Graduate Student Literature Review: What is known about the eliminative behaviors of dairy cattle?*

The eliminative behaviors of dairy cattle include frequencies and distribution over time and space for defecations and urinations, how the animal responds to cow-related and environmental factors by way of altered patterns of defecation and urination, and how an animal carries out and responds to its own acts of elimination. This review discusses the available literature to first define and describe eliminative behaviors of dairy cattle; what follows is a discussion on what can affect eliminative behaviors and methods for managing them. Information regarding these behaviors is sparse for dairy cattle and is largely centered around frequencies and distributions over the day. Relationships exist between eliminative behaviors and activity levels of the animals and activity levels of the people who manage them, suggesting that types of housing systems play a key role in mainly where and when eliminations occur. It also seems that individual animals vary in their elimination frequencies, in which case it may be interesting to determine what aspects of their individuality contribute to these differences. Although aspects of housing are intended to separate animals from their excreta, stalls or cubicles are not necessarily designed with cattle’s natural eliminative behaviors in mind. Re-fining the timing of management routines and training of animals are some options in the next steps toward managing eliminative behaviors.


INTRODUCTION
Eliminative or elimination behaviors include those surrounding defecation and urination. Generally, an eliminative behavior can be defined as "a behavior associated with the elimination of feces and urine from the body" (Semantic Scholar, 2020). The eliminative behaviors of dairy cattle have been described (e.g., Hafez, 1969;Phillips, 2002) and studied (e.g., Brantas, 1968;Aland et al., 2002), but not to the level of detail seen in other domesticated species (e.g., equine textbook: Beaver, 2019; literature review for pigs: Andersen et al., 2020). Research papers focusing on dairy cattle do not appear to provide specific definitions of this group of behaviors, and some only refer to specific components such as observed frequencies of defecations and urinations in their use of the phrase "eliminative behavior." Following the research presently available, this literature review defines eliminative behavior in the context of dairy cattle to include 3 aspects: (1) frequencies and distribution over time and space for defecations and urinations, (2) response to cow-related and environmental factors by way of altered patterns of defecation and urination, and (3) how an animal carries out and responds to its own acts of elimination. Eliminative behaviors are an element of nutritional behavior and relate to body maintenance. Furthermore, they are considered an element of behavioral and physiological needs (Phillips, 2002).
Studying the eliminative behaviors of dairy cattle is important in the context of a husbandry system, one aspect being to improve welfare. For example, a better understanding of how and when eliminations occur might help in improving cleaning routines to manage excreta. It is known that unhygienic environments are risk factors of hoof health issues such as interdigital dermatitis and heel horn erosion, and, in some cases, these environments can exacerbate leg injuries originally associated with uncomfortable stall surfaces, leading to tarsal or carpal bursitis and cellulitis (Bergsten, 2001). High hygiene scores (indicating very dirty) of the udder, lower legs, and composite scores of these areas have also been found to be associated with high somatic cell scores for individual cows, which increases the risk of mastitis (Reneau et al., 2005, using a scoring system developed by modifying the cleanliness scoring Graduate Student Literature Review: What is known about the eliminative behaviors of dairy cattle?* system developed by Chiappini et al., 1994). Other welfare-related outcome measures, such as duration of lying time on uncovered woodchip stand-off pads (O'Connor et al., 2019) and in covered woodchip pens (Schütz et al., 2019), are reduced by the presence of excreta. Ultimately, knowledge of eliminative behaviors has implications in human health and environmental impacts, one example being the emission of ammonia, which is an air pollutant and contributes to eutrophication (Hristov et al., 2011).
The purpose of this literature review is to discuss the current available information related to the eliminative behaviors of dairy cattle, divided into 3 sections: outlining what is known about eliminative behavior, describing what factors affect eliminative behavior, and presenting some methods for and attempts at controlling eliminative behavior. This review aims to examine the literature surrounding dairy breeds specifically. The terms "eliminate" and "eliminations" will be used to refer to defecation and urination events collectively, unless specified in certain studies. The present review was organized around the specific research questions mentioned above, and the literature was searched using McGill University library holdings in conjunction with WorldCat Discovery (https: / / mcgill .on .worldcat .org/ discovery) and Google Scholar (https: / / scholar .google .com/ ) to answer each question. Dates were not narrowed to a specific range to increase the total number of references, as limited literature has been published on this topic.
