Hutchinson’s ecological niche for individuals

We here develop a concept of an individualized niche in analogy to Hutchison’s population-level concept of the ecological niche. We consider the individualized (ecological) niche as the range of environmental conditions under which a particular individual has an expected lifetime reproductive success of ≥ 1. Our concept has primarily an ecological function, as it refers to the match of an individual phenotype to its contemporary environment (niche fit) while we discuss evolutionary fitness as an evaluative parameter of this fit. We address four specific challenges that occur when scaling the niche down from populations to individuals. In particular, we discuss (1) the consequences of uniqueness of individuals in a population and the corresponding lack of statistical replication, (2) the dynamic nature of individualized niches and how they can be studied either as time-slice niches, as prospective niches or as trajectory-based niches, (3) the dimensionality of the individualized niche, that is greater than the population niche due to the additional dimensions of intra-specific niche space, (4) how the boundaries of individualized niche space can be defined by expected lifetime reproductive success and how expected reproductive success can be inferred by marginalizing fitness functions across phenotypes or environments. We frame our discussion in the context of recent interest in the causes and consequences of individual differences in animal behavior.


Introduction
Individuals differ consistently in their behavior and their relations to the environment. We here aim to explore how individual differences can be integrated into the ecological niche concept in order to yield a utile framework of an individualized niche. Many of the individualized niche features, that we discuss here, have metaphorical value that may help in structuring research (or in modelling studies), but some aspects can also be quantified empirically in natural systems. We first discuss recent progress in the study of consistent individual differences in animal behavior. We next briefly review ecological niche concepts and their different definitions. We then discuss the application of the Hutchinsonian ecological niche concept at the level of individuals. Our arguments are based on the idea that since individuals differ phenotypically, they often also differentiate their positions in the environment, eventually generating individualized niches. We structure our discussion of the individualized niche along four key questions: How can we deal with the fact that individuals are not statistically replicated? How can we incorporate time in the study of individualized niches? Which dimensions constitute individualized niches? Where are the boundaries of individualized niches? These four questions, we think, reflect important considerations, when implementing the concept of the niche at the level of individuals.
We write this essay from the perspective of empirically working behavioral ecologists. We therefore envision populations of individually distinct animals such as vertebrates or arthropods. While we are interested in the causes and consequences of individual differences (including, but not limited to, animal behavior), we do not see a particular individual as the object of study. Instead, we strive to understand how individual differences contribute to population-level processes. It is therefore the state and dynamic of population composition that interest us. We, like many other researchers in the field, use statistical summaries at the level of populations to study individual differences. This perspective relies on the law of large numbers and aims to understand general patterns and processes rather than individual life histories.
Being interested in the consequences of individual niche specialization does not mean that we include long-term or evolutionary consequences in the individualized niche definition that we develop here. Whether a particular phenotype will spread in a population depends on how phenotypic variation is inherited and how particular phenotypes perform in comparison to other phenotypes in the population. We see both aspects, inheritance and relative performance, as very important topics, but not directly relevant to the definition of the individualized niche as such. We think of the individualized niche as the current performance of a particular phenotype in the momentary environment. The concept is thus mainly an ecological and functional concept.

Consistent individual differences
Consistent individual differences have been in the spotlight of behavioral ecology for the last two decades (Sih et al. 2004;Réale et al. 2007). There are now hundreds of studies that report on individual differences in behavior across a large array of species, including vertebrates and invertebrates (Bell et al. 2009). A particular interest has been on behavioral traits that represent general reactions towards the environment, especially when these traits are temporally consistent and correlated across contexts (Kaiser and Müller 2021;Dochtermann and Dingemanse 2013). Individually consistent, context-general traits are often called animal personality traits, temperament traits, coping styles or behavioral syndromes (Kaiser and Müller 2021;Réale et al. 2007). A common research framework, relating to consistent individual differences, is the pace-of-life syndrome, which encompasses behavioral, physiological and lifehistory components. The pace-of-life syndrome has been linked to personality and survival (Ricklefs and Wikelski 2002;Réale et al. 2010), while it has laid the foundations for the study of implications of behavioral variation (Wolf and Weissing 2012).
