Anticipatory Behavior in Dogs: Excitement Before Rewards as a Welfare Signal

1. What Anticipatory Behavior Is

1.1 The Appetitive Phase

Ethologists have long divided goal-directed behavior into two phases. The appetitive phase is everything an animal does on the way to a goal: searching, approaching, waiting, getting ready. The consummatory phase is the goal itself: eating, playing, being greeted. Anticipatory behavior belongs to the first phase. A recent review defines it as "a suite of behaviors, expressed by animals during the appetitive phase, or before a desired outcome is acquired" (Krebs et al., 2022).

Two features of this definition matter. Anticipatory behavior is not one behavior but a set: increased activity in one animal, sitting still and watching in another. And it is defined by its position in time, not by its form, so it has to be identified from its relation to an expected event, not from a single snapshot.

1.2 How Researchers Make It Measurable

The standard laboratory approach uses Pavlovian conditioning. A neutral cue, such as a sound or a light, is repeatedly followed by a reward after a fixed delay. Once the animal has learned the association, the interval between the cue and the reward becomes a window in which anticipation can be observed under controlled conditions. Researchers then count behaviors or the number of transitions between behaviors in that window. The idea that this interval could serve as a welfare tool goes back to a concept paper that described "anticipation in Pavlovian conditioning" as "an easy and useful tool" for assessing the state of an animal's reward systems (Spruijt et al., 2001).

In zoos and on farms, the same principle is applied to events that already happen on a schedule. Animals can learn that food arrives at about the same time each day, that a keeper's arrival predicts a training session, or that a particular door opens before access to an outdoor area. A review of this work notes that outcomes "can become predictable to animals either by happening at approximately a similar time every day, being cued (intentionally or not), or some combination of the two" (Krebs et al., 2022). Methods for identifying which behaviors are anticipatory in a given animal have themselves become a topic of research (Krebs et al., 2017).

1.3 The Typical Time Course

Anticipatory behavior has a characteristic shape in time. According to the review literature, it increases as the expected event approaches "and then drops off suddenly when the desired event arrives" (Krebs et al., 2022). The same review notes that, in laboratory studies, food anticipatory activity tends to increase within an hour of expected feedings.

That time course is the most useful single feature for recognizing anticipation. A dog that becomes restless in the half hour before its usual dinner and settles as soon as it has eaten shows the pattern. A dog that is restless all evening, regardless of when it is fed, does not, and its restlessness probably has a different cause. Dogs do not need a clock to show this kind of timing; an entrained daily rhythm and learned cues are enough to produce accurate-looking anticipation (what is known about time perception in dogs).

2. The Reward-Sensitivity Hypothesis

2.1 Welfare as a Balance

The welfare interpretation of anticipatory behavior starts from a particular definition of welfare. Spruijt, van den Bos and Pijlman proposed that welfare can be understood as "the balance between positive (reward, satisfaction) and negative (stress) experiences or affective states" (Spruijt et al., 2001). They linked this balance to brain systems involved in reward, in particular opioid and dopaminergic mechanisms, and argued that the condition of these systems could be read from behavior.

The reasoning runs as follows. Stress and reward systems interact. An animal that lives under chronic stress or deprivation may become more sensitive to rewards, because each reward carries more weight against a negative background. A later review from the same research group described three possible uses of reward-related behavior: measuring welfare through reward sensitivity, improving welfare by presenting rewards, and assessing how animals perceive particular rewards (van der Harst & Spruijt, 2007). The measure of reward sensitivity they used in rats was anticipatory behavior, and they concluded that anticipatory behavior can be evoked in a wide range of other species.

The framework belongs to a broader shift in welfare science toward recognizing positive states (how welfare and quality of life are assessed), and anticipatory behavior seemed to offer a behavioral window onto that positive side.

2.2 Too Much, Too Little and Just Right

The hypothesis has a counterintuitive core. It does not say that more anticipation means better welfare. It says almost the opposite: very intense anticipation may indicate that rewards are scarce or that the animal's baseline is poor, so that each expected reward becomes especially important. In this reading, an animal living in a rich environment, with many rewarding events across the day, has less reason to invest heavily in any single one.

Watters formulated this for zoo animals. As summarized in a later review, animals in a positive state of welfare "are expected to exhibit frequent but low intensity anticipatory behavior toward known rewards", while animals in a more negative state "may show infrequent but intense anticipation toward known positive outcomes" (Watters, 2014, as cited in Krebs et al., 2022). A related paper adds that animals in high-quality environments show lower-intensity anticipation toward single events because they have numerous opportunities throughout the day (Krebs & Watters, 2019).

The framework also has a theoretical other end. If heightened anticipation suggests a reward system that is hungry for input, then a near absence of anticipation toward events that should be rewarding could suggest that rewards have lost their pull, a state sometimes compared with anhedonia in humans. This reading is a logical extension of the reward-sensitivity idea, not a measured finding in dogs, and the rat study in section 3.1, which tested drug treatments, did not find what a simple depression model would predict.

The hypothesis thus describes a curve rather than a straight line, which is elegant but hard to apply to a single animal, because nobody knows where on that curve a given dog's normal level lies.