Because this review does not involve live subjects, IACUC and IRB permissions were not required.

Describing Eliminative Behaviors
Dairy cattle are large bovine ruminants (Hall, 2002). The logical first step would be to examine the eliminative behavior of feral cattle to understand behaviors that are usually displayed, without human influence. Despite the existence of several established feral cattle herds, studies on feral groups of dairy breeds specifically do not appear to exist in the literature, with the exception of a paper on Kerry cows in semiferal conditions, examining grazing patterns (Linnane et al., 2001). Eliminative behaviors in the absence of constraints such as housing involve grazing, studies of which are usually associated with beef breeds and focus on elimination frequencies and distributions for the purposes of nutrient management (e.g., Auerswald et al., 2010;Orr et al., 2012). Extensive pasture systems offer the least restrictive environment for dairy breeds, and their relationship to eliminative behaviors will be further discussed later in this review.
Defecation and urination are natural bodily functions, and animals have natural rhythms that they follow. In a review of 22 studies on overall behavior at pasture, Kilgour (2012) reveals that cattle in general are diurnal, meaning that most activities occur during the day; more specifically, grazing activity peaks during dawn and dusk hours. It follows that resting and rumination occur primarily at night. Outside of these typical activity patterns, it has been suggested that cattle are not selective about where and when they eliminate, nor are they are naturally aware of when they are about to eliminate (Hafez, 1969). It has also been presumed that eliminations are not voluntarily controlled (Hafez, 1969). Following this notion, Whistance et al. (2007) described how other behaviors, such as feeding and transitioning from loafing to feeding, were observed to be performed while defecating. To the author's knowledge, studies attempting to describe pre-eliminative behaviors in cattle (e.g., sniffing to select an area before eliminating) do not exist. Cattle do not seem to exhibit the latrine behavior of repeatedly eliminating in the same location (Whistance et al., 2009), whereas other animals such as pigs eliminate away from resting areas (as reviewed by Andersen et al., 2020), and domesticated horses use defined latrine areas (Beaver, 2019). However, cows do communicate via their urine, apparently increasing the frequency of urinations during estrus to broadcast their status (Phillips, 2002). In perhaps a more involuntary form of communication, urine has also been observed to express negative emotional state. Boissy et al. (1998) found that heifers presented with urine from a stressed companion had a longer latency to feed in a novel environment compared with when presented with urine from a nonstressed companion; similarly, heifers had a longer latency to explore a novel object in a familiar setting if said object was sprayed with urine from a stressed companion compared with urine from a nonstressed companion.
The posture assumed during elimination includes an arched back and raised tail, with the hind legs placed apart. During urination, the back is noticeably more arched (Aland et al., 2002). Dairy cows at pasture most frequently move away from their feces immediately after defecating, whether that be apparently purposefully (e.g., exhibiting feces avoidance) or just by coincidence (e.g., during the process of moving to another area while grazing) (Whistance et al., 2011). However, studies on behaviors following urination have yet to be conducted. The long-standing notion is that cattle do not appear to mind bodily contact with their own feces and urine (Brantas, 1968), nor the excreta of their herd mates (Hafez 1969). This perception appears contradic-  (Michel, 1955), and their preference for clean and dry lying areas over those dirty with manure and those wet with water (Schütz et al., 2019). Both examples involve allowing cattle the opportunity to exhibit feces avoidance, which reasons that an animal paying no mind to manure or urine might be a result of habituation to an environment that consistently entails contact with excreta.

Elimination Frequencies of Dairy Cattle
A comparison of the frequencies of eliminations for dairy cows is shown in Table 1. These articles were selected because they provide data obtained via live or video observation, as opposed to estimation by prediction and after-the-fact methods such as mapping excreta patches. Methodologies aside from live or video observations tend to be in the context of nutrient management under the term "excretion" for defecations and urinations. Few studies covering an entire 24 h of observations for elimination frequencies exist for dairy breeds. Variation exists within these studies, but such differences may be attributed to a combination of several factors. However, Aland et al. (2002) determined that the elimination frequencies of the cows of Fuller's study (data from the 1920s, data published in Aland et al., 2002) and their current observations (data from the 1990s) were similar, even though the latter cows' milk production was approximately double. Other main points to conclude are that cows appear to consistently defecate more frequently than they urinate in a 24-h period, but the numbers of events vary between studies ( Table 1).

