairy Farmers of Canada (Ottawa, Ontario, Canada) as part of the Dairy
Science Cluster initiative. We thank the collaborators, students and co-op students from
Agriculture & Agri-Food Canada (Agassiz, British Columbia, Canada), University of British
Columbia (Vancouver, British Columbia, Canada), University of Calgary (Calgary, Alberta,
Canada), University of Guelph (Guelph, Ontario, Canada), Université Laval (Quebec City,
Quebec, Canada), and Valacta (Sainte-Anne-de-Bellevue, Quebec, Canada). Keri Dunn played a
vital role in collecting data on farm and with the producers in Michigan as part of this study. We
also acknowledge the support and assistance of the dairy farmers in Canada and the U.S. who
participated in this study.

33

APPENDIX

34

TABLES
Table 2.1 Characteristics of AMS farms in the survey

Number of AMS
Number of Stalls/AMS
Number of Cows/AMS
Number of AMS/pen
Months operating the AMS
Average number of milkings/cow/d
Minimum # visits to AMS/cow/d
Maximum # visits to AMS/cow/d
Average feed for cows in AMS/d (kg)
Average feed for heifers in AMS/d (kg)
Number of feed deliveries/d
Number of feed push ups/d
Permitted time between milkings (h)
Permitted yield between milkings (kg)
Number of times cows fetched/d
Number of cows fetched/d
Time elapsed between milkings before
fetching cows (h)
Number of times AMS cleaned/d
Duration of AMS cleaning by a day
(min)

Minimum
1
1
36
1
6
1
1
3
2
2
1
0
3.5
6.8
1
1
8

Median
2
1
55
1
23.5
3
2
5
4.6
4.3
>1
>1
6
8
2.5
12
12

Maximum
8
8
72
3
108
7
3
7
19.8
8
>3
>3
8
30
12
60
16

1
5

2
25

4
60

35

Table 2.2 Descriptive data from the 39 farms surveyed showing stall stocking rate (cow/stall),
feed rail stocking rate (cow/60 cm by 24 headlocks, 19 post-rail, and 1 swing rail), and stocking
rate at the automatic milk system (cow/AMS).

Farm ID
1
2
3A
3B
4
5
6
7
8
9
10
11A
11B
12
13
14
15
16
17
18A
18B
18C
19
20
21
22
23A
23B
24
25A
25B
26
27
28
29
30
31
32
33
34
35
36
37
38
39

# Cows/Pen
148
92
57
63
57
54
60
100
67
47
144
60
63
60
140
59
62
121
99
119
125
127
87
60
56
58
51
51
65
55
52
68
71
42
96
64
50
55
63
190
57
60
48
59
83

# Cows/Stall
0.9
0.96
1.02
0.91
1
0.68
0.95
0.91
0.56
1
1.15
1.05
1.11
0.87
0.67
0.98
1.05
0.98
1.13
1.08
1.17
1.19
0.75
0.92
0.85
0.95
0.84
0.84
0.99
0.97
0.96
0.85
0.93
0.64
1.04
0.96
0.78
0.98
1.26
1.09
0.95
0.95
0.68
1.07
0.66

36

# Cows/60 cm
feed rail
1.92
1.17
1.4
1.9
0.9
0.89
0.96
0.76
0.4
0.8
1.61
0.87
0.9
1.17
0.8
0.93
1.05
1.74
1.14
1.12
1.19
1.21
0.7
0.97
0.9
1.22
0.72
0.6
1.05
0.85
0.8
0.62
1.06
0.77
1.23
1.15
1.12
0.87
1.38
1.43
1.14
1.16
0.85
1.7
1

# Cows/AMS
49
31
57
63
57
54
60
100
34
47
48
60
63
30
140
59
62
61
99
59.5
63
64
44
60
56
58
51
51
65
55
52
34
71
42
96
64
50
55
63
63
57
60
48
59
41.5

FIGURES
Figure 2.1 Correlational analysis showing no relationship between stall stocking rate (SRS) and
feed rail stocking rate (SRF; P = 0.25).

Fit Plot For SRS v. SRF

Fit Plot for SRS
1.4

/Stall
Cows
SRS

1.2

Observations
44
Parameters
2
Error DF
42
MSE
0.0187
R-Square
0.2134
Adj R-Square 0.1946

1.0

0.8

0.6

0.5

1.0

1.5

Cows/60 SRF
cm Feed Rail

Fit

95% Confidence Limits

95% Prediction Limits

37

2.0

Figure 2.2 Correlational analysis showing no relationship between AMS stocking rate (SRA) and
stall stocking rate (SRS; P = 0.08).

Fit Plot For SRA vs. SRS

Fit Plot for SRA
100

Cows/AMS
SRA

80

Observations
44
Parameters
2
Error DF
42
M SE
172.64
R-Square
0.2276
Adj R-Square 0.2092

60

40

20

0.6

0.8

1.0

1.2

Cows/Stall
SRS
Fit

95% Confidence Limits

95% Prediction Limits

38

Figure 2.3. Correlational analysis showing no relationship between AMS stocking rate (SRA)
and feed rail stocking rate (SRF; P = 0.11).

Fit Plot for SRA vs. SRF

Fit Plot for SRA

100

SRA
Cows/AMS

80

Observations
44
Parameters
2
Error DF
42
M SE
215.45
R-Square
0.0361
Adj R-Square 0.0131

60

40

20
0.5

1.0

1.5

Cows/60
SRFcm Feed Rail
Fit

95% Confidence Limits

95% Prediction Limits

39

2.0

Figure 2.4 Descriptive data showing average total lying time for each farm over 24 h (created
using an average from 4 days of data per cow then averaging all 40 focal cows on the farm). Data
are presented according to number of AMS in each pen: orange = 3 AMS/pen, gray = 2
AMS/pen, and blue = 1 AMS/pen.

Total Lying Time
14
12
10

Hours/Day

8
6
4
2
0
1 2 10 34 8 16 1819 26 39 3 4 5 6 9 1113 14 1517 20 2122 23 2425 27 2829 3031 32 3335 36 3738
Farms

40

Figure 2.5 Descriptive data showing the average lying bout duration of each farm (created using
an average from 4 days of data per cow then averaging all 40 focal cows on the farm). Data are
presented according to number of AMS in each pen: orange = 3 AMS/pen, gray = 2 AMS/pen,
and blue = 1 AMS/pen.

Lying Bout Duration (h)
1.8
1.6
1.4

Hours

1.2
1
0.8
0.6
0.4
0.2
0
1 2 1034 8 1618192639 3 4 5 6 9 11131415172021222324252728293031323335363738
Farms

41

Figure 2.6 Descriptive data showing the average number of lying bouts for each farm (created
using an average from 4 days of data per cow, then averaging all 40 focal cows on the farm).
Data are presented according to the number of AMS in each pen: orange = 3 AMS/pen, gray = 2
AMS/pen, and blue = 1 AMS/pen.

Avg. Number of Lying Bouts
0.25

# Lying Bouts/60 min

0.2

0.15

0.1

0.05

0
1 2 10 34 8 16 18 19 26 3 4 5 6 9 11 13 14 15 17 20 21 22 23 24 25 27 28 29 30 31 32 33 35 36 37 38
Farms

42

Figure 2.7 Descriptive data showing presence of a limp (which was scored as either absent (0) or
present (1)) for 40 focal cows at each farm. Data are presented according to number of AMS in
each pen: orange = 3 AMS/pen, gray = 2 AMS/pen, and blue = 1 AMS/pen.

Presence of a Limp
0.6

0.5

Limpe Score

0.4

0.3

0.2

0.1

0
1 2 1034 8 121618192639 3 4 5 6 7 9 11131415172021222324252728293031323335363738
Farms

43

Figure 2.8 Descriptive data showing head bob (which was scored as either absent (0) or present
(1)) associated with lameness for 40 focal cows at each farm. Data are presented according to the
number of AMS in each pen: orange = 3 AMS/pen, gray = 2 AMS/pen, and blue = 1 AMS/pen.

Head Bob
0.4
0.35

Head Bob Score

0.3
0.25
0.2
0.15
0.1
0.05
0
1 2 1034 8 121618192639 3 4 5 6 7 9 11131415172021222324252728293031323335363738
Farms

44

Figure 2.9 Descriptive data showing step asymmetry (which was scored as either absent (0) or
present (1)) associated with lameness for 40 focal cows at each farm. Data are presented
according to number of AMS in each pen: orange = 3 AMS/pen, gray = 2 AMS/pen, and blue = 1
AMS/pen.

Step Asymmetry
1.2

1

Asymmetric Score

0.8

0.6

0.4

0.2

0
1 2 1034 8 121618192639 3 4 5 6 7 9 11131415172021222324252728293031323335363738
Farms

45

Figure 2.10 Descriptive data showing body condition (which was scored on 1-5 scale in 0.25
increments with 1 = a very thin cow and 5 = an obese cow) of 40 focal cows at each farm. Data
are presented according to number of AMS in each pen: orange = 3 AMS/pen, silver = 2
AMS/pen, and blue = 1 AMS/pen.