2.3 Wanting Is Not Liking

The neurobiology behind the hypothesis rests on a distinction that comes from rodent research. Dopamine was long described as the brain's pleasure chemical. Work that manipulated dopamine and measured responses separately led to a different account: dopamine is more closely associated with "wanting", the motivational pull toward a reward, than with "liking", the hedonic impact of receiving it (Berridge & Robinson, 1998). Anticipatory behavior is, in these terms, an expression of wanting.

A second line of work explains why anticipation attaches to cues. Recordings from midbrain dopamine neurons in monkeys showed a burst of activity to unexpected reward, little change when a fully predicted reward arrived on time, and a dip below baseline when a predicted reward was omitted. With learning, the burst moved back in time to the earliest reliable predictor of the reward (Schultz et al., 1997). The cue, not the food, became the event that carried the signal (how prediction error drives learning).

This explains why the leash, the sound of a food container or the owner putting on shoes can trigger so much behavior: these are the earliest reliable predictors in a dog's day. But it is important to keep the evidence base in view. The wanting-liking dissociation was shown in rodents, and the prediction-error pattern was recorded in primates. No canine study has measured dopamine in this way, and the mechanism is applied to dogs rather than demonstrated in them (what dopamine does and does not do in dogs).

3. Evidence from Other Species

3.1 Rats in Standard and Semi-Naturalistic Housing

One clear test of the reward-sensitivity idea compared rats reared and housed long term in two very different environments (Makowska & Weary, 2016). Female rats lived either in standard laboratory cages or in a semi-naturalistic environment. Their behavior was recorded in the interval between a cue signaling a reward and the arrival of that reward, and the measurement was repeated before and after treatment with an antidepressant or an anxiolytic.

In all experiments, total behavioral frequency during the interval was higher in standard-housed rats than in rats from the semi-naturalistic condition. The authors interpreted this as a sign that the standard-housed rats were more sensitive to rewards and experiencing poorer welfare. The form of anticipation also differed: standard-housed rats mostly reared up, while rats from the semi-naturalistic environment mostly sat facing the direction from which the treat would come. The drug treatments had no effect on either the quantity or the form of anticipatory behavior. The authors concluded that the poorer welfare of the standard-housed rats was not analogous to depression or anxiety, or alternatively that the drug interventions were ineffective.

Two points stand out. The better-housed rats still anticipated, but by waiting and watching, so the groups differed in form as well as amount. And the drug results show that heightened anticipation cannot simply be equated with a clinical mood state. Reading higher anticipation as poorer welfare remains an inference from the framework, not a measurement of how the rats felt.

The same research tradition also produced earlier rat studies on standard versus enriched housing, summarized in the review by van der Harst and Spruijt (2007). That review, too, rests on rats and largely on the work of one research group.

3.2 Mice and Access to a Playpen

Anticipatory behavior has also been used the other way round: not to diagnose a poor environment, but to test whether something is rewarding. In a study with 42 female laboratory mice from three strains, some mice received access to a playpen three times a week for 30 minutes over 14 days, while control mice stayed in their home cages (Ratuski et al., 2021). The latency to enter the playpen declined from 168 ± 22 to 13 ± 2 seconds over the trial. Mice with playpen access showed an increase in anticipatory behavior before access (mean ± SE = 19.7 ± 2.6 behavioral transitions), while control mice showed no change.

The authors note that anticipation "is commonly assumed to represent a state of 'wanting'", and they raise a point that is easy to overlook: if animals learn to look forward to a valued opportunity that is later withdrawn, the experience of having had it may make its absence worse. Anticipation can thus serve as evidence that something matters to an animal, and at the same time as a reminder that an expectation, once created, has consequences.

3.3 Farmed Animals and Fur Animals

The concept was carried into farm and fur animal research early on. A short paper on farmed silver foxes, a canid species, proposed reward-related behavior and emotional expressions as the basis for a novel tool to study animal welfare (Moe et al., 2006). A study in mink examined temperament, stereotypies and anticipatory behavior as measures of welfare (Hansen & Jeppesen, 2006). In lambs, the behavioral expression of positive anticipation was studied in relation to food and to an opportunity to play (Anderson et al., 2015). The review literature summarizes such farm animal studies as showing that a variety of animals exhibit anticipatory behavior ahead of enrichment opportunities, not only ahead of food (Krebs et al., 2022).

3.4 Zoo Animals and Dolphins

Zoo research has been particularly active, because zoo animals live with predictable daily routines and many of their positive events are scheduled. Watters (2014) proposed anticipatory behavior as a behavioral indicator of welfare in zoos. A study of bottlenose dolphins reported that cognitive judgment bias was associated with the frequency of anticipatory behavior (Clegg & Delfour, 2018). Another dolphin study concluded that the animals' anticipatory behavior indicated their motivation to participate in specific events with their caretakers (Clegg et al., 2018).

The judgment bias association is worth noting because judgment bias tasks are one of the few other tools used to infer affective state in animals (what cognitive bias tests can and cannot show). An association between two indirect measures does not validate either of them, but it is more than either measure provides alone. The dolphin studies were carried out with animals under human care, and their designs and sample sizes are not detailed here, so they illustrate the concept rather than establish its size or generality.