Amounts of Excreta Produced by Cattle
Studies including measured volume amounts of excreta for cattle from a behavioral perspective are rare. Per 24 h, cattle are estimated to produce 40 to 50 kg of feces and 15 to 25 L of urine. This is equivalent to an average animal under normal conditions producing approximately 6% to 7% of its body weight in excreta (Acatincăi et al., 2011). Fuller's study (data published in Brantas, 1968) showed that 4 lactating Holstein and Jersey cows each produced about 2.4 and 1.5 kg of feces and 1.8 and 1.4 kg of urine per elimination event per day, respectively. Although the elimination frequencies per 24 h were similar between breed (Holstein: 17 defecations, 8 urinations; Jersey: 18 defecations, 9 urinations), Holsteins produced more excreta total.
Reports of the durations of elimination events are rare. There is opportunity for this measurement to be used as a proxy for estimating the amounts of excreta produced. The stanchioned dry cows of the study by Sahara et al. (1990) defecated and urinated for average durations of 7.6 ± 2.6 and 11.7 ± 5.5 s, respectively. Defecation durations did not vary between hours of the day, whereas urinations were of longer durations for events occurring from 0000 to 0800 h, likely due to decreased frequency of events during this period (Sahara et al., 1990).

Cow-Related Effects
Cow-related effects detailed in this section include age or life stage, reproductive stage, and individuality of animals. Dairy calf eliminations over time and space were described in a study by Vaughan et al. (2014a). Female Holstein calves group housed with automatic feeders were observed at 2 time periods (averaged 32 and 61 d old) and, similar to dairy cows in stanchion barns (Sahara et al., 1990;Aland et al., 2002;Acatincăi et al., 2011), showed greater elimination frequency during the active daytime hours. Additionally, large differences in elimination frequencies were observed between calves, as was seen in most studies involving mature animals. Of further interest are the effects of age: the authors found a positive correlation between age and elimination frequencies for unweaned calves. After weaning, differences were attributed to visits to the water feeder, whereas calves also eliminated more frequently on the slatted flooring near the feeding areas compared with the sawdust resting area. This was possibly the result of a combination of more frequent visits to the feeders and decreases in resting time (Vaughan et al., 2014a).
Differences in eliminative behaviors between heifers and cows may exist. Cows showed significantly more defecations, but not urinations, than heifers in a stanchion barn (Aland et al., 2002). In comparing primiparous and multiparous cows in tiestalls, Herlin et al. (1994) described how the former group had significantly higher (poorer) scores for fecal contamination of the lying area. Size in this case was likely a confounding factor, as an average 100-kg difference in live weight between parity groups was noted, meaning that younger animals had a greater range of movement within their stalls. The situation was the opposite in cubicles, where primiparous cows very seldom dirtied their lying areas. Whether this was due to the difference in the importance of lying or differences in body condition between the age groups remains unclear (Herlin et al., 1994). One explanation for tiestall housing may be that multiparous cows have more experience with All animals in dry stage. 6 Involved some cows in dry stage. their environment; they might recognize that defecating in the tiestall means they will have to lie in it, and thus avoid dirtying their stall. Within cows specifically, age does not appear to affect elimination frequencies.
In the case of lactating cows in tiestalls (Fuller study from 1928, data published in Aland et al., 2002) and freestalls (Villettaz Robichaud et al., 2011), neither age nor parity, respectively, were correlated with differences in daily elimination frequencies. In another study (Draganova et al., 2016), lactating cows in a rotationally grazed system were monitored for urination behavior with mechanical sensors. Similar to findings in other studies, the cows varied in their daily frequencies, but this difference was not associated with age (Draganova et al., 2016). Some evidence suggests that a cow's eliminative patterns may change as a result of life stage. For example, only the first experiment by Villettaz Robichaud et al.