From an evolutionary perspective, all individual differences that are heritable can evolve by natural selection. Indeed, individual differences in behavior often have a significant heritable basis (Stirling et al. 2002). It has been shown that animal personality differences can be systematically selected for, thus maintaining inter-individual variation in behavior (Dochtermann et al. 2015;Wolf et al. 2007). Furthermore, intraspecific variation affects interspecific interactions and ultimately species' coevolution (Moran et al. 2021). Consequently, the position of individuals in the environment can both be the cause and the consequence of behavioral differences, owing to the individual x environment interaction being bidirectional (Dingemanse and Wolf 2013).
Individual differences in behavior have ecological consequences, because they have an impact on the way in which individuals interact with their environment. For example, phenotypic variation can affect population dynamics through polymorphism in resource use (Dall et al. 2012). Indeed, intraspecific competition might as well be a fundamental cause of individual differences in behavior (Bergmüller and Taborsky 2010). Reduced competition over resources -as a result of specialization at the level of individuals -can increase the carrying capacity of a habitat and promote resilience of populations (Wolf and Weissing 2012). Thus, in analogy to community dynamics, resource partitioning among phenotypes can reduce intraspecific competition and facilitate population growth and persistence Layman et al. 2015). Indeed, an extensive review of empirical evidence on the consequences of intraspecific variation showed that inter-individual diversity increases establishment success, range size, population stability and resilience, while it decreases extinction risk and vulnerability to climate change Forsman and Wennersten 2016). Intraspecific variation in population-related traits can thus alter population and community dynamics Araújo et al. 2011). We think that an individualized ecological niche concept can provide a fruitful perspective on individual differences.

The ecological niche
The concept of the ecological niche is fundamental in Ecology. The term was initially vaguely defined and used to describe the ecological position, habitat and requirements of species Packard 1894;Grinnell 1917;Allen 1882, see Gibson-Reinemer 2015. First composed definitions of the ecological niche were presented by Elton (1927) and Grinnell (1928). Elton (1927) defined the ecological niche in terms of the species' function within a community and its relations to other species. This view is focused on the functional role of species and is mostly used in community and functional ecology. Grinnell (1928) proposed the ecological niche as the physical place that species are adapted to. According to this definition, niche is a synonym of habitat or position of species in the environment.
The first reviews on the ecological niche emerged quite early (Hutchinson 1978). Since the introduction of the term, the definition of the ecological niche was a topic for debate. Hurlbert (1981) published a collection of more than 20 quotes defining the ecological niche. We expanded this collection of definition quotes to present, resulting in 36 definitions (Table S1). Some of these definitions are only subtly different. However, ecological niche concepts can be broadly categorized into environmentbased concepts and function-based concepts. Environment-based concepts include the ecological niche as the habitat/environment (Grinnell 1917(Grinnell , 1928Gause 1934;Dice 1952;Odum 1959) or as abstract environmental space (Hutchinson 1957;Root 1967;Macfadyen 1957). Function-based concepts include the functional role of a species (Elton 1927;Clarke 1954) or its trophic position (Elton 1950;Weatherley 1963). Some definitions include a combination of environmental requirements and effects on resource availability (Chase and Leibold 2003).
The most popular and widely cited definition of ecological niche was proposed by Hutchinson (1957Hutchinson ( , 1978. Hutchinson defined the ecological niche as a hypervolume in an n-dimensional (abstract) environmental space that allows a population to persist indefinitely. He distinguished between fundamental and realized niches, which correspond to an ecological niche before and after accounting for interspecific competition, respectively. While the Hutchinsonian niche concept is primarily defined in terms of place in the environment, it does include some functional aspects, in particular owing to the distinction between the fundamental and the realized niche. Rosado et al. (2016) claim that Hutchinson built on Grinnell's idea, while others (Colwell and Rangel 2009;Swanson et al. 2015) argued that the concept of the hypervolume was introduced by Gause (1934). Independently of Hutchinson's source of inspiration, the n-dimensional hyperspace is until today a fundamental concept in ecology and evolution. We therefore explore how this concept can be usefully applied at the level of individuals.

