While the focus of bird-feeding studies has been either on the effects that bird-feeding has on avian or non-avian populations and communities9,28,29,30 or the reasons why humans want to feed birds2,6,31,32,33, our approach reveals a more complex diversity of ways in how birds and other species attracted to bird-feeding sites can, in turn, affect how people provide food for birds. The species present or absent at the bird feeder can both affect the bird-feeding practices, but, interestingly, most of the species can, depending on the context, lead to both increases or decreases in feeding birds. Context is highly important for many taxa: for example, cats are often seen as unwanted when a neighbor’s cat is predating birds at a feeding site, but when the owner’s cats are inside and watching birds through the window, they can provide a reason to feed birds. Sometimes, making decisions based on multispecies relations may lead to potential trade-offs, such as a respondent who said that there is no problem with the seeds that fall to the ground and attract rodents as free-roaming cats kill those rodents. The wild predators were less common than cats, but they predominantly led to a reduction in feeding birds. Interestingly, this provides potential for conflicting responses as the small birds on bird feeders attract predators34, though some respondents also felt happy about seeing birds of prey at their feeders. Indeed, our approach of asking about the changes in the feeding practices revealed a plethora of primary reasons that interact with species when people are making decisions on whether to put more or less effort into bird-feeding. Thus, clearly, bird-feeding is not a one-way interaction, but rather a series of reciprocal relations where humans and nonhumans react to each other.
Our results show that the most common species that leads to a decrease in bird-feeding in Finland is the brown rat. The most important reason to increase feeding was the occurrence of wanted species in the bird feeders. These are not surprising results as they correspond well to previous findings1,12,19,35,36. Interestingly, the most often mentioned taxa and most often mentioned reasons were both related to the decrease in bird-feeding rather than those that people were aiming to attract. This significance of unwanted species is also manifested in the apparatus used to provide food as avoiding feeding unwanted species drives the evolution of the bird feeder models used, more than improvements for feeding the species people prefer20.
Closer inspection also suggests that there are more complex phenomena taking place across bird-feeding sites: the same taxa can lead to an increase or decrease in feeding, but the same reason in different contexts can lead to either an increase or a decrease in feeding. The red squirrel (Sciurus vulgaris) is an interesting example, as it can be both a wanted or unwanted species at the bird feeder. Regardless of its status, squirrel appearance could lead to an increase in provisioning to make sure that birds also get food, or it could lead to feeding birds being stopped altogether due to concern that the squirrel may eat too more food than the respondent was prepared to invest in bird-feeding. Indeed, our results corroborate and expand on Dayer et al.’s12 findings that there can be both positive and negative feedback loops between observations of species at the bird feeders, human attitudes, and actions, and subsequently changes in the species community at feeding sites. Similarly, the disappearance of a wanted species from the feeding site can lead to both an increase and decrease in feeding birds. The respondents described ending the feeding completely, increasing the amount of feed or changing feed qualitatively in response to the situation. While some reasons can be expected to lead to a certain direction of change (for example, pathogens always lead to decreasing feeding), there is still a variation in whether the change is cessation, reduction, moving the feeding site or changing the food provided.
Why might people vary so much in how they relate to nonhuman species and whether their perceptions lead to increases or decreases in feeding birds? One possible reason is that our survey data was not obtained from a geographically large area across urban and rural areas. Indeed, in previous work, we found that bird-feeding occurs more often in urban areas, but is declining, while more food is provided in rural areas where it is increasing19. In other countries such as Poland, there can be large differences in the number and type of feeders provided, as well as the species that are attracted to them37. This could, therefore, influence the feedback loops that we have identified here between the species present at feeders and people’s relations to them. Indeed, the values placed on wildlife, as well as feeding birds, can vary widely depending on urbanity, socioeconomic status, household size, and age38, so an important next step will be to tease apart whether the feedback loops we describe here might also vary and potentially contribute to emerging differences in, e.g., urban-rural bird communities29.
A second reason is that people vary in how they appreciate or rate the “attractiveness” of species, and this can sometimes be acknowledged directly or indirectly. For example, the long-tailed tit (Aegithalos caudatus) was considered to be cute, and thus, they were fed. In contrast, some people explained that they attract birds and ensure their survival by reducing the pest insect pressure for the garden in the summer. These findings are not surprising as they resonate with previous research: for example, the long-tailed tit is esthetically pleasing to people39. On the other hand, some species, such as rats, are a typical example of both direct and indirect reasonings as they can be unwanted based on people just not liking them40,41, but also rationalized through reasons of zoonotic or infrastructure risks (even if the actual effect could be minimal or mitigated42,43). Indeed, these perceived risks can lead to regulations and prohibitions, which might lead to stopping feeding, even though the person feeding would have liked to continue (although our questionnaire exposed that regulations or social pressure are not always acted upon, see below). Many respondents were also worried about the ecological effects of feeding birds. If bird-feeding has been initially seen as targeted at improving the welfare of individual birds20, nowadays, respondents were clearly also thinking about the effects on specific, usually locally threatened species, and to biodiversity at large. This can be seen, for example, in the responses where observing tits at the feeding site could reduce feeding. There has been discussion about synanthropic tits (Eurasian blue tit, Cyaniste caeruleus; great tit, Parus major) benefitting at the expense of non-synanthropic tit species (willow tit, Poecile montanus; coal tit, Periparus ater): as their survival increases during the wintertime bird-feeding, they can dominate the community during the breeding season44,45. This knowledge has led to some respondents reporting being worried about willow tits as a reason to stop bird-feeding altogether, although recent results from Finland do not support this interaction among tit species46. Thus, different perceptions of the desirability of species and context could lead to different outcomes when it comes to feeding birds.