(2011) showed a correlation between frequency of urination and days in milk. This may be partially explained by the fact that most cows in the first experiment were pregnant, compared with no pregnancies in this study's second experiment. Aland et al. (2002) calculated significantly more urinations and a tendency for more defecations in lactating cows compared with dry cows in the analyses of Fuller's study. No interaction was found between breed and lactating or dry status, nor was there a significant effect of breed. The observational study by Sahara et al. (1990) gives a baseline for the eliminative behaviors of dry dairy cows. The authors stated that feed intake and moisture content may have affected defecation frequencies of their dry cows, versus feeding conditions of lactating cows of comparable studies (Sahara et al., 1990). Such results from dry cows may be due to differences in feed and energy requirements. However, it has not been determined whether the rate of feed intake-for example, that assumed by higher liveweights and greater supplemental feeding provided to a group (Hirata et al., 2011)-or qualities of the feed, such as fiber content (Acatincăi et al., 2011), affect the frequencies of eliminations.
The other general observation for eliminative behaviors is that individual variation seems to occur between animals. Looking at the standard deviations provided by the papers in Table 1, averages for defecations have a wider spread than averages for urinations in indoor systems. Aland et al. (2002) found that the factor of individual cow had a significant effect on the numbers of eliminations per 24 h. For example, one cow may average nearly double the number of defecations (about 22 events per 24 h) compared with another (about 12 events per 24 h). Over the 9 d of observations, individual cows varied in their frequencies of both defecations and urinations, but no significant differences between different days were found (Aland et al., 2002). This appears to be the same situation for freestall systems, as the ranges in numbers of eliminations per 24 h (Villettaz Robichaud et al., 2011) supported those in Aland et al. (2002). These variations in frequencies were also not related to cow characteristics such as days in milk and feed and water intake (Villettaz Robichaud et al., 2011). The authors suggest that cows which eliminate more frequently do so consistently, based on correlations between days ranging from 0.49 to 0.77 and 0.28 to 0.70 for defecations and urinations, respectively (Villettaz Robichaud et al., 2011). In contrast, the distributions of average eliminations for observations carried out over five 24-h periods on cows at pasture were similar between individuals (White et al., 2001). Defecations for lactating Holsteins and Jerseys varied by 0.5 events (SE), whereas urinations varied by 0.6 events (SE) for both breeds. It should be noted that this data for average eliminations was for only 8 animals (4 Holstein and 4 Jersey) total (White et al., 2001).

Environmental Effects
Environmental effects described in this section include temperature and humidity, housing system or type, and the interaction between housing and the timing of management activities and cow active periods. Several sources state that temperature (Fuller, 1928, via Aland et al., 2002Hafez, 1969;Oudshoorn et al., 2008;Acatincăi et al., 2011) and humidity (Hafez, 1969;Acatincăi et al., 2011) are known to influence elimination frequencies, but with little specification as to how or in what way. The perception may be that higher temperatures lead to less frequent defecations and more frequent urinations. These changes would reflect cattle's decrease in feeding and increase in drinking activity in response to heat stress (Kadzere et al., 2002). Dairy cows increase water intake and draw water from feces to increase evaporative cooling by sweating when under thermal stress. This physiological strategy explains why, in one case, urine outputs were similar and feces volumes were reduced between cool (15-24°C) and hot (32.2°C) periods of observation for lactating Holstein cows (McDowell et al., 1969). Acatincăi et al. (2011) observed their stanchioned lactating cows during the winter and summer seasons, with temperatures averaging 8.1 and 28.4°C, respectively. Defecation frequency per day was not statistically different between the seasons, whereas during the summer significantly fewer urinations took place. Additionally, White et al. (2001) noted how dairy cows at pasture tended to stay closer to the water tank as the temperature-humidity index rose. A greater amount of time spent in the area led to a greater concentration of eliminations. The literature appears to suggest that, in addition to affecting physiology, temperature can influence the immediate movement of cattle when they have the ability to control their position in the environment. As a result, changes occur in the distribution of feces and urine patches.
In comparing housing systems, we begin to see that eliminative behaviors can be affected in ways beyond daily frequencies. However, for calves, the only evidence of a relationship between housing and eliminations lies in frequencies. In Phillips (2004), Friesian-beef breed crossed calves were housed individually or in groups of 3 starting at 1 wk of age. Grouped calves were found to eliminate more frequently than individually housed calves and, although they were subjected to different feed types, were presumed to consume similar amounts of water. Following this, the author believed that the differences in frequencies were a result of social facilitation (Phillips, 2004). If the prediction by Phillips (2004) is true, it may be important to consider the social effect of eliminative behavior for mature animals. If some cows were to start defecating, for example, does this cause other cows to start defecating? However, one would first need to discern whether peaks in elimination frequencies are only associated with daily patterns or with natural rhythms in conjunction with feeding and resting behaviors.