3.5 Horses at Feeding Time

In a small observational study of six stabled horses in one yard, behavior was recorded on video for five days, from one hour before to one hour after feeding (Mata et al., 2024). The authors reported that ration-fed horses exhibited more stereotypic and arousal behaviors, with anticipatory behavior intensifying closer to feeding time; stereotypies increased as feeding approached, with significant increases in the final 20 minutes.

The study is far too small to establish a general pattern, but it illustrates the problem: when anticipation and stereotypic behavior rise together before a meal, the boundary between looking forward to food and coping with a deficit becomes blurred.

4. What Has Been Measured in Dogs

4.1 A Thin Evidence Base

Despite the obvious relevance of the topic to dogs, no published dog study was found that tests the reward-sensitivity hypothesis directly. No study has compared anticipatory behavior in a cue-reward interval between dogs living in poorer and richer conditions, and anticipatory behavior has not been validated as a welfare index in dogs. The major review of anticipatory behavior in animals under human care does not cite a single dog study (Krebs et al., 2022). What follows is therefore not evidence for the hypothesis in dogs, but the dog research that touches on anticipation from different angles.

4.2 Faces During Positive Anticipation and Frustration

The most direct canine data on anticipation come from facial expression research. In a controlled setup, Labrador retrievers were filmed while they expected a food reward or were prevented from reaching it, and their facial movements were coded with DogFACS, a system that records individual muscle movements without assigning meaning to them (Bremhorst et al., 2019). An ear movement coded as ears adductor was more common during positive anticipation. Blinking, lips parting, jaw dropping, nose licking and ear flattening were more common during frustration.

A new sample of 28 dogs, mostly Labrador retrievers, was tested with a similar design using two reward types, food and toys (Bremhorst et al., 2022). The earlier associations were replicated, and all facial actions except the upper lip raiser were independent of reward type. The authors then assessed diagnostic accuracy. Ears adductor had excellent specificity but low sensitivity: when it occurred, it was a good sign of positive anticipation, but it was often absent during positive anticipation. Ear flattening and ears downward showed the opposite pattern. Used individually, none of the facial actions would allow consistent correct classification of the associated emotional state.

These studies show that positive anticipation in dogs is not a vague concept: it can be produced experimentally and is accompanied by measurable facial changes. They also show the limits. The replication came from the same research group, and the setting was a short laboratory task with food or toys, not a household routine. The findings support treating these facial actions as candidate indicators in that kind of situation, not as a code for reading every dog at the front door (what research on canine body language shows).

4.3 When the Dog Makes It Happen

A study of 12 beagles used a yoked design to separate what a dog receives from how it receives it (McGowan et al., 2014). Dogs in the experimental condition learned operant responses that gave them access to food, human contact or contact with another dog. Each experimental dog had a yoked partner that received the same reward after a matched delay, without having to perform the response. The dogs that produced the reward through their own behavior responded differently from those that simply received it, a difference the authors interpreted as positive emotional arousal.

For anticipatory behavior, this matters because the waiting period differs between the two conditions. A dog that knows its own action will produce the outcome is in a different position from a dog that can only wait. The study does not separate control from predictability perfectly, and it is small. But it suggests that the behavior seen before a reward depends not only on the reward but on the dog's role in obtaining it (why control and predictability matter).

4.4 Reunion After Time Alone

Anticipation in the everyday sense also concerns the return of the owner. In a study of 12 privately owned dogs without separation-related problems, each dog was left alone at home for 0.5, 2 and 4 hours, and its behavior was recorded before departure, during the absence and for 10 minutes after the owner returned; heart-rate measures were also collected (Rehn & Keeling, 2011). No significant behavioral differences were detected during equivalent portions of the absence, but several differences appeared at reunion.

Effects of a waiting period may thus become visible when the expected event arrives rather than during the wait. The study is small, the dogs were selected for coping well with being alone, and normal owner routines were retained, so the greeting is not a pure measure of the absence.

4.5 What Owners Report

Owners have their own picture of canine anticipation. In a survey of 445 owners in the UK and Ireland about how they recognize positive emotional states in their dogs, two clusters emerged (Buckland et al., 2014). High-arousal positive states, described as "happy and excited", were linked to the anticipation of walks or food and described in terms of barking, a raised head, wide bright eyes, pricked ears, tail wagging, an open mouth and active play. Low-arousal positive states, described as "happy and relaxed", were linked to resting and gentle stroking. Owners described the high-arousal states in far more detail than the low-arousal ones.

This shows how owners interpret their dogs, not what the dogs experience. But waiting for food and for an outing, the situations studied in other species, are where owners most readily see positive emotion, and quieter positive states may be underrecognized because excitement is easier to see.

5. Anticipation, Frustration and Arousal

5.1 Three Different Things

Much of the confusion around excited dogs comes from treating three different concepts as one. Anticipation describes a phase: the period before an expected outcome. Frustration describes a state that follows when an expected outcome is delayed, reduced or withheld. Arousal describes intensity: the degree of physiological and behavioral activation, regardless of whether the state is pleasant or unpleasant.

