Positive contact with one’s companion animal refers to the positive interactions that pet owners experience with their companion animal, as interspecific relations that involve members of distinct species. This relational factor captures both the frequency as well as the reciprocity and affective valence of these contacts (Davies et al., 2011), in line with the intergroup contact literature. Given that the positive contacts that take place between the members of different social groups predict an increased tendency to help the members of another group (Johnston & Glasford, 2018), herein we test if such contacts also predict more affection and caring behaviors toward one’s own companion animal specifically, as a particularly proximal type of interspecific relationship.

The relational factor pertaining to human-companion animal compatibility refers to the similarities that are shared by a guardian and their companion animal (El-Alayli et al., 2006). Based on similarity attraction principles in social psychology, according to which perceived or actual similarity facilitates the emergence and maintenance of interpersonal relationships (Byrne, 1971), sharing similarities with one’s companion animal could have positive implications, particularly for the guardians’ psychological well-being. Prior research has indeed found that the more pet owners perceive that their self-characteristics are similar to their companion animal’s, the higher their well-being (e.g., lower negative affect; higher life satisfaction; El-Alayli et al., 2006). Similarly, the more dog owners perceive that their personal values (i.e., in terms of hedonism and achievement) are similar to their dog’s, the higher their life satisfaction (Sneddon et al., 2022). Furthermore, perceiving higher behavioral compatibility between one’s own personality and the personality of one’s companion animal has been associated with higher well-being among pet owners (George et al., 1998). Yet, and because similarity is a consistent predictor of relationship satisfaction and quality in human-human interpersonal relationships (Wilson & Cousins, 2003), human-companion animal compatibility could also predict more affection and caring for one’s companion animal.

Relatedly, the quality of the human-companion animal relationship captures how satisfied guardians feel about their relation with their companion animal and experience positive emotions in the context of this relationship (Cassels et al., 2017). Based on the human-human interpersonal relations literature, this relational factor could also potentially have beneficial implications for the guardians’ well-being, as well as fuel more affection and caring for their companion animal, given that experiencing positive emotions and satisfaction within the human-companion animal relation should facilitate commitment and care.

The type of attachment to one’s companion animal has also been found to play a role in predicting guardians’ psychological well-being. Studies that have applied psychological attachment theory (Bowlby, 1969) to the realm of human-companion animal relations have uncovered negative associations between an anxious attachment to one’s companion animal and guardians’ well-being (e.g., higher psychological distress and anxiety: Zilcha-Mano et al., 2011; lower happiness but higher depression and loneliness: Ellis et al., 2024). While such associations between an anxious attachment to one’s companion animal and the well-being outcomes are expected to replicate herein, we will also explore if having an anxious or an avoidant attachment type has implications for guardians’ tendencies to feel affection and care for their companion animal. Although these types of attachment to one’s companion animal were assessed given their theoretical anchor in the human attachment literature, it is worth acknowledging that there exists a diversity of different taxonomies and ways of measuring the human-companion animal attachment construct (e.g., Lexington Attachment to Pets Scale, Pet Attachment Questionnaire, CENSHARE Pet Attachment Survey).

Based on the concept of unconditional positive regard, which involves feeling valued, respected, and accepted unconditionally and non-judgmentally (Rogers, 1959), unconditionally accepting one’s companion animal implies accepting and supporting one’s animal unconditionally and non-judgmentally. Companion animals are often perceived as providing unconditional support and affection to their humans (e.g., Wiens et al., 2024), a message also conveyed in the media (e.g., Lamic, 2022). Some research has confirmed that feeling unconditionally supported and accepted by one’s companion animal has beneficial implications for guardians’ well-being (e.g., Amiot & Bastian, 2023; Teo & Thomas, 2019). Yet, and to our knowledge, no research has focused on the other direction of this unconditional support and acceptance, namely on guardians’ unconditional acceptance of their companion animal. This relational factor should not only have positive implications for feeling affection and caring for one’s companion animal, but it could also predict higher psychological well-being, given that supporting one’s companion animal’s needs—which implies a prosocial inclination toward one’s animal—has been found to have well-being benefits for pet owners per se (Kanat-Maymon et al., 2021) .