The ecological niche of individuals
The recent interest in the study of individual differences highlights current focus on ecological differences between individuals within populations. Here we address the applicability of Hutchinson's niche concept at the level of individuals. Some early work on ecological niches already included discussions on the importance of individual differences within a population. vanValen (1965), for example, pointed out that individuals differ on how they use available resources and that population niche width is driven by the variation between individuals (Niche Variation Hypothesis). Roughgarden (1972) pioneered the idea to use individual differences in trait expres-sion as proxies for resource use. Traits of individuals are here used as substitutes for the environmental dimensions, which are more difficult to measure. Roughgarden's ideas gave rise to a vibrant field of functional trait analyses (Violle et al. 2007). In the meantime, large databases of functional traits have been compiled, in particular for plants (Fraser 2020;Kattge et al. 2020), albeit only part of these data focus on individual differences.
Although the study of intraspecific variation has been neglected for some decades, it revived around the turn of the last century (Bolnick et al. 2003). Individual niche specialization has been studied empirically mostly with a focus on diet, while studies focusing on habitat selection, behavior, or labor division are less numerous (Ingram et al. 2018;Dall et al. 2012;Bolnick et al. 2003). Notably, individualized niches have been even more vaguely defined than concepts of the ecological niche as such Bergmüller and Taborsky 2010;Müller et al. 2020; but see Trappes et al. 2021). This is partly because the concept is broad and encompasses aspects that can better be kept apart. The aim of our essay is to bring structure into the individualized niche concept and provide definitions not only for the individualized niche in the broad sense, but also for facets that are best treated under different (sub-) labels. In our view, there are four main challenges when applying the concept of the ecological niche to individuals: (1) the question of uniqueness, (2) the questions of time, (3) the question of dimensions and (4) the question of boundaries. We first start with a working definition of the ecological niche of individuals before addressing these specific challenges. Hutchinson (1957) defined the (fundamental) ecological niche of a population as the range of environmental conditions in which a population can persist indefinitely. Indefinite persistence implies non-negative population growth rate in the long term. Scaling down to individuals, we propose a working definition of the individualized (ecological) niche as the range of environmental conditions that provides an expected lifetime reproductive success of ≥ 1 surviving offspring to particular individuals. In outcrossing organisms each offspring has two parents and should therefore be counted only as 0.5 for each parent.

Working definition of the individualized niche
Before going into more detailed aspects of our individualized niche concept, we want to highlight two important points: First, lifetime reproductive success (commonly used as a measure of absolute fitness) serves as the currency of the phenotype-environment match in our concept and not as the determinant of contemporary selection. This aligns with the Hutchinsonian niche being an ecological, rather than evolutionary, concept. Research on how the individualized niches evolve might have to consider the comparative performance (relative fitness) of alternative phenotypes, including an adjustment for the mode of reproduction. Second, we highlight that the individualized niche, as used in this manuscript, is defined by the environment that an individual lives in, not by its phenotype. The phenotype can act as a mediator that affects fit to the environment (Trappes et al. 2021), but does not represent a part of the niche itself.

The question of individual uniqueness
One issue when defining niches at the level of individuals is that individuals are (by definition) not identical, impeding statistical replication. The ecological (Hutchinsonian) niche of a population can be estimated by quantifying the (hypothetical) areas where different members of a population can be found in the environmental space. Here, individuals serve as replicates at the level of the population and can thus occupy the same niche. However, individuals themselves can only be found at a particular point of environmental space. (We leave the discussion of integration over time for the following section.) Hypervolumes at the level of populations become points in environmental space at the level of individuals. In loose analogy to Hutchinson's realized niche, we call each of these points the realized individualized niche. However, the point where an individual happens to live almost certainly does not cover the range of environmental conditions under which it could have occurred. The potential individualized niche thus includes all environments where a particular individual would (or could) have an expected lifetime reproductive success of ≥ 1 (Fig. 1). This means that the potential individualized niche is defined by a space of unobservable outcomes. How can we deal with the problem that realized individual- There are at least partial solutions to both issues. A common approach in the study of realized individualized niches is to address the question on the level of populations and to integrate over time. If we collect repeated observations per individual over short but meaningful time intervals, we can use variance decomposition approaches to quantify population-level variability in realized niches. One approach is the estimation of individual-level repeatabilities that quantify the proportion of variation that is explained by individual differences (Nakagawa and Schielzeth 2010;Bell et al. 2009). The idea here is to treat individuals as ephemeral instantiations, but to view the population-level individual variation as a stable population-level feature of the magnitude of individual differences in realized individualized niches.