Affective science treats arousal and valence as two largely independent dimensions (Mendl et al., 2010). A dog spinning at the door before a walk is highly aroused. Whether that arousal is positive anticipation, frustration because the owner keeps stopping to look for keys, or a mixture of both is a question about valence, and arousal alone cannot answer it (how arousal and its regulation work).

5.2 When Expectation Is Violated

Frustration requires prior anticipation. It is defined by an expectation that is not met, and the canine research on it uses exactly that structure. Dogs were trained to gaze at an experimenter's face to obtain food. When gazing that had previously produced food no longer did so, dogs moved away, oriented sideways, lay down, walked around, sniffed and vocalized more, behaviors the authors described as frustration responses (Jakovcevic et al., 2013). In a related design, social buffering was tested: frustration responses increased when food was withheld, and the presence of the owner reduced them; dogs with the owner present also persisted longer in gazing at the experimenter (Dzik et al., 2025).

A milder form of disappointment is a reward that is smaller or worse than expected. In the first study of this in dogs, dogs that had received liver for gazing and were switched to pellets rejected food more often, gazed less than dogs that had always received pellets, and showed more withdrawal and lying down, the classic contrast effect (Bentosela et al., 2009). A later study in two populations of dogs did not find such an effect and even found a reverse one (Riemer et al., 2016). The evidence on reward downshifts in dogs is therefore inconsistent, while outright omission of an expected reward produced frustration responses in the studies cited (what determines the value of a reward).

For anticipatory behavior, the practical consequence is straightforward: strong anticipation creates the conditions for strong frustration. The more a dog has learned to expect a specific outcome at a specific moment, the more it has to lose when the outcome fails to arrive (what is known about frustration in dogs).

5.3 Telling Them Apart

Distinguishing positive anticipation from frustration in a dog is harder than distinguishing calm from excitement. The facial coding studies showed reliable group-level differences but no single action that classified the state on its own (Bremhorst et al., 2022). Physiological measures do not solve the problem either. In a controlled study of 60 dogs from three breeds, six ten-minute scenarios were designed to produce four combinations of valence and arousal, including toy play and treat throwing as positive high-arousal situations (Flint et al., 2024). Cortisol, ACTH, heart-rate variability, panting, whining and body shaking differentiated arousal only within the negative-valence scenarios, while heart rate performed across both, and qualitative behavioral assessment was the best candidate for detecting positive valence.

A review of positive-emotion assessment in dogs reached a similar conclusion: no single behavioral or physiological indicator has been validated as a reliable, context-independent marker of positive emotion in dogs (Csoltova & Mehinagic, 2020). The excited dog at the door is therefore the hardest case to read with confidence, because positive high arousal is exactly the state the available indicators capture least well (why behavior is not a direct readout of emotion).

Context and time course help most. Positive anticipation typically rises toward an event and falls when it arrives. Frustration typically follows a delay or a blocked attempt and often includes behavior directed away from the goal, such as the moving away, lying down and sniffing described in the omission studies.

6. Predictability and Signaled Events

6.1 Temporal and Signaled Predictability

Anticipation depends on predictability. Without some way of knowing that a reward is coming, there is nothing to anticipate. A review of predictability in captive animals distinguished temporal predictability, where events happen at regular times, from signaled predictability, where a specific cue announces an event (Bassett & Buchanan-Smith, 2007). The effects were not uniform. They depended on the type of event, the type of cue, the species and what other forms of control the animals had. Overall, predictable conditions were generally associated with better welfare than unpredictable ones, and unpredictability was one of the conditions under which stereotypies developed. That review covered captive animals other than dogs, and it supports a principle rather than a specific routine.

The distinction matters because the two produce different waiting periods. A dog fed at the same time every day may start anticipating well before any signal, while a dog whose meals are announced by a specific signal can show a more compressed anticipation that begins with it.

6.2 Reliable and Equivocal Cues

Krebs and Watters (2019) refined the concept of signals for welfare purposes. A reliable cue, in their terms, "consistently precedes an event such that it indicates the coming of the event". An equivocal cue is one that resembles two or more reliable cues that carry opposite information, so that the animal cannot tell which outcome it predicts. The authors argue that animals tend to choose predictability over unpredictability and that loss of predictability negatively affects welfare.

Many household signals are equivocal from a dog's point of view. Picking up keys may mean a walk or that the owner is leaving alone. A dog that has learned that such actions sometimes predict a walk will anticipate whenever they occur and meet a violated expectation whenever they lead to something else, which may explain repeated bursts of excitement followed by apparent disappointment.

6.3 Unintentional Cues and Places

Animals do not only learn the cues people intend to give. Krebs and colleagues note that animals can learn to associate unintentional cues, such as the presence of a keeper, with positive events, and that they learn to associate specific places with predictable events occurring there (Krebs et al., 2022). An animal that waits in the same spot before every feeding may look as though it prefers that spot, when in fact it is anticipating what usually happens there.

In a household, the equivalent is the dog that lies by the front door in the late afternoon or sits in the kitchen an hour before dinner: not necessarily a favorite spot, but the place where the next good thing tends to happen.