A power analysis conducted using G*Power prior to conducting the study revealed that a sample size of 471 participants was needed to capture small effect sizes (f² = .04) for multiple regressions with 14 independent variables, α = .05, 1-β = .80. We recruited 585 American pet owners via Prolific. Participants were compensated with 3 British pounds (approximately US$4) upon completion of the study. A total of 50 participants were excluded because: (1) the participants withdrew their participation (n = 23); or (2) failed to provide the right answers to the two attention check items (n = 2); or (3) did not reach the end of the questionnaire or were timed-out by Prolific (n = 9); or (4) did not consent to taking part in the study (n = 1). In addition, 14 participants were excluded because they were not current pet owners, and one participant was excluded given they were not residing in the U.S. The final sample included 535 participants. In terms of gender, 34.2% of participants were males, 62.1% were females, 3.2% were non-binary, 0.4% indicated ‘other’ for their gender, and 0.2% preferred not to answer. Approximately 21.7% of the sample were older than 50 years. More details on the sociodemographic characteristics of the sample are presented in Supplementary Table S1.

The study was approved by the Ethics Committee involving Human Participants of the University of (anonymised) (certificate number 2023-4883). The study was conducted in accordance with the guidelines of the Canadian Tri-Council Policy for the Ethical Conduct of Research Involving Human Participants. The study was administered using the Qualtrics software. After reading the information and consent form detailing the objectives of the study, confidentiality, possible risks involved, and ethical clearance, participants provided sociodemographic information, including information about their companion animals. The questionnaire then included different companion animal-related measures; if participants had more than one companion animal, they were instructed to refer to their preferred companion animal when responding to this section of the questionnaire. As well, participants were asked to either indicate what is the species of their current companion animal, or, if they currently had more than one companion animal, what is the species of their preferred companion animal. Participants then completed the psychological well-being measures.

A detailed description of the measures and their reliabilities, along with the Qualtrics questionnaire used in the study, are provided in the supplementary materials document.

Zero-order correlations between the relational factors (i.e., contact with one’s companion animal, human-companion animal compatibility, quality of the human-companion animal relation, avoidant and anxious attachment to one’s companion animal, unconditional acceptance of one’s animal), the prosocial inclinations toward one’s companion animal variables (i.e., caring for one’s companion animal and affection for one’s animal), and the well-being variables are reported in Table 1. As can be seen in this Table, many significant associations were uncovered between the relational factors and the prosocial inclinations toward one’s companion animal variables. Specifically, contact with one’s companion animal, human-companion animal compatibility, quality of the human-companion animal relation, as well as unconditional acceptance of one’s companion animal, all correlated positively with caring for one’s companion animal and with affection for the companion animal. Avoidant attachment to one’s companion animal correlated negatively with these two prosocial inclination variables, while anxious attachment to one’s companion animal showed positive associations.

As per the associations between the relational factors and the psychological well-being variables, human-companion animal compatibility correlated significantly with nearly all of the well-being variables, showing that the higher human-companion animal compatibility, the more positive the guardians’ well-being (i.e., higher life satisfaction, life meaning, positive relations with others, vitality, and overall well-being, but lower stress). A similar pattern of correlations was uncovered for the quality of the human-companion animal relation and the well-being variables. An inverse pattern of correlations was found however for anxious attachment to one’s companion animal, which correlated with lower well-being (i.e., lower life satisfaction, life meaning, positive relations with others, vitality, and overall well-being, but higher stress and search for meaning), in line with prior work (e.g., Zilcha-Mano et al., 2011). Avoidant attachment to one’s companion animal presented a mixed pattern of associations with the well-being variables, showing positive correlations with life satisfaction and with stress, but a negative correlation with positive relations with others.

Interestingly, a few significant correlations were observed between the prosocial inclinations toward one’s companion animal variables and some of the well-being variables; this suggests that caring and feeling affection for one’s animal can be accompanied by higher well-being (e.g., higher vitality, more positive relations with others, and higher overall well-being). Yet, the modest size of these associations also suggests that these mutual benefits are far from automatic among guardians, and that third variables could moderate the association between caring for one’s companion animal and deriving well-being benefits from such caring behaviors.