Even with replicates over short meaningful time intervals, it is de facto impossible to cover the full potential niche of an individual. At least in observational studies under natural conditions, environmental covariation in space and time will prevent individuals to be observed across the full range of potential environments in which it could have an expected lifetime reproductive success of ≥ 1. Experimental approaches offer a partial solution if individuals can be translocated to a range of different environments (Wilson et al. 2019). Some measure of current performance can then be used as a proxy of reproductive success across a range of environments (sacrificing the value of a fixed boundary for defining the niches, see discussion below). However, experimental approaches are necessarily limited to a few dimensions of environmental space. An ultimate limit to experimental exploration of the potential individualized niche is also set by the lifespan of an individual, since potential individualized niches are almost certainly substantially larger than realized niches.
An alternative approach is to marginalize across phenotypes (or genotypes) when mapping individualized niches (Fig. 2). This is rooted in Roughgarden's (1972) idea to use traits of individuals as proxies for resource use. Individuals are here used as replicates to establish a distribution of phenotype-specific environments. In principle, Fig. 2 Schematic view of the idea of using population-level patterns to predict individualized niches and fitness consequences. Colors show different types of individuals (e.g. females and males). The left plot shows two trait dimensions of which one is informative for occupancy of specific environments. The right plot shows a multivariate fitness distribution that depends on phenotype (here shown by different colors on the abscissa) and environments. Fitness arises from the combination of phenotypes and environments. Darker colors show higher fitness expectations this can be done across many different traits. While individuals are used as tokens of types in particular phenotypic dimensions, individuals are typically unique in their trait combinations. In principle, it would be possible to predict an individual's niche from its unique combination of traits. Such predictions are also possible for nonlinear relationships, provided that the form of the mapping function is known. A limit is set only if interactions between traits are strong and poorly replicated in a population. In such cases, trait combinations in some individuals might be so unique that prediction becomes impossible, a limit that is shared with phenotypic novelties.
Both realized and potential niches might be of interest to ecologists. In some cases, the environmental space that is occupied by an individual might be incidental. In other cases, however, features of an organism might influence the realized niche space that can be occupied. Many insect species, for example, show developmentally plastic wing length polymorphisms (Harrison 1980;Zera and Denno 1997). Wing length affects dispersal abilities and thus the range of environments an individual can reach. Short-winged individuals might, in principle, be able to survive and reproduce in very diverse environments (thus they might have a wide potential niche), but in reality, they are limited to the realized niche at their local patch. The developmental pathway to develop long-winged, dispersive phenotypes might not affect the potential niche as defined above, but might result in a much wider array of realized individual niches. Sampling of environments is only possible for an individual with sufficient mobility.
We may distinguish a third form of the individualized niche, the fundamental individualized niche. The difference to the potential individualized niche is very subtle and probably not too relevant in practical applications, so the two might often be used interchangeably (see Trappes et al. 2021). Huntchinson's fundamental ecological niche is the environmental space that is occupied by a population in the absence of specific environmental factors (competitors, predictors, dispersal barriers). This is appropriate for populations, because if a species is absent from a potentially suitable habitat, it is so for a reason. Individuals, however, exist only as a single copy and can be absent from many suitable environments, not for specific, but for arbitrary or random reasons (e.g. being born in a specific place). The term fundamental individualized niche might thus be used when there is an absence of particular external (usually intraspecific or interspecific) factors, while the term potential niche does imply coincidental absence from some environments -simply because individuals cannot be at multiple places at a time. The reference space of the potential individualized niche is usually the realized niche of the population, while the reference space for the fundamental environmental niche are all possible environments. The distinction is specific to the individualized niche, since replication is less of an issue for the niche of the population.