Even within types of indoor housing systems, eliminative behaviors may vary. Herlin et al. (1994) compared the fecal contamination of tiestalls, feeding cubicles (lying stalls with a feed table in front), and cubicles, listed in order of most to least time occupied by the cow. Cubicles and stalls were cleaned during 2 periods daily, directly after morning and afternoon contamination scores were recorded. For multiparous cows, tiestalls had a greater percentage of contaminated stalls in the afternoon compared with the other systems, whereas feeding cubicles had a lower percentage of contaminated stalls in the morning compared with the other systems. It is important to consider that tiestall cows spend their entire day in the stall. With that being said, the authors reasoned that the cleanliness of feeding cubicles was a result of concentrates being delivered at a separate location from the forage at the feeding cubicles (Herlin et al., 1994). Evidence also suggests that the structures of housing systems affect the feces avoidance element of eliminative behavior. Cows in a more restrictive cubicle system remained lying or resumed lying after defecation more frequently, as opposed to walking away from their own manure in a straw yard system. Behavioral sequences showing incidental (e.g., moving away from feces while performing another activity) and intentional (e.g., stopping a behavior to defecate) avoidance were more frequent in the straw yard system, and avoidance was not as affected by milk yield clas-sification as it was in the cubicle system (Whistance et al., 2007). It has been hypothesized that consistent exposure to excreta "may cause habituation to its offensiveness" (Phillips, 2002, p. 150). Consistent exposure in this case was described as a feature associated with indoor, loose-housing systems, but in the sense that cattle become accustomed to the odor of excreta rather than physical contact with it. Housing systems without open areas bring up the question of excreta avoidance, in which case it could be interesting to differentiate between avoidance of an animal's own waste or to that of its herd mates.
Different outdoor housing systems for dairy cattle vary in terms of duration of access and the actual presence of pasture. Oudshoorn et al. (2008) found that when lactating cows were given time-limited pasture access, frequencies of eliminations per cow per hour of outdoor time remained similar through 4-h, 6.5-h, and 9-h treatments. Averaged results from visual observations and GPS logging showed 0.36, 0.37, and 0.38 average defecations and 0.29, 0.25, and 0.25 average urinations per hour for these 3 time treatments. Cows altered their activity and time budgets to maximize amounts of grazing time, but this still did not affect elimination frequency. Furthermore, feces and urine were distributed without specific hot spots for all time treatments (Oudshoorn et al., 2008). Observations on lactating cows in a rotationally grazed pasture resulted in seasonal variation in elimination concentrations (White et al., 2001). The time spent in a location was highly correlated with the number of eliminations in that location, such that warmer seasons saw significantly more eliminations within 30 m of the single water tank. Other seasons resulted in more even distributions of excreta. Ultimately 84.7% of total manure and 84.1% of total urine was deposited during the 20 h of pasture time, and only 10% of the total pasture was covered by excreta over the year of the study, which the authors use as arguments in favor of maximizing outdoor grazing periods (White et al., 2001). Whistance et al. (2011) noted that, in conducting their feces avoidance studies on cows at a cubicle system, in a straw yard system, and at pasture, pastured cows did not eliminate as promptly as housed cows and were quicker to change their behavior after defecating. These observations might be a function of the comfortability of lying surfaces and the decreased space and movement opportunity associated with cubicle systems (Whistance et al., 2007). To the author's knowledge, studies involving eliminative behavior in dry lot systems do not exist. One might propose that dairy cows in dry lot systems would behave similarly as in indoor, loose-housed systems, but with the added factor of temperature extremes requiring heat stress management (Tresoldi et  Housing type affects the schedules of barn staff and animals; therefore, some aspects of eliminative behaviors in dairy cattle are associated with activity periods. Dairy breeds in stanchion barns seem to show more eliminations during times when cows and humans are most active, during the times just before milking and during feeding. It then follows that frequencies are lower during resting periods after milkings and during the evenings (Aland et al., 2002;Acatincăi et al., 2011). Stanchion-housed dry cows were reported to have eliminated more frequently during the day, from 0800 h to 2000 h, compared with at night, from 2000 h to 0800 h, but specific diurnal patterns were not described (Sahara et al., 1990). Freestall systems for dairy cows may differ in this activity-elimination pattern. Whistance et al. (2007) observed most defecations to occur while cows were in the cubicle alleys, not lying or resting, which appears to align with the patterns of the studies previously mentioned. But consideration must be given as to what classifies as being active, as the authors suggest that standing in the cubicle alleys as opposed to the feed alleys was a result of lower comfortability of the stalls. Had lying areas been of a more suitable quality, this idle time might instead have been used for lying (Whistance et al., 2007). The other freestall study on eliminative behaviors by Villettaz Robichaud et al. (2011) showed patterns of standing and time at the feeder consistent with diurnal rhythms, but these were not closely associated with frequencies of defecations and urinations. It was concluded that eliminations were distributed evenly across 24 h (Villettaz Robichaud et al., 2011). Knowing that elimination frequencies may change throughout a day means that observational studies examining a specific segment of time per 24 h could be affected.