To systematically test how each of the relational factors uniquely contributes to predicting the prosocial inclinations and the well-being variables relative to the others, and over and above the sociodemographic and social resources variables, hierarchical multiple regressions were conducted. In these regressions, Step 1 included the sociodemographic and social resources variables, and Step 2 included the relational factors. All variance inflation factors values were below 3, confirming that multicollinearity was not a problem in these analyses. Nevertheless, supplementary analyses, which account for the intercorrelations among the independent variables when predicting the outcomes (i.e., relative weight analyses) (Tonidandel & LeBreton, 2011) were conducted and are presented in the supplementary materials document. Although some differences could be noted across the findings observed in these supplementary analyses and those observed in the multiple regression analyses presented herein, such disparities are not uncommon in the literature (Tonidandel et al., 2009).

As can be seen in Table 2 (at Step 2 of the hierarchical regressions), when predicting affection for one’s companion animal, some of the sociodemographic and social resources variables significantly predicted this prosocial inclination outcome. Specifically, age predicted lower affection for one’s companion animal. Furthermore, relationship status, number of companion animals, and resources available to care for one’s companion animal predicted high affection for one’s companion animal. In terms of the relational factors, contact with one’s companion animal, quality of the human-companion animal relation, and unconditional acceptance of one’s companion animal—as relational factors denoting a more positive relationship with one’s animal—all positively predicted affection for one’s companion animal; anxious attachment with the companion animal also positively predicted affection for one’s companion animal. In contrast, avoidant attachment - which also denotes lower engagement toward one’s companion animal - predicted lower affection. When predicting caring for one’s companion animal, the resources available to care for the companion animal positively predicted this prosocial inclination variable, along with the relational variables of: contact with one’s companion animal, quality of the human-companion animal relation, and unconditional acceptance of one’s companion animal.

When predicting the well-being variables, and as can be seen in Table 3 (at Step 2 of the hierarchical regressions), some of the sociodemographic and social resources variables significantly predicted these variables. Specifically, age predicted lower stress and search for meaning. Education level positively predicted life satisfaction, life meaning, vitality, and overall well-being. Number of children also positively predicted life satisfaction, life meaning, vitality, and overall well-being. Relationship status (i.e., being in a relationship) positively predicted life satisfaction, life meaning, positive relationships with others, and overall well-being. Perceived social status predicted positively life satisfaction, life meaning, vitality, and overall well-being but lower stress. Finally, resources available to care for companion animals positively predicted life satisfaction, life meaning, positive relations with others, vitality, and overall well-being but lower stress.

Over and above these sociodemographic and social resources variables, the relational factors also contributed to predicting the well-being outcomes, as can be seen in Table 3. Specifically, human-companion animal compatibility predicted nearly all of the well-being variables (i.e., higher life satisfaction, meaning in life, positive relations with others, vitality, and overall well-being but lower stress). Quality of the human-companion animal relation predicted three well-being variables (i.e., higher life satisfaction, and vitality, but lower stress) as well as higher overall well-being. While unconditional acceptance of one’s companion animal negatively predicted search for meaning, contact with one’s companion animal positively predicted stress, suggesting less beneficial implications of such contacts for pet owners’ well-being. In terms of attachment, an anxious attachment to one’s companion animal predicted lower well-being (i.e., higher stress, search for meaning and lower overall well-being). In contrast, avoidant attachment to one’s companion animal positively predicted higher well-being (i.e., higher life satisfaction, life meaning, vitality, and overall well-being).

To test the moderating role of the relational factors in the associations between guardians’ caring for their companion animal and their psychological well-being, six moderation analyses (see Table 4) were conducted using PROCESS analyses (Model 1 ; Hayes, 2022) with bias-corrected 95% confidence intervals, 5,000 bootstrapped samples. The overall well-being measure was used as the dependent variable in these analyses. All variables used in these analyses were standardized. We probed the significant interaction by testing the conditional effect of caring for one’s companion animal at two levels of the moderator—i.e., one standard deviation (SD) below the mean and one SD above the mean.

As can be seen in Table 4, the main effects observed for the relational factors on guardians’ psychological well-being revealed positive effects for: human-companion animal compatibility and quality of the human-companion animal relation, and negative effects for: anxious attachment to one’s companion animal and unconditional acceptance of one’s companion animal. In terms of the moderating role of the relational factors, only one significant moderation effect was uncovered, for human-companion animal compatibility. Simple slope analyses revealed that the effect of caring for one’s companion animal on psychological well-being was not significant at low (M-1SD; β = -0.01, P = .871, 95% CI [−0.10, 0.08]), nor at moderate levels of human-companion animal compatibility (M; β = 0.06, P = .101, 95% CI [−0.01, 0.14]). However, at high levels of human-companion animal compatibility (M+1SD), caring for one’s companion animal had a significant positive effect on guardians’ well-being (β = 0.13, P = .017, 95% CI [0.02, 0.24]).