Body Condition Score
4.5
4

Body Condition Score

3.5
3
2.5
2
1.5
1
0.5
0
1 2 1034 8 121618192639 3 4 5 6 7 9 11131415172021222324252728293031323335363738
Farms

46

Figure 2.11 Correlational analysis showing no relationship between feed rail stocking rate and
frequency of feed delivery (P = 0.57).

Fit Plot for Feed Delivery

Fit Plot for FED

4

FED
Day
Number of Feed Deliveries/

3
Observations
33
Parameters
2
Error DF
31
M SE
0.536
R-Square
0.2605
Adj R-Square 0.2366

2

1

0

-1
0.50

0.75

1.00

1.25

1.50

Cows/60
cm Feed Rail
SRF
Fit

95% Confidence Limits

95% Prediction Limits

47

1.75

Figure 2.12 Correlational analysis showing no relationship between frequency of feed push up
and feed rail stocking rate (P = 0.47).

FitFit
Plot
forfor
Push
Plot
PUUp

Push-Ups/Day
Number of FeedPU

4

Observations
33
Parameters
2
Error DF
31
M SE
1.4335
R-Square
0.0105
Adj R-Square -0.021

2

0

0.50

0.75

1.00

1.25

1.50

Cows/60 SRF
cm Feed Rail
Fit

95% Confidence Limits

95% Prediction Limits

48

1.75

Figure 2.13 Correlational analysis showing no relationship between body conditions score and
frequency of feed delivery (P = 0.16).

Feed Fit
Delivery
vs.FED
Body Condition
Plot for

Number of Feed
FED Deliveries/Day

4

3

Observations
33
Parameters
2
Error DF
31
M SE
0.7108
R-Square
0.0193
Adj R-Square -0.012

2

1

0

2.4

2.6

2.8

3.0

3.2

BCS
BCS
Fit

95% Confidence Limits

95% Prediction Limits

49

3.4

Figure 2.14 Correlational analysis showing no relationship between BCS and frequency of feed
push up (P = 0.9).

Feed Push
vs.for
Body
FitUp
Plot
PU Condition

3
Observations
33
Parameters
2
Error DF
31
M SE
1.448
R-Square
0.0005
Adj R-Square -0.032

2

PU

Number of Feed Push Ups/Day

4

1

0

-1
2.4

2.6

2.8

3.0

3.2

BCS
BCS
Fit

95% Confidence Limits

95% Prediction Limits

50

3.4

Figure 2.15 Correlational analysis showing no a relationship between push up and feed delivery
(P = 0.14).

Number
PUof Feed Deliveries/Day

Feed Delivery
vs. PU
Feed Push Up
Fit Plot for

4

Observations
33
Parameters
2
Error DF
31
M SE
1.2356
R-Square
0.1471
Adj R-Square 0.1195

2

0

1.0

1.5

2.0

2.5

FED
# Feed Push
Ups/Day
Fit

95% Confidence Limits

95% Prediction Limits

51

3.0

Figure 2.16 Descriptive data showing leg cleanliness (which was scored on a 0-3 scale, with 0
representing clean and 3 representing very dirty) for 40 focal cows on each farm. Data are
presented according to the number of AMS in each pen: orange = 3 AMS/pen, gray = 2
AMS/pen, and blue 1 = AMS/pen.

Leg Cleanliness
2.5

2

Score

1.5

1

0.5

0
1 2 1034 8 121618192639 3 4 5 6 7 9 11131415172021222324252728293031323335363738
Farms

52

Figure 2.17 Descriptive data showing udder cleanliness (which was scored on a 0-3 scale, with 0
representing clean and 3 representing very dirty) for 40 focal cows on each farm. Data are
presented according to the number of AMS in each pen: orange = 3 AMS/pen, gray = 2
AMS/pen, and blue 1 = AMS/pen.

Udder Cleanliness
1.2

1

Score

0.8

0.6

0.4

0.2

0
1 2 1034 8 121618192639 3 4 5 6 7 9 11131415172021222324252728293031323335363738
Farms

53

Figure 2.18 Descriptive data showing flank cleanliness (which was scored on a 0-3 scale, with 0
representing clean and 3 representing very dirty) for 40 focal cows on each farm. Data are
presented according to the number of AMS in each pen: orange = 3 AMS/pen, gray = 2
AMS/pen, and blue = 1 AMS/pen.

Flank Cleanliness
0.7

0.6

0.5

Score

0.4

0.3

0.2

0.1

0
1 2 1034 8 121618192639 3 4 5 6 7 9 11131415172021222324252728293031323335363738
Farms

54

REFERENCES

55

REFERENCES

Anderson, N. 2008. Cow behaviour to judge free-stall and tie-stall barns. Pages 1-8 in Food and
Rural Affairs. Ontario Ministry of Agriculture.
Armstrong, D.V. 1994. Heat stress interactions with shade and cooling. J. Dairy Sci. 77:20442050.

Bach, A., N. Valls, A. Solans, and T. Torrent. 2008. Association between nondietary factors and
dairy herd performance. J. Dairy Sci. 91:3259-3267.

Brouk, M. 2005. Evaluating and selecting cooling systems for different climates. Pages 33-40 in
Proc. 7th Western Dairy Management Conference. Reno, NV.

Chapinal, N., A. M. de Passillé, D. M. Weary, M. A. G. von Keyserlingk, and J. Rushen. 2009.
Using gait score, walking speed, and lying behavior to detect hoof lesions in dairy cows. J. Dairy
Sci. 92(9):4365-4374.

Chapinal, N., A. M. de Passillé, and J. Rushen. 2010a. Correlated changes in behavioral
indicators of lameness in dairy cows following hoof trimming. J. Dairy Sci. 93: 5758-5763.

Chapinal, N., A. M. de Passillé, J. Rushen, and S. Wagner. 2010b. Automated methods for
detecting lameness and measuring analgesia in dairy cattle. J. Dairy Sci. 93: 2007-2013.

Charlton, G.L., D.B. Haley, J. Rushen, and A. M. de Passillé. 2014. Stocking density, milking
duration, and lying times of lactating cows on Canadian freestall dairy farms. J. Dairy Sci.
97:2694–2700.

Christopher T. H., P. D. Krawczel, H. M. Dann, C. S. Ballard, R.C. Hovey, W. A. Falls, R. J.
Grantemail. 2008. Effect of stocking density on the short-term behavioural responses of dairy
cows. J. Appl. Anim. Behav. Sci. 117:144-149.

Collier, R. J., and R. B. Zimbelman. 2007. Heat stress effects on cattle: What we know and what
we don't know. Pages 76-83.

56

Deming, J. A., R. Bergeron, K. E. Leslie, and T. J. DeVries. 2013. Associations of housing,
management, milking activity, and standing and lying behavior of dairy cows milked in
automatic systems. J. Dairy. Sci. 96:344-351.

DeVries, T. J., and M. A. G. von Keyserlingk. 2005. Time of feed delivery affects the feeding
and lying patterns of dairy cows. J. Dairy Sci. 88:625-631.

DeVries, T. J., M. A. G. von Keyserlingk, and K. A. Beuchemin. 2003. Short communication:
diurnal feeding pattern of lactating dairy cows. J. Dairy Sci. 86:4079-4082.

DeVries, T. J., M. A. G. von Keyserlingk, and D. M. Weary. 2004. Effect of feeding space on the
inter-cow distance, aggression, and feeding behavior of free-stall housed lactating dairy cows. J.
Dairy Sci. 87:1432–1438.

DeVries, T. J., J. A. Deming, J. Rodenburg, G. Seguin, K. E. Leslie, and H. W. Barkema. 2011.
Association of standing and lying behavior patterns and incidence of intramammary infection in
dairy cows milked with an automatic milking system. J. Dairy Sci. 94:3845-3855.

Elanco Animal Health. 2009. The 5-point body condition scoring system. Bulletin AI 10752.
Elanco Animal Health, Greenfield, IN.

Fregonesi, J. A., C. B. Tucker, and D. M. Weary. 2007. Overstocking reduces lying time in dairy
cows. J. Dairy Sci. 90:3349-3354.

Ferguson, J.D., D.T. Galligan, and N. Thomsen. 1994. Principal descriptors of body condition in
Holstein dairy cattle. J. Dairy Sci. 77:2695-2703.

Galindo, F., and D. M. Broom. 2000. The relationships between social behaviour of dairy cows
and the occurrence of lameness in three herds. Res. Vet. Sci. 69:75–79.

Gibbons, J., E. Vasseur, J. Rushen, and A. M. de Passille. 2012. A training programme to ensure
high repeatability of injury scoring of dairy cows. J. Animal Welf. 21: 379-388.

Gonzalez, M., A. K. Yabuta, and F. Galindo. 2003. Behaviour and adrenal activity of
firstparturition and multiparous cows under a competitive situation. Apple. Anim. Behav. Sci.
83: 259-266.
57

Grant, R. J. and J. L. Albright. 2001. Effect of animal grouping on feeding behavior and intake of
dairy cattle. J. Dairy Sci. 84(E. Suppl.): E156-E163.