Exercise
The amount of exercise cows achieve is a function of the housing environment and management, and can relate to the overall health of the animals (e.g., Davidson and Beede, 2009). It has been suggested that exercise may affect the gastrointestinal tract in ways such as circulating blood flow away from the intestines and stimulating colonic motility in humans (Peters et al., 2001), but eliminative behaviors do not appear to be noted in such exercise-related research for dairy cattle.
Anecdotal notes suggest a relationship between exercise and eliminative behaviors. One example is the description of the "bovine treadmill," which states that "digestive upsets" can be affected, and explains that "walking movement massages the entire bovine digestive tract" (Smith, 2010). In another case (Anderson et al., 1977), a circular exercise mechanism was created to subject heifers and dry cows in late gestation to a controlled exercise regimen. The goal was to conduct exercise trials on the notion that, for dry cows, the digestive, cardiovascular, and respiratory systems decline in condition, muscle tone, and functional efficiency before parturition (Anderson et al., 1977). Future research involving exercise for dairy cattle, whether it be forced or offering varying levels of movement opportunity, should at minimum record elimination frequencies.
Such information would be helpful in both confirming the statements above and determining the practicality of managing excreta in exercise areas.

General Arousal and Affective State
Dairy cattle respond to negative experiences via changes in eliminative behavior. Monitoring defecations and urinations is a common method of determining animals' responses to stimuli, and the relationship between eliminations and negative states such as fear and anxiety (as reviewed by Forkman et al., 2007) has been well studied. For dairy cows specifically, it has been documented that more eliminations occur when tiestall cows were placed in an unfamiliar room for milking. These effects of visual and olfactory isolation from conspecifics were significantly reduced by brushing by a familiar human (Rushen et al., 2001). More frequent eliminations were also the case when tiestall cows were subjected to an aversive handling treatment (struck on the head and muzzle) compared with a gentle handling treatment (offered hay or grass or concentrates, stroked, gently spoken to). A total of 32 treatment repetitions took place over 5 total days. Eliminations did not occur after the first 6 aversive and 6 gentle treatments over the first 2 d of experimentation, which the authors attribute to adaptation once cows learned to discriminate between handlers (Munksgaard et al., 1997). In a study by Kilgour (1975), defecation likely occurred in response to being separated from the rest of the group when dairy cows were to enter an open-field testing arena. Defecation occurred before actually entering the arena in many cases (Kilgour, 1975). In addition to defecating more frequently, cattle under stress will produce feces that are more liquid in consistency (Phillips, 2002).
In contrast, eliminative responses to events of positive valence are less clear. For instance, returning to Kilgour's open-field test (1975), researchers noted that most cows appeared to explore their new temporary environment, while some even pranced, a term which they described as being "excited." Considering the activation of the autonomic nervous system, which does not necessarily differentiate between positive and negative triggering, one might propose that increased eliminations could happen as a result of a more generally "aroused" state. Villalba and Manteca (2019) argue that situations such as fighting and play both may result in a similar physiological stress response, even though they are aversive and rewarding, respectively. They also note that the degree of control an animal has over its situation and environment should be the factor determining whether an event is perceived as negative or positive. However, the authors' notions are based on grazing animals, and opportunities for providing enrichment that presents beneficial challenges associated with this environment (e.g., navigating terrain and vegetation) (Villalba and Manteca, 2019) are limited for dairy cattle in total confinement housing systems. Therefore, future research should consider the different contexts of animals' situations when interpreting results where eliminative behaviors are included as key or additional measures. The same notion of considering context should also apply to how researchers interpret the existing literature surrounding eliminative behavior.