6.4 Anticipation or Stereotypy

One of the most important practical warnings in this literature is that intense anticipation can look like abnormal repetitive behavior. The review literature states that "intense anticipation may appear similar to an abnormal repetitive behavior such as pacing", and that considering the timing, context and location of the behavior may make it possible to tell them apart (Krebs et al., 2022). The horse data show the overlap from the other side: stereotypic behavior increased in the run-up to feeding in ration-fed horses (Mata et al., 2024).

Timing is the key. Anticipatory behavior is tied to an expected event and drops when the event arrives. Repetitive behavior that occurs in many contexts, persists after the event or is difficult to interrupt is a different pattern and deserves a different analysis (what the research on compulsive behavior shows).

7. What Owners See

7.1 Excitement Before a Walk

Pre-walk excitement typically begins with an early cue, such as the owner picking up a jacket or reaching for the leash, and escalates as further cues follow. Owners in the survey described such situations with barking, a raised head, wide bright eyes, pricked ears, tail wagging and an open mouth (Buckland et al., 2014). The facial coding studies suggest that, in a positive anticipation situation, the ear movement coded as ears adductor is one of the more specific signs, although it is often absent (Bremhorst et al., 2022).

Nothing in the research suggests that a dog that is excited before walks has a welfare problem. The reward-sensitivity hypothesis does not say that anticipation is bad; it says that unusually intense anticipation may be informative when compared with what is normal for that animal. Most pre-walk excitement in dogs is exactly what one would expect from an animal anticipating an event it values.

7.2 Barking at Feeding Time

Barking, whining or pacing before meals is another everyday form of anticipation. It is also one of the most common sources of irritation for owners. From the perspective of anticipatory behavior research, two questions are useful. The first is about time course: does the behavior start at a consistent interval before feeding and stop once the dog has eaten? The second is about context: is feeding time a highlight in an otherwise uneventful day, or one of many rewarding events?

A third question concerns learning. If barking has regularly been followed by the meal, it is also a reinforced behavior, a training history rather than a welfare signal. Separating the two requires looking at what the behavior produces, not only at how intense it is (what is known about vocal communication in dogs).

7.3 Reading Frequency and Intensity

If the zoo framework is applied to dogs, the most informative pattern would be the combination of frequency and intensity. A dog that shows moderate anticipation toward several events in the day, such as walks, meals, play sessions and the return of family members, would fit the picture of an animal with many rewarding experiences. A dog that is largely inactive through the day and then shows extreme anticipation toward a single event would fit the picture that the hypothesis treats as a possible warning sign (Watters, 2014, as cited in Krebs et al., 2022; Krebs & Watters, 2019).

This applies a zoo framework to dogs; it is a way to look, not a validated way to diagnose.

8. In Practice

8.1 Signals and Routines

Because anticipation is built on predictability, the most direct lever owners have is the structure of signals and routines. The predictability research suggests that reliable signals are generally preferable to equivocal ones (Bassett & Buchanan-Smith, 2007; Krebs & Watters, 2019), and the frustration research shows that expectations that are created and then not fulfilled produce frustration responses (Jakovcevic et al., 2013).

In practice: A clear, specific signal that reliably precedes the walk, such as a particular word or taking a specific leash from its hook, often helps dogs distinguish between "we are going out together" and "someone is leaving". Actions that are often equivocal for dogs, such as picking up keys, then lose some of their predictive power. This is an observation from training practice, not a finding from controlled studies.

Very rigid timing may stretch the anticipation period, while signaled routines can compress it. Which arrangement suits a dog depends on the household and the dog; the research does not establish one as universally better (how predictability and variety interact in enrichment).

8.2 Calming Rituals Before Walks

Many owners respond to pre-walk excitement by waiting for calm before opening the door. This is compatible with what is known about anticipation and frustration, with one condition: the waiting must be short and the criterion achievable. A delay that the dog cannot resolve is exactly the structure that produces frustration in the experimental studies.

In practice: A short, predictable pre-walk ritual, such as a sit while the leash is clipped on and a brief pause at the door, followed by the walk as soon as the dog manages it, usually reduces the intensity of excitement over time. What helps is that the dog can bring about the walk through its own behavior, which links to the findings on control (McGowan et al., 2014). This comes from practice; it has not been tested as a protocol.

8.3 When Strong Anticipation May Hint at a Deficit

The reward-sensitivity hypothesis offers one useful question for owners: is a single event carrying too much weight in the dog's day? If a dog spends most of the day with little to do and then shows extreme, hard-to-interrupt excitement before the one walk it gets, the intensity may say something about the rest of the day.

In practice: When anticipation seems excessive, it is often worth looking first at the overall day rather than at the anticipation itself: how much exercise, exploration, social contact, chewing, sniffing and rest the dog gets, and whether these are spread across the day. If those needs are better met, extreme anticipation of a single event often becomes less intense. This is an observation from practice; in dogs, it has not been shown experimentally.