Grant, R. 2009. Stocking density and time budgets. pages 7-17 in Proceedings of Western Dairy
Management Conference.

Huzzey, J. M., T. J. DeVries, P. Valois, and M. A. G. von Keyserlingk. 2006. Stocking density
and feed barrier design affect the feeding and social behavior of dairy cattle. J. Dairy Sci.
89:126-133.

Huzzey, J. M., T. R. Overton, M. A. G. von Keyserlingk. 2013. Effect of stocking density on
feeding strategies and health. Pages 51-62 in Proceedings of the Tri-State Dairy Nutrition
Conference. Fort Wayne, IN.

KRAWCZEL P.D., L.B. KLAIBER, R.E. BUTZLER, L.M. KLAIBER, H.M. DANN, C.S.
MOONEY, and R.J.GRANT. 2012. Short-term increases in stocking density affect the lying and
social behavior, but not the productivity, of lactating Holstein dairy cows. J. Dairy Sci. 95:42984308.

Lexer, D., K. Hagen, R. Palme, J. Troxler, and S. Waiblinger. 2009. Time budgets and
adrenocortical activity of cows milked in a robot or a milking parlour: Interrelationships and
influence of social rank. Anim. Welf. 18:73–80.

Munksgaard, L., J. Rushen, A. M. de Passill, and C. C. Krohn. 2011. Forced versus free traffic in
an automated milking system. Livest. Sci. 138:244–250.

National Farm Animal Care Council (NFACC). 2009. Code of practice for care and handling of
Dairy Cattle. Dairy Farmers of Canada, Ottawa, ON, Canada.

National Milk Produces Federation (NMPF). 2010. National Dairy Animal Care Manual
Available at: http://www.nationaldairyfarm.com/animal-care-resources.html.

Nielsen, B. L. 1999. On the interpretation of feeding behaviour measures and the use of feeding
rate as an indicator of social constraint. Appl. Anim. Behav. Sci. 63:79–91.

58

Olofsson, J. 1999. Competition for total mixed diets fed for ad libitum intake using one or four
cows per feeding station. J. Dairy Sci. 82:69-79.

Proudfoot, K. L., D. M. Weary, and M. A. G. von Keyserlingk. 2009. Competition at the feed
bunk changes the feeding, Standing, and social behavior of transition dairy cows. J. Dairy Sci.
92:3116-3123.

Talebi, A., M. A. G. von Keyserlingk, E. Telezhenko, and D. M. Weary. 2014. Reduced stocking
density mitigates the negative effects of regrouping in dairy cattle. J. Dairy Sci. 97:1358-1363.

Tucker, C. B. 2009. Behaviour of cattle. Pages 151-160 in The Ethology of Domestic Animals.
P. Jensen, ed. CABI, Oxfordshire, UK.

von Keyserlingk, M. A. G, and D. M. Weary. 2010. Review: feeding behavior of dairy cattle:
Measures and applications. Can. J. Dairy Sci. 90:303-309.

59

CHAPTER 3
IMPACT OF STALL AND FEED RAIL STOCKING RATE ON COW BEHAVIOR AND
PRODUCTION IN AUTOMATIC MILKING SYSTEMS

Interpretive summary: Stall and feed rail stocking rate affects behavior in AMS. Witaifi
This experimental study examined the impact of stall and feed rail stocking rates at AMS on cow
behavior and welfare. Three experimental treatments (at stalls or feed rail separately) were 1:1
(100%), 1:1.2 (120%), and 1:1.5 (150%) stocking rates. These stocking rates were related to stall
and feed rail displacement (after feed delivery and push up) and time budgets for stalls and feed
rails; however, only after fresh feed delivery did displacement increase due to high stocking
rates. Cow behavior and welfare in AMS does not appear as tightly linked to stall and feed rail
stocking rates as in parlor-milked dairies.

OUR INDUSTRY TODAY
Impact of Stall and Feed Rail Stocking Rate on Cow Behavior and Production in
Automatic Milking Systems
A.A. Witaifi* and J.M. Siegford*1
*

Animal Behavior and Welfare Group, Department of Animal Science, Michigan State

University, East Lansing, 48824.
1

Corresponding author: Janice M. Siegford

Department of Animal Science, Michigan State University, 1290 Anthony Hall, East Lansing,
MI 48824, Phone: 517-432-1388, Fax: 517-353-1699, Email: siegford@msu.edu

60

ABSTRACT
The number of stalls and amount of space at the feed rail provided per cow (stocking rate) on
dairy farms is important because overstocking may impact animal welfare if there are not enough
comfortable lying spaces or space for feeding. Research has shown the impact of high stall
stocking rates on cows in parlor-milked systems; however, no studies have examined stocking
rates for cows milked with automatic milking systems (AMS). Though cow behavior is no longer
synchronized by a fixed milking routine in AMS, cows may still wish to lie down or eat
synchronously). Therefore, it is important to examine the impact of stocking rates on dairy cows
in AMS to determine effects on production and behavior. This experiment examined the effect
of reducing the number of stalls available or the feeding space per cow on displacement and
other behaviors in an AMS dairy farm. Each of two pens contained 60 cows/pen and 58 freestalls and 60 headlocks. The impact of stocking rate was examined using three treatments: 1) 1:1
treatment with 58 stalls or 60 headlocks available for 60 cows; 2) 1:1.2 treatments with 50 stalls
or 50 headlocks available for 60 cows; and, 3) 1:1.5 treatment with 40 stalls or 40 headlocks
available for 60 cows. Each stall or feed rail treatment was applied separately, and each was
applied for 1wk in a randomized order that was different for each pen, with a 1-wk washout
period between treatments. Observations of displacement behavior were made from video
recordings on the last 2 d of each treatment week to examine whether treatment affected how
often cows were displaced from stalls. Displacement data were recorded at the feed rail by live
observation for the 45 min after fresh feed delivery and for the 45min after feed push up. Results
revealed no difference in number of displacements at the stalls between the treatments (F2 = 5.22,
P = 0.16). However, a difference in number of displacements among three treatments was found
in the period immediately after feed delivery (F2 = 86.33, P = 0.01), when a difference was seen

61

between treatments 1:1.1 and 1:1.2 ((18 ± 1.56 vs. 24.5 ± 1.56 displacements, respectively; F2 =
0.71, P = 0.02) and between 1:1.1 and 1:1.5 (27 ± 1.56 displacements for 1:1.5; F2 = 0.71, P =
0.01). The number of displacements observed after push up of feed was not different among the
three feed rail stocking rates (F2 = 6.09, P = 0.14). Stall stocking rate treatments also did not
impact activity or time spent ruminating (F2 = 1.39, P = 0.43 and F2 = 0.2, P = 0.84,
respectively). There was no difference among the treatments for the average number of cows
lying in stalls (F2 = 4.95, P = 0.17), standing in stalls (F2 = 11.99, P = 0.08), perching in stalls
(F2 = 0.28, P = 0.76), standing in the alley (F2 = 9.04, P = 0.1), or walking in the alley (F2 = 1.09,
P = 0.36). There was no difference among the treatments for number of cows eating (F18 = 0.5, P
= 0.93), with heads in the headlocks but not eating (F18 = 0.96, P = 0.53), standing at the feed rail
with heads not in headlocks (F18 = 0.31, P = 0.99), or standing in the alley behind the feed rail
(F18 = 0.72, P = 0.76). There were no differences among the treatments for number of cows
eating (F18 = 1.36, P = 0.27), with heads in the headlocks but not eating (F18 = 1.11, P = 0.42; F),
standing at the feed rail with heads not in headlocks (F18 = 1.14, P = 0.39), or standing in the
alley behind the feed rail (F18 = 0.91, P = 0.59). Additionally, activity and rumination results in
response to feed rail stocking rates showed no difference among treatments (F2 = 0.55, P = 0.65
and F2 = 0.12, P = 0.89, respectively). This study is the first to examine the impact of stocking
rate in stalls and at the feed rail on displacement and general behavior of lactating cow in an
AMS dairy. Our results suggest that increasing the stocking rate at either feed rail or stalls to
1:1.2 and 1:1.5 in AMS dairies may not affect cow behavior to the degree that these stocking
rates do in parlor-milked dairies. Results may help producers determine optimal stocking rates
for AMS dairies, allowing them to optimize facility design and management in ways that
maintain good welfare and production without wasting space or resources.

62

Key Words: robotic milking, stocking density, lying, displacement, rumination

63

INTRODUCTION

Stocking rate of stalls in parlor-milked dairy herds can have effects on cow behavior, health and
production, and there is evidence of a positive relationship between the stocking rate and lying
time at the stalls and milk production (Huzzey et al., 2013). Stocking rate is a significant element
of cow comfort, with particular importance related to its effect on lying time (von Keyserlingk et
al., 2009). Echoing this sentiment, the European Food Safety Authority stated that overstocking
is one of the most important risks to the welfare of cattle in free-stall housing (EFSA, 2009a).
However, while dairy producers are often advised to keep stocking density below 120% (e.g.,
NFACC, 2009), overstocking is commonly found on commercial dairy farms in the North
America (USDA, 2007; von Keyserlingk et al., 2012). This is true despite the fact
thatoverstocking can lead to lower milk production on farms with fewer stalls per cow (Bach et
al., 2008).