Three relational factors experienced within the human-companion animal relationship consistently predicted the guardians’ prosocial inclinations toward their companion animal (i.e., their affection for their companion animal and caring behaviors) in the hierarchical multiple regressions, namely: positive contact with the companion animal, quality of the human-companion animal relation, and unconditional acceptance of the companion animal. Both the quality of the human-companion animal relation and the positive contact with the companion animal variables capture the positive emotions (e.g., satisfaction, happiness, fun, joy) experienced in the relationship and while interacting with one’s companion animal. As well, unconditional acceptance of one’s companion animal involves a deep sense of acceptance of the animal. Together, these relational factors promoted a feeling of affection and love toward one’s companion animal and fueled a greater tendency to provide care and exert continuous efforts to ensure the well-being of the animal.

In addition, an anxious attachment to one’s companion animal predicted more affection toward this animal; this association aligns with prior work showing that feeling psychologically connected to and committed toward other animals—in the form of solidarity with animals—is associated positively with an anxious attachment to one’s companion animal (Amiot & Bastian, 2017), suggesting that anxiously attached individuals can also experience a high level of personal involvement and care toward their companion animal, possibly fueled by the fear of losing a valued companion or of harm to the human-companion animal relationship. In contrast, an anxious attachment to one’s companion animal, given the distance that this attachment type involves putting between oneself and one’s companion animal, predicted lower affection toward this animal.

When predicting the guardians’ own levels of psychological well-being, the relational factor pertaining to human-companion animal compatibility emerged as a particularly consistent predictor of the well-being variables. This positive role of human-companion animal compatibility in predicting pet owners’ well-being aligns with prior work (e.g., El-Alayli et al., 2006; George et al., 1998; Sneddon et al., 2022). Interestingly, this more ‘cognitive’ relational factor did not significantly predict guardians’ tendency to care or feel affection for their companion animal (i.e., as prosocial inclinations toward their animal). This pattern of findings suggests that this relational factor may provide a sense of familiarity, reassurance and comfort to the guardians which could also have beneficial implications for their own well-being, but without necessarily promoting positive attitudes and behaviors toward their companion animal per se.

It should be noted that the measure of human-companion animal compatibility employed herein involved perceptions of similarity reported by the guardians themselves. Our findings show that this subjective perception of being similar to one’s companion animal plays a clear role in predicting guardians’ well-being. Yet, more objective measures of similarity may not have yielded the same findings. Future research could use more objective measures of human-companion animal similarity, and directly compare the role of perceived similarity as reported by different informants (e.g., owner, breeder, family members) in predicting guardians’ well-being. Similarity in the ‘eye of the beholder’—possibly also based on small cues observed on a day-to-day basis when interacting with one’s companion animal—may likely remain a potent predictor of well-being in such future work.

Quality of the human-companion animal relation, however, predicted both higher affection and caring for one’s companion animal, as well as higher well-being for guardians, possibly because of the positive emotions (e.g., satisfaction, happiness) experienced when reporting higher relationship quality with one’s companion animal. Indeed, the experience of positive emotions, as ‘broadening’ emotions, has been associated with more prosocial tendencies toward others, but it could also have positive implications for individual well-being (Fredrickson, 2013). As well, unconditional acceptance of one’s companion animal significantly positively predicted both affection and caring for one’s animal, along with a lower search for meaning, as an indicator of ill-being.

In line with prior research, anxious attachment to one’s companion animal was associated with lower well-being in the current study (e.g., Zilcha-Mano et al., 2011; Ellis et al., 2024). However, avoidant attachment to the companion animal was associated with higher well-being, and more specifically: higher life satisfaction, meaning in life, and vitality. It should be noted that our correlations uncovered a mixed pattern of associations between avoidant attachment to one’s companion animal and some well-being variables (i.e., positive associations with life satisfaction but also with stress, and a negative association with positive relations with others). This suggests ambiguity in the role played by an avoidant attachment to one’s companion animal. Because it involves establishing a form of distance with one’s animal, this attachment style may actually help to ‘free’ guardians from being overinvested in their relationship with their companion animal and potentially allow them to nurture their own individual well-being. Indeed, in a recent systematic review of the studies testing the associations between attachment to companion animals and depression (Ellis et al., 2024), when attachment to one’s companion animal is conceptualised and assessed as a strong attachment bond between pet owners and their animals, 12 studies have found a positive association between such attachment and depression, while 4 studies have found a negative association, suggesting some possible psychological benefits from experiencing a less intense bond with one’s companion animal.