A closer look is warranted when anticipation begins very early and lasts long, does not drop when the event arrives, comes with repetitive movement that also appears in other contexts, or turns into frequent frustration when routines change. Such patterns can have many causes, including pain, so persistent changes are a reason to involve a veterinarian or a qualified behavior professional (how stress is measured in dogs).

8.4 Avoid Pathologizing Normal Excitement

The opposite mistake is equally possible. A dog that is delighted when its owner reaches for the leash is not showing a symptom. The hypothesis was developed to compare groups of animals in different environments, not to evaluate the joy of a single dog before a walk. In dogs, the hypothesis has not been tested at all.

In practice: The most useful reference point is the dog's own baseline. A sudden increase or a clear decrease in anticipation toward events the dog usually enjoys is more informative than the absolute level of excitement. A dog that no longer reacts to the leash or the food bowl the way it used to may be unwell, in pain or under strain, and that change deserves attention. This reading of change against baseline comes from practice, not from validated studies.

9. Summary at a Glance

Anticipatory behavior is behavior shown in the appetitive phase, before an expected outcome arrives. It typically rises as the event approaches and drops off when the event happens (Krebs et al., 2022). The reward-sensitivity hypothesis proposes that anticipation reflects the state of the brain's reward systems and therefore the balance between positive and negative experiences (Spruijt et al., 2001; van der Harst & Spruijt, 2007).

Heightened anticipation has been interpreted as a sign of increased reward sensitivity and poorer welfare. In rats, standard-housed animals showed more anticipatory behavior than rats from a semi-naturalistic environment, and drug treatments did not change this (Makowska & Weary, 2016). Mice developed anticipatory behavior toward a valued playpen (Ratuski et al., 2021), and anticipatory behavior has been studied as a welfare tool in foxes, mink, lambs, zoo animals, dolphins and horses (Moe et al., 2006; Hansen & Jeppesen, 2006; Anderson et al., 2015; Watters, 2014; Clegg & Delfour, 2018; Mata et al., 2024).

In dogs, the hypothesis has not been tested. Dog research shows facial differences between positive anticipation and frustration (Bremhorst et al., 2019, 2022), behavioral differences when dogs produce a reward themselves (McGowan et al., 2014), reunion effects after time alone (Rehn & Keeling, 2011), and that owners link excitement mainly to anticipation of walks and food (Buckland et al., 2014). Anticipation, frustration and arousal are distinct, and the available indicators capture positive high arousal least well (Flint et al., 2024; Csoltova & Mehinagic, 2020).

10. Research Gaps and Critical Appraisal

10.1 Transfer Across Species

The central concept of this article was developed in rats and applied in mice, farmed and fur animals, zoo animals, dolphins and horses. Dogs share the basic reward circuitry with these species, but their environments, social relationships and daily routines with people are different. The mechanism background is even further removed: the wanting-liking distinction comes from rodent research (Berridge & Robinson, 1998), and the prediction-error pattern was recorded from monkey neurons (Schultz et al., 1997). None of these mechanisms has been measured in dogs.

10.2 Missing Dog Studies

No dog study has used the cue-reward interval paradigm to compare anticipatory behavior between groups of dogs in different living conditions. The dog studies that touch on anticipation answer other questions: what faces look like during positive anticipation and frustration, how dogs respond to producing a reward themselves, what happens at reunion and how owners perceive their dogs. A study comparing, for example, kennel-housed dogs before and after an enrichment program, or companion dogs with different daily routines, using a standardized cue-reward interval, would be the obvious next step.

10.3 Sample Sizes and Designs

The samples are small. The beagle study on achievement used 12 dogs (McGowan et al., 2014), the reunion study 12 dogs (Rehn & Keeling, 2011), the facial replication 28 dogs (Bremhorst et al., 2022) and the horse study six animals (Mata et al., 2024). The dolphin data come from animals under human care. The largest dog dataset relevant here is the owner survey with 445 respondents (Buckland et al., 2014), which measures perception rather than behavior. The facial coding replication was carried out by the same research group as the original study, which is valuable but not independent confirmation.

10.4 Measures and Their Meaning

Most anticipatory behavior studies count behavioral transitions or the frequency of behaviors in a defined interval. Higher counts are interpreted as higher reward sensitivity, but this interpretation is part of the framework rather than an independent measurement. The rat study showed that the form of anticipation, rearing versus sitting and watching, can differ between groups as much as its quantity (Makowska & Weary, 2016), which means that a simple count may miss important information. In dogs, physiological indicators differentiated arousal mainly within negative-valence situations (Flint et al., 2024), so they cannot easily confirm whether intense anticipation is positive.

10.5 The Direction of the Signal

The hypothesis predicts that both very high and very low anticipation can indicate a problem. That makes it flexible, but it also makes it difficult to falsify for an individual animal. Without a known baseline, the same level of anticipation could be read as healthy interest, as a sign of deprivation or as reduced responsiveness. The rat drug study did not support a simple equation of heightened anticipation with depression or anxiety (Makowska & Weary, 2016), and the anhedonia end of the curve has not been demonstrated in dogs.

10.6 Household Conditions

Almost all controlled data come from laboratories, farms, zoos and stables. Companion dogs live with partly unpredictable routines, receive many unintentional cues and often have a training history in which excited behavior was reinforced, which makes household anticipation harder to interpret. Whether the zoo-derived pattern of "frequent, moderate anticipation" applies to them is unknown.