Current recommendations suggest that most cows in comfortable free stalls spend an average
about 12 h/d lying down (EFSA, 2009b). Although, animal welfare standards for dairy farms
have generally been related to resource-based criteria, such as stall stocking rate, there is
increased interest in assessing the consequences of these recommendations on the animals
themselves (Rushen et al., 2011). Some studies showed that increased time spent away from the
pen can also decrease lying time (Botheras, 2006; Gomez and Cook, 2010). Other studies have
shown the relationship between high stocking rate and reduced time of lying in stalls (Fregonesi
et al., 2007; Krawczel et al., 2012; Telezhenko et al., 2012). Some studies showed that increased
time spent away from the pen can also decrease lying time (Botheras, 2006; Gomez and Cook,

64

2010).in North America survey studies have shown average lying times of 10-11 h/d for highproducing lactating cows (Ito et at., 2009; Von Keyserlingk et at., 2013; Ito et al., 2014).

Reducing time cows spend lying down can lead to both behavioral and physiological stress
(Munksgaard and Simonsen, 1996). According to Fregonosi, and colleagues (2007) lying time
decreased 1.7 hr/per d, when stall stocking rate was 150% compared with a 100% stocking rate.
In addition, short lying times may also lead to lameness (Bell et al., 2009; Chapinal et al., 2009;
Cook et al., 2004; Proudfoot et al., 2010). However, Chapinal and colleagues did not prove that
overstocking was a main factor of lameness but effectively, there is a positive relationship
among lying time and lameness, where the healthy cows lie less than lame cows (Chapinal et al.,
2009; Chapinal, 2010a,b).

Stocking rate at the feed rail is another key element that must be considered to ensure cow
comfort and adequate feeding behavior in dairy farms. It is commonly recommended that each
cow have approximately 60 cm of feeding space at the feed rail (Grant and Albright, 2001),
although one study suggested that this recommendation is overly generous, as cows had similar
feed intake when given less feed space than recommended (Menzi and Chase, 1994). Generally,
however, feeding space available for each cow was related to the level of competition at the feed
rail (DeVries et al., 2004); and displacement at the feed rail increased when stocking rate was
increased at the feed rail (DeVries et al., 2004; Huzzey el at., 2006; Proudfoot et al., 2009).

65

Cows often prefer to feed at the same time, as they are social animals that naturally show
synchronized behavior, which is another important factor to consider when evaluating the impact
of space at the feed rail (Von Keyserlingk and Weary, 2010). For example, in the 90 min after
fresh feed delivery, feeding activity was greatest compared to the rest of the day (Huzzey el at.,
2013). This means cows are motivated to feed at the same time, and so ensuring that all animals
have the space to feed simultaneously will allow cows with less competitive ability (e.g., fresh or
primiparous cows) to feed with the herd.

Thus, competition at the feed rail affects feeding behavior and welfare as increased competition
at high feed rail stocking density could limit the ability of some cows to access to the feed rail at
feeding time this could translate into altered feeding behavior, reduced feed intake and lying time
and increased time spent standing (Grant and Albright, 2001; DeVries et al., 2004; Huzzey et al.,
2006; Collings et al., 2011; Krawczel et al., 2012). However, reducing feed bunk space has not
been directly linked with lower milk yield (Bach et al., 2008; Proudfoot et al., 2009).

At present little is known about how stocking rate affects cow behavior in herds milked with
automatic milking systems (AMS). Cows in AMS have free movement to and from the robot and
may milk asynchronously as opposed to cows milking as a single group in parlor-based systems.
On the other hand, stall stocking rate in AMS dairies may also be influenced by other factors not
present in parlor-milked dairies, such as the design of the barn and traffic system (forced, free, or
guided) and the number of cows per AMS. For example, according to Uetake and colleagues
(1997), the AMS milking and holding areas impact synchronization of cows eating and resting
and can reduce time spent eating relative to cows that are parlor milked. As return from the

66

milking parlor appears to be one of the factors that synchronizes feeding (DeVries et al., 2003), it
is possible that cows in AMS who are milking individually may also eat and lie down more
asynchronously compared to their parlor-milked counterparts. However, there is some evidence
to suggest that cows may still wish to lie down synchronously or are most motived to lie down at
certain parts of the day based on their circadian rhythms (Rushen et al., 2001; Tucker, 2009)

The objective of this research was to examine the effect of reducing the number of lying stalls
available per cow or the amount of available space at feed rail on cow behavior in an AMS dairy.
Cows should spend less total time lying and have shorter lying bouts when stalls were
overstocked. Further, more displacement of cows from stalls and the feed rail should occur at
higher stocking rates compared to a 1:1 stocking rate.

MATERIALS AND METHODS

Prior to the start of the study, all protocols were submitted to and approved by the Michigan
State University Institutional Animal Care and Use Committee.

Animals and Facility
The study was conducted at Kellogg Biological Station Pasture Dairy Research and Education
Center of Michigan State University. This study occurred during the winter months when the
cows were kept in the barn and not given access to pasture. This farm consisted of one barn with
two pens each with 60 cows, 58 free stalls, and 65 headlocks. No cows entered or left either pen

67

during the 12 wk of the study. Cows were balanced between two pens according to stage of
lactation (1st, 2nd, 3rd) and breed (i.e., Holstein or New Zealand Friesian). One AMS (Astronaut
A3, Lely, Maassluis, and The Netherlands) was present in each pen. A total mixed ration was
delivered twice a day at the feed rail at 04:00 and 12:00, with feed push-ups occurring four times
a day. In addition, cows received a concentrate mix in the AMS each time they were milked. The
amount of concentrate received over a 24-h period was based upon expected milk yield and stage
of lactation.

Treatments
Three stocking rate treatments applied to the 2 pens in a random order were used to alter the
number of free-stalls or feed space available per cow. The treatments were as follows: 1:1
treatment = 58 stalls or 60 headlocks for 60 cows; 1:1.2 treatment = 50 stalls or 50 headlocks for
60 cows; and 1:1.5 treatment = 40 stalls or 40 headlocks for 60 cows. When reducing the number
of stalls to change the stocking rate, stalls were blocked by hanging double-stranded chain
barriers across the rear of the stall to prevent cows from entering. When reducing the number of
headlocks to increase the stocking rates at the feed rail, headlocks were blocked by boarding
them up in order prevent cows for putting their heads in to the headlocks. The first six weeks
(12- wk experimental period) examined alterations to stall stocking rate and the subsequent six
weeks examined alterations to feed rail stocking rate. Data were collected during the last 2 d of
each weeklong treatment. A washout period of 1 wk was applied between treatments. During the
washout period, all stalls or headlocks in the pen were unblocked.

68

Data Recording
Two point-tilt-zoom cameras (HAD 480 PTZ Speed Dome SPE-CAMCCD55, Sony Super, and
Sony Corporation, Tokyo, Japan) and two-fixed angle cameras (1/3” Sony Super HAD Color
CCD 540, Sony Corporation, Tokyo, Japan) were used per pen to allow for recording of areas of
interest. Digital video was recorded using a 16-channel digital video recorder (G-Max 9000
Series, Skyway Security, Mauldin, SC). Individual cows were identified on video by matching
their unique spot patterns to a picture database containing photos of each cow’s head, rear, and
right and left sides. This method of identification has been used extensively to identify wild
animals (Peterson, 1972; Wursing and Jefferson, 1990; Kelly, 2001) and has been validated by
comparing accuracy of photo-identification against various forms of electronic identification
(Irvine et al., 1982; Scott et al., 1990).

Video was recorded during the last 2 d of each stall stocking rate treatment. Video was decoded
using instantaneous scan sampling to capture time budgets of cows in the stall area. Every 10
minutes during the morning (0400-1000), afternoon (1000-1600), and evening (1600-2200), a
scan sample was taken to record the number of cows in a pen that were lying, standing, walking
and perching in the free stall area. Lying and standing were further classified according to
whether they occurred in a stall or in the alley. All displacements from stalls in the last 2 d of
treatment were recorded. A displacement was defined as occurring when one cow (actor) came
into contact with another cow (reactor), resulting in the reactor leaving the stall.

69

At the feed rail, live observations were made using instantaneous scan sampling in the 45 min
after a feed delivery and in the 45 min after a push up. Every 5 min we recorded the number of
cows doing the following behaviors: cows eating at the feed rail, cows with heads in headlocks
but not eating, cows with heads out of headlocks but remaining at the rail, and cows standing in
the alley immediately behind the feed rail. All instances of displacement from the feed rail were
recorded during live observations.