Definition A: The realized individualized niche is the place in environmental space in which a particular individual is found and has an expected lifetime reproductive success of ≥ 1. The realized individualized niche can be quantified empirically.

Definition B:
The potential individualized niche is the volume in environmental space in which a particular individual could be found with an expected lifetime repro-ductive success of ≥ 1. The potential individualized niche cannot directly be quantified, but significant parts of the niche space can usually be statistically inferred.

The question of time
We have alluded to the integration across intervals of time above. This raises the more general question about whether the individualized niche refers to slices of time or to entire lifespans. The ecological niche of a population is focused on entire lifespans. The ecological niche of a forest-dwelling frog, for example, includes a network of forests and ponds, since adults require shelter in woodlands while in its juvenile stage, as a tadpole, the frog requires ponds for survival and growth. Population persistence can only be achieved if both habitats are available. One might argue that the equivalent is also true for individuals: that the individualized niche is a lifetime niche. However, there are arguments why this simple application of lifetime niches misses important intricacies of the individualized niche.
Throughout an individual's life, developmental decisions influence the niche space later in life (West-Eberhard 2003). The development of long wings in grasshoppers, for example, is triggered by increased population density (Poniatowski and Fartmann 2009). All (or at least most) individuals seem to have the potential to develop the long-wing phenotype under high population density, but remain short-winged under low population density. Potential niches of long-and short-winged individuals are therefore no different at birth, since all (or at least most) individuals have the potential to develop into either phenotype. It is a specific time during development when niches of short-and long-winged phenotypes split. Another example is given by match-based phenotypic adjustments. Some species of grasshoppers, for example, are able to change their body coloration during development (Rowell 1972;Dearn 1990). Since body color affects background-dependent crypsis, individuals of different color morphs have different individualized niches in the sense of environmental conditions under which they can survive and reproduce. At birth, these individuals have the same potential for alternative body colors, therefore they have the same potential individualized niches. However, after phenotypic adjustment, their niches become different. A focus on lifetime niches misses the importance of such critical developmental decisions.
We therefore think that the individualized niche (whether realized or potential) is most fruitfully viewed from two perspectives. A time-slice perspective looks for individual niches within certain life stages or other relevant periods of time (such as different seasons). The study of such time-slice individualized niches (Fig. 3) allows insights into individual differences in the use of niche space and short-term phenotypic adjustments. A now-and-in-the-future perspective looks at individual niches with a focus on sensitive phases or developmental switch-points and their lifelong consequences (Sachser et al. 2020). We define this ("now-and-in-the-future") perspective, the prospective individualized niche (Fig. 4), as the space of environments in which an individual can survive and reproduce given its current phenotype and its developmental opportunities. The prospective individualized niche is the timestructured space of potential niches.
The prospective individualized niche does not give a lifetime perspective except for the special case of a zygote. Potential individualized niches are affected by previous development (and by accidents). Certain areas of environmental space might not be available if irreversible developmental plasticity in early life-stages prevents an individual from developing a matching phenotype (Nyman et al. 2018). Development has manifest consequences for the individualized niche. The potential niche from a prospective perspective therefore changes as individuals age. In fact, with the possible exception of accidents, it always shrinks, as potentials are widely available at early stages and can only be reduced by individual decisions during development. The potential time-slice individualized niche, in contrast, might vary across lifetime and might shrink or expand as an individual keeps adjusting its phenotype.
Accidents and 'bad luck' represent a special case that should be taken into consideration. Purely coincidental events that might affect any individual with equal probability shall not be considered as affecting the expectations of lifetime reproductive success. However, not all risks are equally distributed across environments. If individualized niches are unequally risky, then (some) accidents are in fact nonrandom and genuinely affect fitness expectations. Some individuals may select risky environments with high variance in reproductive success while others select safer environments (Moran et al. 2021). All individuals may have the same probability of being killed by a storm, while choosing to nest in areas with high predator density (or not) affects the reproductive success non-randomly.
