We experimentally demonstrated that great tits were able to remember and project the apparent nest-site feature preferences of conspecifics and other guild members in the previous breeding season onto nest-site choice one year after. Great tits that settled early or late strongly avoided the nest-site features associated with both tits and flycatchers and preferred the feature associated with unoccupied nest boxes. Such a shift towards apparently unused niche space was strongest among philopatric individuals, i.e. those which had bred in the same study plot in the previous year. The preference for the symbol associated with unoccupied nest boxes in immigrant females, which were not expected to respond to treatments, probably reflects male’s impact on nest-site selection decision  if they were mated with a philopatric male. These results emphasize that individuals can shift their resource use depending on the observed resource use of con- and heterospecifics. Importantly, this occurred without strong resource limitation (because there were plenty of empty nest-boxes available), indicating that social information use can have independent impact on resource use and niche dynamics within communities. In addition, most (c. 56%) great tits made their nest box choice before the first collared flycatchers returned from migration and initiated nest building, which demonstrates that they can actively avoid the features apparently preferred by flycatchers in the previous breeding season without flycatchers being present when the choice is made. Such information use, where decisions can be influenced without immediate association, may lead to more lasting and widely spreading shifts in resource use compared to a situation when imminent signal is needed to produce a response. Because our experimental design of using cross-controlled abstract geometric symbols as a substitute of nest-site characteristics effectively controls for innate or learned preferences, our results provide strong inference for the effects of social information use on resource use in animals.
Our experimental design cannot unequivocally distinguish whether great tits preferred for the unused niche space or avoided the preference of conspecifics and flycatchers. The avoidance of features preferred by conspecifics may reflect the costs of intraspecific competition and/or the avoidance of ectoparasites living in nest material. Classical niche theory predicts that niche width is a result of the expanding and reducing forces linked to intra- and interspecific competition, respectively . Intraspecific competition can indeed expand the resource use of individuals in a population , and changes can take place rapidly through behavioral plasticity as a response to resource availability . Our results highlight that long-term evolutionary processes are not necessarily needed for niche shifts to occur (cf. ): the perceived resource use of con- and heterospecifics, even without strong resource limitation, can also trigger niche shifts within individuals’ lifetime. However, if the strategy of preferring previously unused resources is driven by reducing intra- and interspecific competition, its prevalence and benefits may depend on the population density. High population density, and in particular high number of philopatric individuals, may increase competition for previously unused resources. At low or intermediate densities, or if population includes a low proportion of philopatric individuals, a strategy of preferring previously unused resources may result in reduced competition over the resource. Great tits may also have avoided nest features associated to tits because potential heterospecific competitors, the blue tits , exhibited the same apparent preference for a given symbol as great tits. However, the effect of blue tits is plausibly minimal because great tits dominate blue tits in the selection of nest-sites . Another plausible force driving niche shift in nest site selection could be the presence of nest parasites. Shift towards unused nesting resource may be reinforced by ectoparasite loads frequently present in old tit nests, which can decrease nesting success . In our study, nest boxes were cleaned after each breeding period, so visual signals of the presence of old nests or ectoparasites could not be utilized directly by the individuals – instead, avoidance mechanism was indirect, via responding to a nest-site feature associated to con- or heterospecifics. Parasitism could be a main force selecting for the use of social information to avoid settling in a potentially previously occupied site.
The quadratic effect of day on great tit nest site feature choices remains unknown, but plausible explanation is the varying intensity of competition over high-quality nest sites over settlement period. During the peak of the settlement period, from late April to early May, time constraints and competition for the best nest sites  and mates are at their highest, which is probably strengthened by the appearance, and fast accumulation, of flycatchers. Conceivably, decisions are then likely to be based on more immediate factors such as the occupancy status and owners of neighbouring boxes. At the end of the breeding period, competition decreases again and may allow individuals to either use social information gathered in the previous or current year. The choices of philopatric and immigrant birds were distributed rather evenly over the season suggesting that it cannot explain the observed pattern.
Great tits may also have avoided symbols that were apparently preferred by collared flycatchers in the preceding breeding season. Avoiding the reciprocal negative effects of direct interspecific competition between our two study species may explain this result ,,, but see . Apparent competition , driven by shared nest predators, has also been shown to be a strong selective force causing divergence in nest-site use in birds ,. Additionally, "information parasitism" of tits by flycatchers could also explain the great tit response. In the pied flycatcher (Ficedula hypoleuca), a closely related species to the collared flycatcher, individuals have been shown to prefer to breed in the vicinity of tit nests and thereby gain fitness benefits , while great tits suffer from the proximity of pied flycatchers in terms of reduced nesting success . In addition, flycatchers can copy apparent novel nest-site feature preferences of tits ,, implying that they may actively penetrate into the niche space of tits. Conceivably, counter-behaviors may have been selected in great tits to escape the negative effects of flycatchers by avoiding the nest site features that they apparently prefer. In line with this hypothesis, Loukola et al.  demonstrated that one function of the egg-covering behaviour with hair in Paridae during egg-laying can be preventing flycatchers to obtain the important clutch size information ,.
Irrespective of the ultimate mechanism(s) leading to avoidance of apparent con- and heterospecific resource use, our results add support for earlier findings that, in addition to the usually considered “copying rules” , learning strategies based on actively avoiding others’ choices also exist ,,. Avoiding using the same resources than others may be adaptive if it reduces the costs of overlapping resource use, such as competition and indirect negative effects caused by shared predators and parasites (apparent competition), or if it allows selecting higher quality resources depending on the quality of the demonstrators .
Collared flycatchers and blue tits showed no response to the artificial nest site feature preference of con- and heterospecifics in this study, yet, these species are just as likely as great tits to be able to respond to the association between symbols and the occupancy status of the box, and more generally to use intra- or interspecific information ,,,,-. The sample size for blue tits was perhaps too small to detect an effect. Flycatchers probably rely more on the up-to-date information provided by tits that already have initiated breeding activities by the time flycatchers arrive in their nest-site feature selection decisions cf. , than one year old cues. A recent study also showed that collared flycatchers do use information from the previous breeding season in their small-scale nest-site selection . Philopatric individuals preferred to breed at a site where conspecifics were breeding in the previous breeding season. Also a high breeding success of conspecifics increased the likelihood of settlement close to such a site . Thus, in across-year information use, it seems that flycatchers prefer to use small-scale intraspecific spatial location information about nest locations in their nest box selection rather than relying on more large-scale information about nest-site niche preferences, which our symbols were reflecting.
To conclude, our experiment showed that the perceived resource use of con- and heterospecifics can influence the resource use of animals later on, even without any apparent resource limitation and in the absence of heterospecifics. This result complements our understanding of the division of resources among coexisting species by showing that social information use in the form of avoiding the resource use of others, both within and between species, may affect the realized niches of coexisting species. This may have implications for the rate of phenotypic change of coexisting species because the observed effect on resource use was parallel with the theoretical predictions of intra- and interspecific competition. Hence, information use in interspecific context may complement the evolutionary effects of competition, and enhance the speed of the niche divergence among species using overlapping resources.