Activity (ATV) and rumination (RU) data were collected using hardware built into AMS
identification tag worn by each cow (Lely Qwes-HR, Lely, Maassluis, The Netherlands). At
milking, data from these specific devices was read by the AMS and downloaded to the farm
computer. Data were collected and cleaned by Time for Cows (T4C; Lely, Maassluis, The
Netherlands), the management program associated with the AMS. In brief, ATV measures time
spent moving by the cow, specifically by measuring vertical movements of the cow’s head, such
as head bobs that occur during walking or during mounting (Ben Smink, Lely USA Inc.,
Madison, WI, personal communication). The ATV data were reported in “activity units.” Time
spent ruminating was recorded using a microphone built into the AMS to tag detect regurgitation
of a bolus for rumination (Ben Smink, personal communication). Both types of data were
recorded in 2-h blocks (e.g., 0600 - 0800, and 0800 - 1000 , so on), and the management
program deleted duplicate data to create a single data set of unique 2-h blocks. Previous work in
our lab has linked higher activity unit values to live observation of greater activity by cows as
well shown a correlation between live observation of rumination and values obtained from the
Lely Qwes- HR (Elischer et al., 2013).

70

Analysis
All analyses were completed using SAS (SAS version 9.4, and SAS Institute Inc., Cary, NC) and
significance was declared at P < 0.05. Treatments were applied to pens, all analysis were
conducted at the level of pen. The effect of stocking rate in the stalls was examined using video
recordings of the final 2 d of each stall stocking treatment. The effect of stocking rate at the feed
rail was examined using live observation of cow behavior during the 45 min after one push up
event and in the 45 min after one feed delivery event for each feed rail stocking rate treatment in
each pen.

Analyses of frequency of displacement at the stall and feed rail (at feed delivery and push up)
were conducted using a mixed model (PROC MIXED) after verifying that data were meet
assumption of normality and equality of variance. The model was as follows:
Yijkl = μ + PNi + TPj + Ck×GCBL +eijk
Where PNi is the random effect of pen, Ck is the fixed effect of treatments (1:1, 1:1.2, and 1:1.5),
TPj is the repeated measure of time period, GCBLis the covariate of (displacements and scanning
data of stall and feed rail), Ck×GCBL is the interaction between treatments and covariate, and eijk
is the residual. Counts of cows performing specific behaviors at the feed rail collected via
instantaneous scan sampling were analyzed using a mixed model (PROC MIXED) after
verifying that data met assumption of normality and equality of variance. Counts of cows
performing specific behaviors in the stall area in response to stall stocking rate treatments
collected via instantaneous scan sampling were analyzed using a mixed model (PROC MIXED)
for the average number of cows lying in stalls, standing in stalls, or standing in the alley. TukeyKramer tests were used as needed following mixed model analysis to determine which

71

treatments were different. However, the average number of cows perching in stalls and the
number of cows walking in the alley were analyzed with a Kruskal-Wallis test (PROC
NPAR1WAY) because the data were not normal. The model was as follows:
Yij1 = μ + Ti + Vj + eijk
Where Ti is the fixed effect of treatments, Vj is the covariate of (Stall scanning data), and eijk is
the residual.

RESULTS

Stall Stocking Rate
There was no effect of stocking rate treatment on number of displacements at the stalls (F2 =
5.22, P = 0.16; Figure 3.1). In addition, overstocking did not appear to affect the time budgets of
cows in the stall area (P < 0.05). In particular, there was no effect of treatment on average
number of cows lying in stalls (F2 = 4.95, P = 0.17; Figure 3.2), standing in stalls (F2 = 11.99, P
= 0.08; Figure 3.3), perching in stalls (F2 = 0.28, P = 0.76; Figure 3.4), standing in the alley (F2 =
9.04, P = 0.10; Figure 3.5), or walking in the alley (F2 = 1.09, P = 0.36; Figure 3.6). Stall
stocking rate treatments did not affect the activity level of cows (F2 = 1.39, P = 0.34; Figure 3.7)
or the time spent ruminating (F2 = 0.12, P = 0.89; Figure 3.8).

Feed Rail Stocking Rate
Overstocking at the feed rail led to differences in displacement at the feed rail during fresh feed
delivery between the treatment (F2 = 86.33, P = 0.01; Figure 3.9). Specially, more displacement
occurred in both overstocking treatments (1:1.2 = 24.5 ± 1.56 displacements; 1:1.5 = 27 ± 1.56

72

displacements) compared to the 1:1 treatment (18 ± 1.56 displacements; vs. 1:1.2: F2 = 0.71, P =
0.02; vs. 1:1.5: F2 = 0.71, P = 0.01). However, overstocking at the feed rail did not affect the
number of cows performing behaviors of interest at the feed rail. Specifically, there were no
differences among the treatments for number of cows eating (F18 = 0.5, P = 0.93; Figure 3.10A);
with heads in the headlocks but not eating (F18 = 0.96, P = 0.53; Figure 3.10B), standing at the
feed rail with heads out headlocks (F18 = 0.31=, P = 0.99; Figure 3.10C), or standing in the alley
behind the feed rail (F18 = 0.72, P = 0.76; Figure 3.10D).

Stocking rate did not affect number of displacements at the feed rail in response to push up of
feed (F2 = 9.06, P = 0.14; Figure 3.9). Additionally, there were no differences among the
treatments for number of cows eating (F18 = 1.36, P = 0.27; Figure 3.11A), with heads in the
headlocks but not eating (F18 = 1.11, P = 0.42; Figure 3.11B), standing at the feed rail with heads
out headlocks (F18 = 1.14, P = 0.39; Figure 3.11C), or standing in the alley behind the feed rail
(F18 = 0.91, P = 0.59; Figure 3.11D). Activity levels of cows were not affected by feed rail
stocking rates (F2 = 0.55, P = 0.65; Figure 3.7), nor was time spent ruminating (F2 = 0.2, P =
0.84; Figure 3.8).

DISCUSSION

Stall Stocking Rate
Typically, studies of parlor-milking dairies show more displacement of cows from stalls and less
lying time when cows are overstocked (Cook et al., 2005; Fregonesi et al., 2007). However,

73

when we analyzed similar data from an AMS dairy, we did not find significant differences in
frequency of displacement as a result of stall overstocking. In addition, time budget analysis via
instantaneous scan sampling of behavior in the stall area showed no significant difference
between the stall stocking rate treatments in number of cows lying, standing or perching in stalls
or the number of cows standing or walking in alleys. This suggests that the stocking rate
treatments used in this study did not impact cows’ use of stalls. Further, there was no effect of
stall stocking rate on activity or rumination as recorded by the AMS monitors. Suggesting that
the cows did not change the amount of time spent performing these behaviors as a result of
overstocking.

Our findings are unlike those of other studies, which have shown that stocking rates of 120% or
greater lead to an increase in standing time of 15 to 25% (Grant, 2001). Conversely, in previous
studies of parlor-milked cows, increased stocking rate at the stalls has led to decreased lying time
(Cook et al., 2005; Fregonesi et al., 2007). Therefore, our findings may suggest that cows milked
with AMS may adapt their behavior to lie more at other times of day. However, our data do not
suggest different patterns of stall use, at least not across the daylight hours; in response to
overstocking.

Feed Rail Stocking Rate
Displacement of cows from the feed rail has been found to increase when stocking rate is
increased at the feed rail (DeVries et al., 2004; Huzzey el at., 2006; Proudfoot et al., 2009). In
our study, there was a significant difference in number of displacements from the feed rail
between the 1:1 treatment and the two overstocked treatments after fresh feed delivery, while

74

there was no difference between 1:1.2 and 1:1.5 treatments. However, there were no significant
differences between the treatments in number of displacements from the feed rail following feed
push up. This may emphasize the synchronicity of cows’ response to fresh feed delivery and
competition between cows for access to fresh feed when any degree of feed rail overstocking
leads to displacement.

However, there were no differences among the treatments particularly with respect to the number
of cows eating but also with heads in headlocks but not eating, standing at the feed rail with
heads out of headlocks, or standing in the alley behind the feed rail. This suggests that the impact
of increasing stocking rate on cow feeding behavior was minimal. These findings are unlike
these from parlor studies that have shown increase in stocking rate at the feed bunk leads to
increased competition and changed feeding behavior (Collings et al., 2011). Particularly, our data
from the feed rail found no effect on eating behavior when the stocking rate increased, while
parlor-based studies have found feeding time decreased when the stocking rate at the feed rail
increased (Huzzey et al., 2006).

CONCLUSIONS

Stall and feed rail stocking rates for cows in AMS systems may have little direct impact on cow
behavior and lying time, unlike what is seen when parlor-milked farms are overstocked.
However, AMS dairies may still experience some effects of overstocking, but not to a greater
degree than in in parlor-milked systems such as in displacements; when fresh feed is deliverd.

75

This study showed there was no real impact on cow behavior and welfare when the stocking rate
increased either at the stalls or feed rail, which could mean that cows are adapting to stocking
rates by becoming more asynchronous. Overstocking in AMS may affect cows differently from
what is seen in the parlor dairies due to cow having control over their milking routines as well as
over feeding and lying times. Thus studies on cow behavior and welfare related to stocking rate
in parlor dairies may not be directly applicable to dairies milking with AMS.