Interestingly, the positive contact with companion animal relational factor predicted higher stress among guardians, while it also positively predicted the variables pertaining to affection and caring for one’s companion animal. This relational variable hence appears particularly beneficial to these prosocial inclinations toward one’s animal, as behavioral and attitudinal variables that involve intergroup helping (i.e., companion animals being from distinct species relative to humans), but not to human well-being per se.

The sociodemographic factors as well as the variables capturing the resources available to care for one’s companion animal also played a role in predicting pet owners’ prosocial inclinations toward their companion animal and their own well-being. Specifically, age predicted lower affection for one’s companion animal, suggesting more affinity with animals among younger participants. Furthermore, relationship status (being in a committed relationship), number of companion animals, and resources available to care for one’s companion animal predicted higher affection for one’s companion animal, suggesting that having a social network comprised of both humans and animals could facilitate feelings of affection for one’s companion animal per se. Resources available to care for the companion animal also positively predicted caring for one’s animal.

The sociodemographic and social resources variables significantly predicted the well-being variables. The associations uncovered herein generally aligned with prior research, showing that: being older is associated with higher well-being and better mental health (Orpana, 2008); having higher socioeconomic status, education level, and being married is associated with higher well-being (Diener et al., 1993; Orpana, 2008). While being a parent has been associated with higher well-being in some studies but not others (Nelson et al., 2014), it was associated positively with some of the well-being measures in the current sample (i.e., life satisfaction, meaning in life, vitality, overall well-being). In line with prior work, perception of social status also positively predicted well-being (Adler et al., 2000). As well, access to more resources to care for one’s companion animal predicted higher well-being.

While research investigating human-companion animal relations has focused mainly on human well-being and health, an important perspective to take is the companion animal’s welfare, and the possible interplay between the guardians’ own well-being and their tendency to care for their companion animal. When exploring which relational factor may facilitate a positive linkage between the guardians’ tendency to care for their companion animal and their own psychological well-being, we found a significant role for human-companion animal compatibility. Specifically, guardians who reported higher levels of compatibility with their companion animal displayed a significant positive association between caring for their animal and their own psychological well-being. Conceptually, human-companion animal compatibility itself involves perceiving a link between one’s own personal characteristics and the characteristics of one's companion animal. In this sense, this relational variable could facilitate deriving a personal well-being advantage from caring for one’s (more similar) companion animal and experiencing benefits from such behaviors, specifically by creating a cognitive passage between the self and the companion animal, and easing this flow of benefits, from one’s companion animal to oneself. Despite being a promising variable to investigate to better understand human-companion animal relations, which could also potentially be used to ensure a good human-companion animal ‘match’ when adopting a new animal, knowledge on human-companion animal compatibility remains limited. The current findings extend our knowledge of the role of such compatibility for pet owners’ well-being, while also taking into account the direct linkages that exist between human and animal wellness, in line with emerging research on the interplay between companion animal and owner health and wellness (e.g., Koskela et al., 2024).

While the current study relied on a large and diverse dataset, the design employed remains correlational, and no causality can be inferred based on the current findings. Longitudinal designs should hence be employed in future research to provide an indication of the direction of these associations. Furthermore, other measures or methods for capturing mutuality within human-companion animal relationships could potentially be developed and employed. For example, self-reported or objective measures (e.g., veterinary assessments) could be used to assess animals’ well-being more directly, and the associations with their guardian’s well-being could be directly tested, also in line with the One Health Approach. As well, future work could further investigate the role played by the guardians’ unconditional acceptance of their companion animal, including how far it is appropriate to accept some of their animal’s behaviors. Indeed, some behaviors are clearly unacceptable (e.g., aggressivity) and need to be modified and managed by the guardians. While the current measure of unconditional acceptance did not capture such level of details, future research should keep these considerations in mind and develop fined-grained measures of unconditional acceptance.