11. Conclusion

The excitement a dog shows before a walk or a meal is not trivial. Welfare science has developed a serious hypothesis about what such behavior may reveal: the intensity of anticipation reflects how sensitive the animal's reward systems are, and that sensitivity depends on the balance of positive and negative experiences in its life. The idea has empirical support in rats, has been applied in mice, farmed animals, zoo animals and horses, and fits well with what is known about the neurobiology of wanting and prediction.

For dogs, the hypothesis remains untested. What has been measured in dogs shows that positive anticipation is a real, observable state, that it can be distinguished from frustration at group level but not reliably from single signs, and that owners associate their dogs' excitement mainly with walks and food. The most responsible use of the concept is therefore modest: as a reason to look at the whole day when a single event seems to carry too much weight, and as a reminder that a change in a dog's anticipation is more informative than the anticipation itself. Normal excitement before good things is part of a good life, not a symptom.

Key Insights (Takeaways)

• Anticipatory behavior is behavior shown before an expected reward; it typically rises as the event approaches and drops when it arrives (Krebs et al., 2022).

• The reward-sensitivity hypothesis proposes that anticipation reflects the state of reward systems and the balance between positive and negative experiences (Spruijt et al., 2001; van der Harst & Spruijt, 2007).

• In rats, standard-housed animals showed more anticipatory behavior than rats from a semi-naturalistic environment, interpreted as higher reward sensitivity and poorer welfare (Makowska & Weary, 2016).

• The hypothesis has not been tested in dogs; most evidence comes from rats, mice, farmed animals, zoo animals, dolphins and horses.

• Dog faces differ between positive anticipation and frustration at group level, but no single facial action classifies the state reliably (Bremhorst et al., 2019, 2022).

• Anticipation, frustration and arousal are different concepts; positive high arousal is the state current indicators capture least well (Flint et al., 2024; Csoltova & Mehinagic, 2020).

• Reliable signals and achievable routines are compatible with the research on predictability and frustration (Bassett & Buchanan-Smith, 2007; Krebs & Watters, 2019).

• A change in a dog's usual anticipation is more informative than the level of excitement itself; normal pre-walk joy is not a symptom.

References

Anderson, C., Yngvesson, J., Boissy, A., Uvnäs-Moberg, K., & Lidfors, L. (2015). Behavioural expression of positive anticipation for food or opportunity to play in lambs. Behavioural Processes, 113, 152–158. https://doi.org/10.1016/j.beproc.2015.02.003

Bassett, L., & Buchanan-Smith, H. M. (2007). Effects of predictability on the welfare of captive animals. Applied Animal Behaviour Science, 102(3–4), 223–245. https://doi.org/10.1016/j.applanim.2006.05.029

Bentosela, M., Jakovcevic, A., Elgier, A. M., Mustaca, A. E., & Papini, M. R. (2009). Incentive contrast in domestic dogs (Canis familiaris). Journal of Comparative Psychology, 123(2), 125–130. https://doi.org/10.1037/a0013340

Berridge, K. C., & Robinson, T. E. (1998). What is the role of dopamine in reward: Hedonic impact, reward learning, or incentive salience? Brain Research Reviews, 28(3), 309–369. https://doi.org/10.1016/S0165-0173(98)00019-8

Bremhorst, A., Mills, D. S., Würbel, H., & Riemer, S. (2022). Evaluating the accuracy of facial expressions as emotion indicators across contexts in dogs. Animal Cognition, 25(1), 121–136. https://doi.org/10.1007/s10071-021-01532-1

Bremhorst, A., Sutter, N. A., Würbel, H., Mills, D. S., & Riemer, S. (2019). Differences in facial expressions during positive anticipation and frustration in dogs awaiting a reward. Scientific Reports, 9, 19312. https://doi.org/10.1038/s41598-019-55714-6

Buckland, E. L., Volk, H. A., Burn, C. C., & Abeyesinghe, S. M. (2014). Owner perceptions of companion dog expressions of positive emotional states and the contexts in which they occur. Animal Welfare, 23(3), 287–296. https://doi.org/10.7120/09627286.23.3.287

Clegg, I. L. K., & Delfour, F. (2018). Cognitive judgement bias is associated with frequency of anticipatory behavior in bottlenose dolphins. Zoo Biology, 37, 67–73. https://doi.org/10.1002/zoo.21400

Clegg, I. L. K., Rödel, H. G., Boivin, X., & Delfour, F. (2018). Looking forward to interacting with their caretakers: Dolphins' anticipatory behaviour indicates motivation to participate in specific events. Applied Animal Behaviour Science, 202, 85–93. https://doi.org/10.1016/j.applanim.2018.01.015

Csoltova, E., & Mehinagic, E. (2020). Where do we stand in the domestic dog (Canis familiaris) positive-emotion assessment: A state-of-the-art review and future directions. Frontiers in Psychology, 11, 2131. https://doi.org/10.3389/fpsyg.2020.02131