ACKNOWLEDGMENTS

We would like to thank the Michigan State University Department of Animal Science for
supporting this research through an award from the Elwood Kirkpartrick Dairy Research Science
Endowment to Janice Siegford (East Lansing, MI). We are grateful for the support and
assistance of Howard Straub III (Hickory Corners, MI). Keri Dunn (East Lansing, MI) played a
vital role in collecting data on farm.

76

APPENDIX

77

Figure 3.1 Average counts of stall displacement (± SEM) by time period during the final 2 d of
each treatment for the three stocking rate treatments.

Stall Displacement
60

Avg. # of Displacements per 6 h

50

40

morning

30

afternoon
evening
20

10

0
1:1

1:1.2
Treatments

78

1:1.5

Figure 3.2 Average number of cows lying down in stalls (± SEM) presented by time period
during the final 2 d of each treatment for the three stocking rate treatments.

Cows Lying Down in Stalls
20
18

# of Cows per 6 h

16
14
12
Morning

10
8

Afternoon

6

Evening

4
2
0
1:1

1:1.2

1:1.5

Treatments

79

Figure 3.3 Average number of cows standing in stalls (± SEM) presented by time period during
the final 2 d of each treatment for the three stocking rate treatments.

Cows Standing in Stall
7
6

# of Cows per 6 h

5
4
Morning
3

Afternoon
Evening

2
1
0
1:1

1:1.2

1:1.5

Treatments

80

Figure 3.4 Average number of cows perching in stalls (± SEM) presented by time period during
the final 2 d of each treatment for the three stocking rate treatments.

Cows Perching in Stalls
7
6

# of Cows per 6 h

5
4
Morning
3

Afternoon
Evening

2
1
0
1:1

1:1.2

1:1.5

Treatments

81

Figure 3.5 Average numbers of cows standing in alley (± SEM) presented by time period during
the final 2 d of each treatment for the three stocking rate treatments.

Cows Standing in Alley
2.5

# of Cows per 6 h

2

1.5
Morning
Afternoon

1

Evening
0.5

0
1:1

1:1.2

1:1.5

Treatments

82

Figure 3.6 Average number of cows walking in the alley (± SEM) presented by time period
during the final 2 d of each treatment for the three stocking rate treatments.

Cows Walking in Alley
4
3.5
3

# of Cows per 6 h

2.5
2
Morning
1.5

Afternoon

1

Evening

0.5
0
1:1

1:1.2

1:1.5

-0.5
-1

Treatments

83

Figure 3.7 Cow activity in response to stocking rate during the final 2 d of each stocking rate
treatment.

Cow Activity
29

Activity Units per Day

28
27
26
Stall

25

Feed
24
23
22
1:1

1:1.2

1:1.5

Treatments

84

Figure 3.8 Descriptive data showing the time cows spent ruminating during the final 2 d of
stocking rate treatments at stalls and feed rail.

Cows Rumination

Hours of Rumination per Day

7.4
7.3
7.2
7.1
Stall

7

Feed
6.9
6.8
6.7
1:1

1:1.2
Treatments

85

1:1.5

Figure 3.9 Average number of displacements (± SEM) was occurring during feed delivery and
push up at the feed rail during the final 2 d of each stocking rate treatment.

Displacement
40

Avg..# of Displacements

35
30
25
20

Feed delivery

15

Push up

10
5
0
1:1

1:1.2

1:1.5

Treatments

86

Figure 3.10A. Descriptive data showing number of cows eating at the feed rail for each treatment
after delivery of fresh feed. Numbers are presented as total number of cows eating along the
entire length of the feed rail (averaged between the 2 pens).

Eating after Feed Delivery

12

# of Cows

10
8

1:1.5

6

1:1.2
4

1:1

2
0
0

5

10

15

20

25

Time (min)

87

30

35

40

45

Figure 3.10B. Descriptive data showing number of cow with heads in headlocks at the feed rail
for each treatment after delivery of fresh feed. Numbers are presented as total number of cows
eating along the entire length of the feed rail (averaged between the 2 pens).

Heads in Headlocks after Feed Delivery
14
12

# of Cows

10
8
1:1.5
6

1:1.2
1:1

4
2
0

0

5

10

15

20

25

Time (min)

88

30

35

40

45

Figure 3.10C. Descriptive data showing number of cow with heads out of headlocks but standing
at the feed rail for each treatment after delivery of fresh feed. Numbers are presented as total
number of cows eating along the entire length of the feed rail (averaged between the 2 pens).

Heads out of Headlocks after Feed Delivery
6

# of Cows

5
4
1:1.5

3

1:1.2
2

1:1

1
0

0

5

10

15

20

25

Time (min)

89

30

35

40

45

Figure 3.10D. Descriptive data showing number of cow standing behind the feed rail for each
treatment after delivery of fresh feed. Numbers are presented as total number of cows eating
along the entire length of the feed rail (averaged between the 2 pens).

# of Cows

Behind Feed Rail after Feed Delivery

10
9
8
7
6
5
4
3
2
1
0

1:1.5
1:1.2
1:1

0

5

10

15

20

25

30

Time (min)

90

35

40

45

Figure 3.11A. Descriptive data showing number of cows eating at the feed rail for each treatment
after push up of feed.. Numbers are presented as total number of cows eating along the entire
length of the feed rail (averaged between the 2 pens).

Eating after Push Up
8
7
# of Cows

6
5
4

1:1.5

3

1:1.2

2

1:1

1
0
0

5

10

15

20

25

Time (min)

91

30

35

40

45

Figure 3.11B. Descriptive data showing number of cow with heads in headlocks at the feed rail
for each treatment after push up of feed. Numbers are presented as total number of cows with
head in headlocks but not eating along the entire length of the feed rail (averaged between the 2
pens).

Heads in Headlocks after Push Up
8
7

# of Cows

6
5
4

1:1.5

3

1:1.2
1:1

2
1
0
0

5

10

15

20

25

Time (min)

92

30

35

40

45

Figure 3.11C. Descriptive data showing number of cow with heads out of headlocks but standing
at the feed rail for each treatment after push up of feed. Numbers are presented as total number
of cows with heads out of headlocks but standing at the feed rail along the entire feed rail
(averaged between the 2 pens).

Heads out Headlocks after Push Up
5
4.5

4
# of Cows

3.5
3
2.5

1:1.5

2

1:1.2

1.5

1:1

1
0.5
0
0

5

10

15

20

25

Time (min)

93

30

35

40

45

Figure 3.11D. Descriptive data showing number of cow standing behind the feed rail for each
treatment after push up of feed.. Numbers are presented as total number of cows standing behind
the feed rail along the entire feed rail (averaged between the 2 pens).

# of Cows

Behind Headlocks after Push Up

8
7
6
5
4
3
2
1
0

1:1.5
1:1.2
1:1

0

5

10

15

20

25

30

Time (min)

94

35

40

45

REFERENCES

95

REFERENCES

Bach, A., N. Valls, A. Solans, and T. Torrent. 2008. Associations between non-dietary factors
and dairy herd performance. J. Dairy Sci. 91:3259–3267.

Bell, N. J., M. J. Bell, T. G. Knowles, H. R. Whay, D. J. Main, and A. J. F. Webster. 2009. The
development, implementation and testing of a lameness control programme based on HACCP
principles and designed for heifers on dairy farms. Vet. J. 180:178–188.

Botheras, N. A. 2006. The behaviour and welfare of grazing dairy cows (Bos taurus): Effects of
time away from pasture and position in the milking order. PhD Thesis. Univ. of Melbourne,
Melbourne, Australia.

Chapinal, N., A. M. de Passillé, D. M. Weary, M. A. G. von Keyserlingk, and J. Rushen. 2009.
Using gait score, walking speed, and lying behavior to detect hoof lesions in dairy cows. J. Dairy
Sci.92:4365–4374.

Chapinal, N., A. M. de Passillé, J. Rushen, and S. Wagner. 2010a. Automated methods for the
detection of lameness and analgesia in dairy cattle. J. Dairy Sci. 93:2007–2013.

Chapinal, N., A. M. de Passillé, J. Rushen, and S. Wagner. 2010b. Effect of analgesia during
hoof trimming on gait, weight distribution and activity of dairy cattle. J. Dairy Sci. 93:3039–
3046.

Blowey, R., and P. Edmondson. 1995. Mastitis Control in Dairy Herds. An Illustrated and
Practical Guide. Farming Press, Ipswich, UK. L.K.M.

Collings, D.M. Weary, N. Chapinal, and M.A.G. von Keyserlingk. Temporal feed restriction and
overstocking increase competition for feed by dairy cattle. J. Dairy Sci. 94:5480–5486.

Cook, N. B., T. B. Bennett, and K. V. Nordlund. 2005. Monitoring indices of cow comfort in a
free-stall-housed dairy herds. J. Dairy Sci. 88:3876–3885.

96

Cook, N. B., T. B. Bennett, and K. V. Nordlund. 2004. Effect of free stall surface on daily
activity patterns in dairy cows with relevance to lameness prevalence. J. Dairy Sci. 87:2912–
2922.