Importantly, the current findings will need to be replicated using larger representative samples. Most of our predictions were confirmed. Overall, findings aligned with prior empirical work on human-companion animal relations to suggest a generally beneficial role for the relational factors tapping into more positive and mutual human-companion animal relationships. However, future studies will need to reproduce the specific pattern of associations uncovered herein to ensure its generalizability. Such studies could include a cross-cultural component to capture possible similarities but also differences across cultures and samples. Confirming the stability of the current findings in other samples will also be important given the high intercorrelations observed among the predictors included in our regression models.

Indeed, and despite the acceptable variance inflation factor indices observed in the current study, some of the independent variables were highly correlated (i.e., up to a correlation of .70 observed between some of the relational factors). In the current research, we were particularly interested in teasing the role of these specific relational factors apart and identifying which specific ones predicted both human well-being outcomes and caring for one’s companion animal. Given this objective, a multiple regression approach, which kept each of these relational factors distinct and allowed a fine-grain analysis of their roles in predicting these outcomes, was hence most appropriate. However, our statistical approach may have not fully dealt with the overlapping variance existing between the independent variables, nor considered issues such as colliders, confounders, and suppressors. Future research could directly test the underlying factorial structure of these relational factors, and whether these specific relational factors can be represented by one general latent factor, as well as which of these relational factors are most important in defining this latent factor. Statistical approaches based on latent variables models within the structural equation modeling framework (Kline, 2023), or the Schmid-Leiman solution (Schmid & Leiman, 1957; Wolff & Preising, 2005), could be particularly useful to this aim. This approach would allow tests of whether this latent factor predicts positive relations with companion animals and human well-being.

Finally, given the particularly clear role uncovered for human-companion animal compatibility in the current study, understanding the specific compatibilities that are most important when predicting the well-being and prosocial inclination outcomes could be further investigated. In other words, what are the most important specific compatibility attributes? Also, the implications of this relational factor could be further tested when predicting additional and consequential outcomes, such as guardians’ motivation to engage in activities that involve their companion animal (training, sports), and companion animal relinquishment. Future work could also be conducted to test if this relational factor also predicts actual care behavior (e.g., taking one’s companion animal to the vet when sick, regular feeding, making sure one’s companion animal has appropriate levels of socialisation and activities). Such findings would add more evidence for the need to devise empirically-based methods to promote human-companion animal compatibility so as to maximise successful companion animal adoption and their integration in human lives.

While a main focus in research investigating human-companion animal relations has been on the health and well-being experienced by humans within this relationship (e.g., Amiot & Bastian, 2015; Herzog, 2011), commentators are recognizing the importance of accounting for animals’ own needs and welfare within these relationships (e.g., Amiot & Santerre-Bélec, 2022). Doing so has the potential to improve the care given to animals in our societies, and, more broadly, to raise the social status of non-human animals. Such changes could have direct benefits to the health and welfare of animals. The current findings contribute to this discussion, by first identifying which relational factors predict increased prosociality toward companion animals (i.e., greater affection and more caring behaviors for the companion animal). Our findings identified three such relational factors, namely: the quality of the human-companion animal relation, unconditional acceptance of one’s companion animal, and positive contact with one’s animal. Promoting these factors socially and individually and developing strategies to nurture these factors within the human-companion animal relationship could have direct benefits for companion animals’ welfare. Second, the relational factor pertaining to human-companion animal compatibility was found to accentuate the positive association between caring for one’s animal and reporting higher well-being among guardians. This finding suggests that prospective guardians should mindfully select a companion animal with whom they are likely to be compatible; doing so may allow them to derive more benefits from their companion animal’s presence and to potentially develop a more mutualistic relation with their animal.

The role of companion animals in promoting human health and well-being is contested in the empirical literature. The current study contributed to this debate by focusing on which specific relational factors, experienced within the human-companion animal relationship, contribute to the guardians’ well-being as well as to their prosocial inclinations toward their companion animals. By doing so, the current findings allowed to identify the relational factors that promote greater mutuality within human-companion animal relations. While companion animals are ubiquitous and normative in many human societies, uncovering which specific aspects within the human-companion animal relationship could be individually and socially promoted appears useful for building more sustainable human-companion animal relations.