Third, our results highlight that interpersonal relations among people can also influence how they interact with nonhuman animals. For example, social conventions, relations and norms affected respondents’ decisions of whether to change how they feed birds. These norms could be explicit, either as written rules or neighbors shouting over the fence, or implicit, such as perceived acceptability in the neighborhood31. In Finland, urban apartment buildings are commonly administered as housing associations, where every apartment owner has a vote, and these associations commonly prohibit bird-feeding in the common yard due to, e.g., concerns over the presence of rats47. Importantly, however, we found that these norms did not always lead to consistent change, with some respondents even finding social pressures to be an inspiration to increase or even start to feed birds. Nevertheless, the most common relation that affected bird-feeding among our respondents was human-to-other species relation, and this was expected, as bird-feeding is generally expected to function on humans’ terms and conditions. For example, magpie can be seen as “stealing” food from the bird feeder: this might be due to people not wanting to feed magpies as they want to feed rather small birds or also because magpies, as large-sized birds, eat much more and while people might not have anything against magpies as such, they cannot afford feeding magpies. Some similar reasoning could be also seen through how people care: many people care for nonhuman animal-to-animal relations, such as evidenced by reasons for feeding birds being to provide entertainment for one’s pets. It seems that cultural norms, such as small birds being in more apparent need of care than corvids, are in play here.
Our findings should be understood within the Finnish context. For example, in Finland bird-feeding has long been more common during the winter, in particular when there is snow on the ground. Feeding birds when there is no snow on the ground is much less common and is more often directed to waterfowl or to birds in urban parks19,48. The total number of respondents to our questionnaire (i.e., more than 9000) was close to 0.5 percent of all households in Finland (i.e., 2,793,636 in 2021)49. About half of the households in the UK feed birds1, and although this data is not available for Finland, it seems that feeding birds is likely to be as popular here. For example, the amount of food (per household) imported to both countries intended for feeding wild birds is similar (UK: 28.4 million households purchase 150 million kg50; Finland: 2.7 million households with 15.8 million kg imported51). As the answers to our open questions were informative, even if sometimes short, such as “My heart was so broken when the magpie found the suet balls,” we would expect that our qualitative data should contain any important phenomena that would occur. We suggest, therefore, that the underlying phenomena we have found in this study might be quite universal, such as is shown by the global surge in bird-feeding during the pandemic52. On the nonhuman animal side, even if the particular species are different in different locations, their ecological niches vary similarly to Finland and are thus likely to present similar observations and interactions in bird-feeding sites. Thus, we suggest that while the specific species and proportions of different reasons vary from country to country, these relations are likely to be present in some form across the globe. More broadly, entering into these relations through bird-feeding creates a cycle of action and reaction: for example, feeding birds lead to behavioral changes in birds that not only modulates the pleasure that humans are able to derive from observing birds, but also relate to interactions among the bird community22,23.
Asking about the changes in bird-feeding practices and their underlying reasons directly links human actions and the relevant observations that humans make in their bird feeders. This approach does depend on the human perception about the change in the feeding. We do not know about the changes that are perceived as neutral, nor can we be sure whether the humans’ perception of the direction of the change is the real effect on birds. For example, a study showed that while using a squirrel guard increased the amount of feed going to birds, it reduced bird visits to the bird feeder and had differential effects on different bird species32. Similarly, we are not able to assess how relevant different perceptions of diseases are and whether they correlate with actual risk to humans or birds: while they are a highly common reason to reduce feeding, there are already a number of ways of taking these into account, such as bird-feeder design20.
In conclusion, we suggest that the relationship between bird-feeding and humans’ animal attitudes is a highly dynamic process and also context-dependent: the same species can in different contexts lead to either increase or decrease of provisional feeding. Similarly, the same reasons can be used to justify increases and decreases in feeding. Thus, the avian and non-avian species that are directly or indirectly affected by bird-feeding also affect humans and become actionable participants in the process. This phenomenon has both ecological and social dimensions, which are tightly interlinked. While bird-feeding influences species communities, it is also important to acknowledge and understand how other species influence feeding practices and behaviors and are active participants in bird-feeding practices.