National Farm Animal Care Council (NFACC). 2009. Code of practice for care and handling of
Dairy Cattle. Dairy Farmers of Canada, Ottawa, ON, Canada.

DeVries, T. J., M. A. G. von Keyserlingk. 2005. Time of feed delivery affects the feeding and
lying patterns of dairy cows. J. Dairy Sci. 88:625-631.

DeVries, T. J., M. A. G. von Keyserlingk, and K. A. Beauchemin. 2003. Diurnal feeding pattern
of lactating dairy cows. J. Dairy Sci. 86:4079-4082.

DeVries, T. J., M. A. G. von Keyserlingk, and D. M. Weary. 2004. Effect of feeding space on the
inter-cow distance, aggression, and feeding behavior of free-stall housed lactating dairy cows. J.
Dairy Sci. 2004; 87: 1432–1438.

EFSA (European Food Safety Authority). 2009a. Scientific Opinion of the Panel on Animal
Health and Welfare on a request from the Commission on the risk assessment of the impact of
housing, nutrition and feeding, management and genetic selection on behaviour, fear and pain in
dairy cows. EFSA J. 1139:1–68. EFSA, Parma, Italy.

EFSA (European Food Safety Authority). 2009b. Scientific Opinion of the Panel on Animal
Health and Welfare on a request from the Commission on the risk assessment of the impact of
housing, nutrition and feeding, management and genetic selection on metabolic and reproductive
problems in dairy cows. The EFSA J. 1140:1–75. EFSA, Parma, Italy.

Elischer, M. F., Arceo, M. E., Karcher, E. L., Siegford J. M. 2013. Validating the accuracy of
activity and rumination monitor data from dairy cows housed in a pasture-based automatic
milking system. J. Dairy Sci. 96:6412–6422.

Fregonesi, J. A., C. B. Tucker, and D. M. Weary. 2007. Overstocking reduces lying time in dairy
cows. J. Dairy Sci. 90: 3349–3354.

Gomez, A., and N. B. Cook. 2010. Time budgets of lactating dairy cattle in commercial freestall
herds. J. Dairy Sci. 93:5772–5781.

97

Grant, R. J. and J. L. Albright. 2001. Effect of animal grouping on feeding behavior and intake of
dairy cattle. J. Dairy Sci. 84: E156–E163.

Huzzey, J. M., T. J. DeVries, P. Valois, and M. A. G. von Keyserlingk. 2006. Stocking density
and feed barrier design affect the feeding and social behavior of dairy cattle. J. Dairy Sci.
89:126-133.

Huzzey, J. M., T.R. Overton, and M. A. G. Von Keyserlingk. 2013. Effect of stocking density
on feeding strategies and health. Pages 51-62 in Proceedings of the Tri-State Dairy Nutrition
Conference. Fort Wayne, IN.

Irvin, A. B., R. S. Wells, and M. D.Scott. 1982. An evaluation of techniques for tagging small
odontocete cetaceans. Fish. Bull. 80:135-143.

Ito, K., D. M. Weary, and M. A. G. von Keyserlingk. 2009. Lying behavior: Assessing withinand between-herd variation in free-stallhoused dairy cows. J. Dairy Sci. 92:4412–4420.

Ito, K., N. Chapinal, D. M. Weary, and M. A. G. von Keyserlingk. 2014. Associations between
herd-level factors and lying behavior of freestall-housed dairy cows. J. Dairy Sci. 97:2081-2089.

Kelly, M. J. 2001. Computer-aided photograph matching in studies using individual
identification : an example from Serengeti cheetahs. J. Mammal. 82:440-449.

Krawczel, P. D., L. B. Klaiber, R. E. Butzler, L. M. Klaiber, H. M. Dann, C. S. Mooney, and R.
J. Grant. 2012. Short-term increases in stocking density affect the lying and social behavior, but
not the productivity, of lactating Holstein dairy cows. J. Dairy Sci.95:4298–4308.

Menzi, W. Jr. and Chase, L. E. 1994. Feeding behavior of cows housed in free stall barns. Pages
829–831 in Dairy Systems for the 21st Century. Am. Soc Agric. Eng., St. Joseph MI.

Munksgaard, L., and H. B. Simonsen. 1996. Behavioral and pituitary adrenal-axis responses of
dairy cows to social isolation and deprivation of lying down. J. Anim. Sci. 74:769–778.
Peterson, J. C. 1972. An identification system for zebra. East Afr. Wildl. J. 10:59-63.

98

Proudfoot, K. L., D. M. Veira, D. M. Weary, and M. A. G. von Keyserlingk. 2009. Competition
at the feed bunk changes the feeding, standing, and social behavior of transition dairy cows. J.
Dairy Sci. 92:3116-3123.

Proudfoot, K. L., D. M. Weary, and M. A. G. von Keyserlingk. 2010. Behavior during transition
differs for cows diagnosed with claw horn lesions in mid lactation. J. Dairy Sci. 93:3970–3978.

Rushen J., L. Munksgaard, P. G. Marnet, A. M. dePasillé. 2001. Human contact and the effects
of acute stress cows at milking. Apple. Anim. Behav. Sci.73: 1–14.

Rushen, J., A. Butterworth, and J. C. Swanson. 2011. Animal Behavior and Well-Being
Symposium: Farm animal welfare assurance: Science and application. J. Anim. Sci. 89:1219–
1228.

Scott, M. D., R. S. Wells, A. B. Irvine, and B. R. Mate. 1990. Tagging and marking studies on
small cetaceans. Pages 489-514 in The Bottlenose Dolphin. S. Leatherwood and R. R. Reeves,
ed. Academic Press, San Diego, CA.

Telezhenko, E., M. A. G. von Keyserlingk, A. Talebi, and D. M. Weary. 2012. Effect of pen
size, group size, and stocking density on activity in freestall-housed dairy cows. J. Dairy Sci.
95:3064–3069.

Tucker, C. B. 2009. Behaviour of cattle. Pages 151-160 in The Ethology of Domestic Animals.
P. Jensen, ed. CABI, Oxfordshire, UK.

USDA. 2007. Facility characteristics and cow comfort on U.S. dairy operations. No. 524.1210.
USDA-Animal and Plant Health Inspection Service (APHIS)-Veterinary Services (VS), Centers
for Epidemiology and Animal Health (CEAH), Fort Collins, CO.

Uetake, K., J.F. Hurnik, and L. Johnson.1997. Behavioral pattern of dairy cows milked in a
two-stall automatic milking system with a holding area. J. Anim. Sci. 75:954-958.

von Keyserlingk, M. A. G., A. Barrientos, K. Ito, E. Galo, and D. M. Weary. 2012.
Benchmarking cow comfort on North American freestall dairies: Lameness, leg injuries, lying
time, facility design, and management for high-producing Holstein dairy cows. J. Dairy
Sci. 95:7399–7408.

99

von Keyserlingk, M. A. G., J. Rushen, A. M. de Passillé, and D. M. Weary. 2009. Invited
review: The welfare of dairy cattle— Key concepts and the role of science. J. Dairy Sci.
92:4101–4111.

von Keyserlingk, M. A. G, and D. M. Weary. 2010. Review: feeding behavior of dairy cattle:
Measures and applications. Can. J. Dairy Sci. 90:303-309.

Wursing, B., and T. A. Jefferson. 1990. Methods of photo-identification for small cetaceans.
Rep. Int. Whal. Commn. 12:43-52.

100

CHAPTER 4
SUMMARY AND CONCLUSIONS

SUMMARY
Survey Stocking Rates at AMS Dairies
In first study presented in this thesis, few relationships were found between stocking rates (at
stalls, the feed rail or the AMS) and variables of interest including lying behavior or measures of
lameness. In fact, only leg cleanliness was significantly correlated with stall stocking rate. This is
in contrast to what has generally been observed in parlor-milked systems where increases in the
stocking rate at stalls and feed rail to a level greater than 100–113% have been found to lead to
changes in behavior of lactating dairy cows, especially with regard to lying time (Christopher et
al., 2008).

Stall stocking rate observed at the surveyed farms ranged from 56 to 126%, with an average of
94% (i.e., less than one cow per stall); while stocking rate at the feed rail ranged from 0.4 to 1.92
cows per 60 cm of feed rail with an average of 1.07 cows/60 cm (i.e., 56 cm of feeding space
per cow). Stocking rate at the AMS ranged from 36 to 72 cows per AMS with an average
stocking rate of 55±8.6 cows per AMS. Thus, in all cases, the average stocking rates were at or
below recommendations of various codes of practice and scientific studies (e.g., NFACC, 2009;
Anderson, 2008; Armstrong, 1994; Brouk, 2008; Collier et al., 2007). There were no cases of
extremely high stall stocking rate and only four cases when an AMS was stocked with more than
71 cows. However, 20% of the surveyed farms provided less than 45 cm of feed rail space per
cow, which could be expected to cause reduced feed intake and thus lower body condition
101

scores. It was surprising then that our results found no effect of high feed rail stocking rate on
body condition, measures of lamenes, or gait scores (Grant, 2009).