Housing and Management
Methods of controlling eliminative behaviors for dairy cattle begin with the housing system. Tiestall and cubicle housing are designed to separate the animal from its excreta by use of brisket boards and neck rails to limit forward movement in the stall, curbs that keep the lying surfaces separated from excreta as it is scraped or flushed, and, in some cases, gutters directly behind stalls. Whether it be scraping and flushing alleys, cleaning out deep-bedded systems, or tilling compost-bedded pack, studies suggest a minimum of 2 maintenance periods per day (as reviewed by Bewley et al., 2017). The idea of a "hygienic relief area," essentially providing loose-housed cows access to pasture or anywhere off of concrete, has also been suggested, because the physical grass and ground textures aid in cleaning the feet (Bergsten, 2001, p. 18).
Electric trainers are used to maintain cleanliness of the tiestall. The devices deliver a shock if touched when the back is arched for defecation or urination and train the cow to step backward for elimination (Phillips, 2002). Trainers are still used in many tiestall farms (e.g., a previous epidemiological study noted that trainers were used in 71.1% of 100 farms examined in Ontario and Quebec, Canada, during 2011;Charlton et al., 2016), but their effectiveness is not certain. For example, evidence suggests that the use of electric trainers was associated with dirty hind limbs (up to or over the hock joint) and slightly dirty udders when compared with stalls without trainers in a study involving 317 Ontarian tiestall farms, which the authors hypothesized to be due to improper placement (Zurbrigg et al., 2005). Regarding this matter, Villettaz Robichaud et al. (2018) reported that only 65% of the same farms studied by Charlton et al. (2016) met the criterion for appropriate trainer placement, keeping in mind that this percentage also included farms that did not use electric trainers and, thus, automatically met criterion for proper placement. More importantly, walking backward is not a natural form of movement for cattle (Phillips, 2002). This fact may contribute to the difficulties cows face in eliminating off of or beyond the stall curb in tiestall and freestall systems. Other factors to consider, particularly for tiestall and stanchion barns, are the forward positioning of the cow while feeding, the tendency to defecate while feeding (Aland et al., 2002), and the natural process of cattle walking forward while grazing (Phillips, 2002) or after defecating (Whistance et al., 2007(Whistance et al., , 2011. One straightforward concept of controlling eliminative behavior is to influence timing by way of management practice. Brantas (1968) described the training of cows to a cafeteria stable, a partial loose-housed system involving a bedded lying area and individual feeding stalls with headlocks. Following the tendency to eliminate after rising from resting and during standing periods, cows ideally were to eliminate into the excreta passage during 4 daily feeding sessions. Eliminations per cow per hour were to some degree greater during feeding (defecations: 0.68; urinations: 0.40) compared with during loose resting periods (defecations: 0.52; urinations: 0.24). However, rates drastically rose (defecations: 2.92; urinations: 1.84) approximately halfway through the periods between sessions, consequently soiling the bedded area. Six feeding sessions might have resolved the issue, but this was deemed to be an impractical level of management (Brantas, 1968). An experiment by Villettaz Robichaud et al. (2013) involving the use of footbaths was conducted on lactating cows, to assess whether defecation can be stimulated. Neither treatments of walking through or standing in water-filled footbaths, nor having feet sprayed with water or air, proved reliable in prompting this behavior. It appeared that the defecations observed were a result of novelty, as the total percentage of cows that defecated for control and any treatments combined reduced over the 26 d of experimentation (Villettaz Robichaud et al., 2013). Whistance et al. (2009) attempted to toilet train 14-to 16-mo-old heifers via 4 phases involving classical and operant conditioning. All heifers were successfully trained in the first 3 phases, and phase 4 showed that eliminations on concrete occurred about 4 times as often as on straw. However, because concrete-to-straw movement before eliminating did not decrease as elimination on concrete increased, and straw-to-concrete movement before eliminating did not increase as elimination on straw decreased, actual toilet training was not achieved. Heifers showed awareness of their own eliminations, but the incompletion of the final phase may have been due to the inability to associate flooring type with eliminations (Whistance et al., 2009). In another study by Vaughan et al. (2014b), separate sets of 1-to 2-mo-old female calves were subjected to 2 experiments in an attempt to train urination to a specific stall. Operanttrained calves urinated more frequently in the specified stalls than their controls, whereas the opposite was the case for some classically trained