Dzik, M. V., Jousset, M., Vietti, J., & Bentosela, M. (2025). Effects of the presence of a human on frustration in domestic dogs: Is there a social buffering phenomenon? Applied Animal Behaviour Science, 283, 106503. https://doi.org/10.1016/j.applanim.2025.106503

Flint, H. E., Weller, J. E., Parry-Howells, N., Ellerby, Z. W., McKay, S. L., & King, T. (2024). Evaluation of indicators of acute emotional states in dogs. Scientific Reports, 14, 6406. https://doi.org/10.1038/s41598-024-56859-9

Hansen, S. W., & Jeppesen, L. L. (2006). Temperament, stereotypies and anticipatory behaviour as measures of welfare in mink. Applied Animal Behaviour Science, 99, 172–182. https://doi.org/10.1016/j.applanim.2005.10.005

Jakovcevic, A., Elgier, A. M., Mustaca, A. E., & Bentosela, M. (2013). Frustration behaviors in domestic dogs. Journal of Applied Animal Welfare Science, 16(1), 19–34. https://doi.org/10.1080/10888705.2013.740974

Krebs, B. L., Chudeau, K. R., Eschmann, C. L., Tu, C. W., Pacheco, E., & Watters, J. V. (2022). Space, time, and context drive anticipatory behavior: Considerations for understanding the behavior of animals in human care. Frontiers in Veterinary Science, 9, 972217. https://doi.org/10.3389/fvets.2022.972217

Krebs, B. L., Torres, E., Chesney, C., Kantoniemi Moon, V., & Watters, J. V. (2017). Applying behavioral conditioning to identify anticipatory behaviors. Journal of Applied Animal Welfare Science, 20, 155–175. https://doi.org/10.1080/10888705.2017.1283225

Krebs, B. L., & Watters, J. V. (2019). Assessing and enhancing the welfare of animals with equivocal and reliable cues. Animals, 9(9), 680. https://doi.org/10.3390/ani9090680

Makowska, I. J., & Weary, D. M. (2016). Differences in anticipatory behaviour between rats (Rattus norvegicus) housed in standard versus semi-naturalistic laboratory environments. PLOS ONE, 11(1), e0147595. https://doi.org/10.1371/journal.pone.0147595

Mata, F., Boyton, G., & Young, T. (2024). Anticipatory behaviour during the approach to feeding times as a measure of horse welfare. Animals, 14(24), 3677. https://doi.org/10.3390/ani14243677

McGowan, R. T. S., Rehn, T., Norling, Y., & Keeling, L. J. (2014). Positive affect and learning: Exploring the "Eureka Effect" in dogs. Animal Cognition, 17(3), 577–587. https://doi.org/10.1007/s10071-013-0688-x

Mendl, M., Burman, O. H. P., & Paul, E. S. (2010). An integrative and functional framework for the study of animal emotion and mood. Proceedings of the Royal Society B: Biological Sciences, 277(1696), 2895–2904. https://doi.org/10.1098/rspb.2010.0303

Moe, R. O., Bakken, M., Kittilsen, S., Kingsley-Smith, H., & Spruijt, B. M. (2006). A note on reward-related behaviour and emotional expressions in farmed silver foxes (Vulpes vulpes): Basis for a novel tool to study animal welfare. Applied Animal Behaviour Science, 101, 362–368. https://doi.org/10.1016/j.applanim.2006.02.004

Ratuski, A. S., Makowska, I. J., Dvorack, K. R., & Weary, D. M. (2021). Using approach latency and anticipatory behaviour to assess whether voluntary playpen access is rewarding to laboratory mice. Scientific Reports, 11, 18683. https://doi.org/10.1038/s41598-021-98356-3

Rehn, T., & Keeling, L. J. (2011). The effect of time left alone at home on dog welfare. Applied Animal Behaviour Science, 129(2–4), 129–135. https://doi.org/10.1016/j.applanim.2010.11.015

Riemer, S., Ellis, S. L. H., Ryan, S., Thompson, H., & Burman, O. H. P. (2016). A reappraisal of successive negative contrast in two populations of domestic dogs. Animal Cognition, 19(3), 471–481. https://doi.org/10.1007/s10071-015-0947-0

Schultz, W., Dayan, P., & Montague, P. R. (1997). A neural substrate of prediction and reward. Science, 275(5306), 1593–1599. https://doi.org/10.1126/science.275.5306.1593

Spruijt, B. M., van den Bos, R., & Pijlman, F. T. A. (2001). A concept of welfare based on reward evaluating mechanisms in the brain: Anticipatory behaviour as an indicator for the state of reward systems. Applied Animal Behaviour Science, 72(2), 145–171. https://doi.org/10.1016/S0168-1591(00)00204-5

Van der Harst, J. E., & Spruijt, B. M. (2007). Tools to measure and improve animal welfare: Reward-related behaviour. Animal Welfare, 16(S1), 67–73. https://doi.org/10.1017/S0962728600031742

Watters, J. V. (2014). Searching for behavioral indicators of welfare in zoos: Uncovering anticipatory behavior. Zoo Biology, 33, 251–256. https://doi.org/10.1002/zoo.21144