Cows synchronize their behavior as a result of circadian rhythms and group dynamics, with the
result that reducing the number of stalls may lead to less lying and increased standing (Fregonesi
et al., 2007), and may also impact cows’ feeding or milking behavior. Stocking rate at the stalls,
feed rail, and AMS can each affect cow health and lying time individually, and it is also possible
that effects of the various stocking rates may interact within a dairy. Thus, we examined the
correlation between stall, feed rail, and AMS stocking rates to better understand whether they
were related, which would have impacted how we assessed the effects of these three stocking
rates on variables of interest such as lying behavior, body condition score, measures of lameness
(gait scores), or cow cleanliness. However, we found no relationship between the stocking rates
in our study, which may mean these variables have stronger individual rather than interactive
effects on the cow behavior and welfare.

On average, cows spend between 8-12 hours a day lying when the stocking rate in free stalls is
1:1 (Tucker, 2009). Further, lying time decreased 1.7 h/day when stalls were stocked at 150%
compared with a 100% stocking rate in parlor-milked farms (Fregonosi et al., 2007).
Specifically, in the overnight hours and during the mid-day period, lying time decreased as a
result of such stall overstocking (Fregonosi et al., 2007). However, in the present study, we
found no relationship between stocking rate at the feed rail, stalls, or AMS with lying time, the
number of lying bouts or lying bout duration.

102

According to our analysis, there was no relationship between gait scores and the stocking rates at
the 39 farms that were examined. Specifically, the data showed no relationship between presence
of a limp, step asymmetry or head bob and the stocking rate at the stalls, feed rails or AMS at the
surveyed farms. According to Galindo and Broom (2000), cows with less access to stalls and
feed rail had increased standing time, which increased their risk of becoming lame. However, it
is important to note that there is also a positive relationship between lameness and lying time,
where lame cows lie more than healthy cows (Chapinal et al., 2009; Chapinal, 2010a,b).

The average body condition score of cows in the first study was 2.88 (on a 5 point scale). The
ideal body condition score for a dairy cow is 2.50-3.25 for the first third of lactation, the stage
targeted for focal cows in our study (NFACC, 2009). We found that neither stocking rate at the
feed rail or AMS had an effect on BCS. Given that we did not see an impact of high feed rail
stocking rates on BCS in the farms we surveyed, it was possible that managers had increased
frequency of feeding or feed push up frequency to compensate. Therefore, we analyzed the data
to evaluate relationships between stocking rate at the feed rail and body condition score,
frequency of fresh feed delivery, or push up of feed to the rail. We did not find any significant
relationships between these variables; however, producers were typically delivery fresh feed or
pushing up feed more than three times per day. This feeding management resulted in over 90%
of cows having feed in front of them during all of our observations at each farm. Thus, good
feeding management may have been compensating for high feed rail stocking rates at the eight
farms with limited feed rail space. Another possibility is that the cows at farms with high feed
rail stocking rates could be feeding asynchronously or adapting their feeding to times when the

103

rail is less crowded. It is also important to remember that cows in AMS receive concentrate in
the AMS; which may mitigate the impacts of high feed rail stocking rate.

According to Krawczel and colleagues (2012) there is no relationship between udder and leg
hygiene and stocking rate. However, other studies have reported that lower ranking cows are
more likely to lie down outside of free stalls when there is overstocking (Gonzalez et al., 2003)
and might be expected to be dirtier. In our study, stocking rate at the stall and feed rail did not
correlate with udder and flank cleanliness, which may mean the producers were keeping the stall
bedding clean and dry. Leg cleanliness, however, showed a significant correlation with stall
stocking rate, but there was no relationship of leg cleanliness with feed rail stocking rate; perhaps
indicating that time spent scraping alleys has an effect on leg cleanliness.

Experimental Overstocking at AMS Dairy
In second study presented in this thesis, we examined the effect of reducing the number of stalls
available or the feeding space per cow on displacement and other behaviors in an AMS dairy
farm. Each of two pens contained 60 cows/pen and 58 free-stalls and 60 headlocks. The impact
of stocking rate was examined using three treatments: 1) 1:1 treatment with 58 stalls or 60
headlocks available for 60 cows; 2) 1:1.2 treatments with 50 stalls or 50 headlocks available for
60 cows; and, 3) 1:1.5 treatment with 40 stalls or 40 headlocks available for 60 cows. Each stall
or feed rail treatment was applied separately, and each was applied for 1wk in a randomized
order that was different for each pen, with a 1-wk washout period between treatments.

104

Stall Stocking Rate
Typically, studies of parlor-milking dairies show more displacement of cows from stalls and less
lying time when cows are overstocked (Cook et al., 2005; Fregonesi et al., 2007). However,
when we analyzed our data from an AMS dairy, we did not find significant differences in
frequency of displacement as a result of stall overstocking. In addition, time budget analysis via
instantaneous scan sampling of behavior in the stall area showed no significant difference
between the stall stocking rate treatments in number of cows lying, standing or perching in stalls
or the number of cows standing or walking in alleys. This suggests that the stocking rate
treatments used in this study did not impact cows’ use of stalls. Further, there was no effect of
stall stocking rate on activity or rumination as recorded by the AMS monitors. This may mean
that the cows did not change these behaviors as a result of overstocking.

Our findings are unlike those of other studies, which have shown that stocking rates of 120% or
greater in parlor-milked dairies lead to an increase in standing time of 15 to 25% (Grant, 2001).
Also, in previous studies of parlor-milked cows, increased stocking rate at the stalls has led to
decreased lying time (Cook et al., 2005; Fregonesi et al., 2007). Therefore, our findings may
suggest that cows milked with AMS may adapt their behavior to lie more at other times of day.

105

Feed Rail Stocking Rate
Displacement of cows from the feed rail has been found to increase when stocking rate is
increased at the feed rail (DeVries et al., 2004; Huzzey el at., 2006; Proudfoot et al., 2009). In
our study, there was a significant difference in number of displacements from the feed rail
between the 1:1 treatment and the two overstocked treatments after fresh feed delivery, while
there was no difference between the 1:1.2 and 1:1.5 treatments. However, there were no
significant differences between any treatments in number of displacements from the feed rail
following feed push up. This may emphasize the synchronicity of cows’ response to fresh feed
delivery and competition between cows for access to fresh feed, where any degree of feed rail
overstocking leads to displacement.

No differences were seen among the treatments particularly with respect to the number of cows
eating but also, with heads in headlocks but not eating, standing at the feed rail with heads out of
headlocks, or standing in the alley behind the feed rail. This suggests that the impact of
increasing stocking rate on cow feeding behavior was minimal. These findings are unlike these
from parlor studies which have shown that increase in stocking rate at the feed bunk leads to
increased competition and changed feeding behavior (Collings et al., 2011). particularly on our
farm data from the feed rail stocking rate has no effect on the eating behavior when the stocking
rate increased, while other parlor studies the feeding time decreased when the stocking rate at the
feed rail increased (Huzzey et al., 2006).

106

CONCLUSIONS
In our first study, we focused on examining cow behavior and welfare rather than on milk
production. Our data showed no strong relationships between stocking rates and lying behavior,
body condition, measures of lameness (gait scores), or cleanliness scores. However, cows in
AMS have free movement to and from the AMS and may milk asynchronously as opposed to
cows milking as groups in parlor-based systems; this freedom may result in less synchrony of
feeding and lying behavior. Also, the stocking rate in AMS systems may also be influenced by
the design of barn and traffic system (forced, free, or guided). Cows are also feed concentrate in
the AMS, while milking; which may mitigate high stocking rates at the feed rail. Typically
producers who adopt AMS alter management routines in their entiretyspending less time milking
cows more time devoted to cow health, feeding or breeding routines. This may allow producers
to make management changes that can compensate for any issues caused by high stocking rates.
For example, delivering or pushing up feed frequently to provide a constant source of feed for
cattle when little feed rail space is available may reduce competition and mitigate any impacts on
body condition score.

The second study in this thesis suggests that stall and feed rail stocking rates for cows in AMS
systems may have little direct impact on cow behavior and lying time, unlike what is seen when
parlor-milked farms are overstocked. However, AMS dairies may still experience some effects of
overstocking, but not to a greater degree than in parlor-milked systems such as in displacement;
when fresh feed is delivered. This study showed there was no real impact on cow behavior and

107

welfare when the stocking rate increased either at the stalls or feed rail, which could mean that
cows are adapting to stocking rates by becoming more asynchronous.

Overstocking in AMS may affect cows differently from what is seen in the parlor dairies due to
cow having control over their milking routines as well as over feeding and lying times. Thus
studies on cow behavior and welfare related to stocking rate in parlor dairies may not be directly
applicable to dairies milking with AMS.

Finally, both studies in the thesis looked at stocking rates at the stalls and feed rail in AMS in
comparison with parlor system. However, these two studies indicate there may be less effect of
stocking rates on cow behavior and welfare at AMS dairies, which means producers could
increase number of cows in their barns without impacting cow welfare in a negative way.

108