calves. On average, operant-trained calves also urinated approximately twice as promptly as classically trained calves, but the former group had a large variation in how quickly they learned (Vaughan et al., 2014b). Dirksen et al. (2020a) state that a multitude of steps must be achieved in toilet training, and suggest that operant conditioning methods may prove superior in toilet training processes because defecation and urination involve both involuntary and voluntary muscle control. Dairy cattle have the capacity to be toilet trained under the correct conditions, but the practicality of managing these processes for industry settings should be considered (Dirksen et al., 2020a). Following these ideas, Dirksen et al. (2020b) outlined a behavioral chain for toilet training urination. Five female calves around 91 d of age were first operant trained for voluntary behaviors of responding to a vibration signal to enter a latrine, a flashing light to signal a food reward, and the empty reward bowl to signal leaving the latrine. These steps were combined with the reflexive responses of internal signaling of urination and behaviorally showing pre-urination, interrupting or withholding urination, and reinitiating urination. Calves were successful in responding to the vibration signal, in that 95% of sequences initiated outside of the latrine led to latrine entry, with 65% of these sequences resulting in urination in the latrine. Sequences where calves began in and urinated in the latrine were considered ambiguous, due to the inability to accurately determine whether their placement was due to anticipating urination or just wanting to be near the reward, and occurred 31% of the time (Dirksen et al., 2020b).

Training Latrine Behavior
One technological advancement toward toilet training is the use of an automatic urinal system. This consists of a separate stall, which cows in loose-housing systems enter voluntarily to receive feed. Sensors determine the placement of the urine collection cup, which moves to stimulate the nerve superior to the suspensory ligament of the udder and activates the urinary reflex. Urine is collected separately from manure and can be stored for future use, such as fertilizer or as a source of hydrogen for energy. Preventing the mixing of manure and urine also reduces ammonia emissions, thus providing an improved environment for animals and barn staff. Currently there is no published evidence of this system's effectiveness, as it is expected to be commercially available to the Dutch market as of writing the present review (Hanskamp, n.d.). Manual stimulation of the escutcheon nerve is a well-known technique and is recognized in the literature for the collection of urine samples (e.g., Andersson and Larsson, 1952;LeBlanc et al., 2005;Løvendahl and Sehested, 2016). However, it does not seem to be referred to in textbooks or research specifically relating to eliminative behaviors.

RESEARCH GAPS AND FUTURE DIRECTIONS
A more thorough look into the eliminative behaviors of dairy cattle would be beneficial for both managementrelated and animal-related matters. One place to start may be to research baseline behaviors in environments with as little human interference as possible, as the present review has shown that eliminative behaviors can be affected by environmental factors such as housing (e.g., stalls or cubicles) and activity levels (e.g., the timing of management activities such as milking). The first component of this involves observing behaviors just before elimination, whereas the second component involves full 24-h observations and comparisons of different breeds' eliminative behaviors outside of just elimination frequencies. These points would lead to a better understanding of the entire sequence of eliminative behaviors, which may lead to opportunities in improving the timing of barn cleaning or even refining toilet training techniques. Another suggestion is for researchers to record eliminative behaviors during any type of study that is conducted involving dairy cattle. Simple records of elimination frequencies can provide some insight to factors such as activity period, exercise, and affective state.

CONCLUSIONS
Of the literature regarding dairy cattle eliminative behavior, most focus is on frequency and distribution over time and space. Even within this information, variances occur due to numerous factors specific to each situation, such as an animal's life stage or the amount of available space for movement. Ultimately, stalls and cubicles are limited in their ability to prevent the animal from eliminating in the lying space, with the knowledge that such systems were also not designed with the animal's natural eliminative behaviors in mind. Timely, more frequent management routines and toilet training are options in managing eliminative behaviors. The dairy industry would also largely benefit from more research in pre-and post-eliminative behaviors, as this information will improve the management of excreta for indoor housing systems. Additionally, a better understanding of the eliminative behavioral sequence may lead to improvements in terms of implementing less-restrictive housing.