However, there is room for a lifelong perspective. We think it is usually meaningless to reconstruct realized individual niches post-mortem for its own sake, since in biology we are rarely interested in unique individuals that represent an ephemeral phenomenon. Rather we aim to understand general patterns and mechanisms. A compilation of individual lifetime niche trajectories (with dynamic changes throughout life) can expose alternative developmental trajectories as bundles of alternative realized niches that change across age (Fig. 5). Such a trajectory-based lifetime perspective helps to answer the question how individualized niches arise during development. We therefore call specific life-history trajectories in environmental space the trajectory-based individualized niche.
Definition D: The prospective individualized niche is a volume in environmental space in which a particular individual has an expected lifetime reproductive success of ≥ 1 that includes the current and future potential niches. The prospective individualized niche provides a focus on particular developmental decisions, which affect future niche space and can be quantified empirically. Definition C: The time-slice individualized niche is the environmental space in which a particular individual occurs during a particular part of its development and has an expected lifetime reproductive success of ≥ 1. Aspects of the time-slice individualized niche can be quantified empirically, with repeated measurements.
Definition E: The trajectory-based individualized niche is a time-structured volume in environmental space that allows for expected lifetime reproductive success of ≥ 1 and that is different from alternative developmental trajectories. The trajectory-based individualized niche provides a focus on alternative developmental trajectories that affect potential niche space and can be quantified empirically. Hutchinson (1957) defined the ecological niche as an n-dimensional space of environmental dimensions: abiotic (scenopoetic) and biotic (bionomic) factors. Attributes of the focal species, such as specific phenotypes, are not dimensions of the environmental niche. Instead, traits are features that allow a species to occupy a specific environment, for example by providing the ability to exploit particular resource (and traits can be used as proxies for resource use, Roughgarden 1972). Hutchinson distinguished the fundamental niche, the space that can be occupied by a particular species in principle, from the realized niche, the space occupied by a particular population as a consequence of competition. Since the presence of the other species is just a particular dimension of environmental space, the main function of the realized vs. fundamental niche distinction is to highlight how a particular inter-species interaction can affect niche use (a clearly functional perspective). The realized niche is thus the niche of a species in n -1 environmental dimensions.

The question of dimensions
In analogy to Hutchinson's ecological niche, we define the individualized niche in terms of environmental dimensions, explicitly including all biotic and abiotic factors that are external to an individual. There is no need to restrict the factors to those that are causally relevant to an individual's reproductive success. Some environmental dimensions might have little influence on reproductive success, however this is an empirical finding and should not condition the use of particular environmental dimensions. It is sometimes argued that niche dimensions should be independent, i.e. orthogonal (Blonder et al. 2018). Often they will not be orthogonal and some subspaces will not be realized in any real physical location. It is thus impossible to infer whether some environmental combinations represent part of the niche of an individual (or population). However, it is most useful to define niche space by evidence for presence of an individual rather than lack of evidence for an absence. Combinations of environmental dimensions that are not realized in the real world should thus not be regarded as part of the ecological niche of individuals (or populations). While niche dimensions might not be orthogonal in the real world, it is fair to treat them as orthogonal in hypothetical environmental space.
When scaling down from populations to individuals, the intraspecific context becomes external to the individual. The presence or absence of conspecifics (includ-ing potential mates) or conspecifics with particular trait values become an explicit part of the individualized niche. The social context, for example, is part of the individualized niche, like the interspecific community context in the ecological niche of populations. The social conditions that allow an individual to realize a non-zero inclusive fitness are also known as the social niche Blonder et al. 2018;Saltz et al. 2016, see below for a discussion of setting the boundaries). The fact that the intraspecific (including social) context is part of individualized niche dimensions represents one of the most important differences to the population niche. The individualized niche, thus, consists of n + s dimensions, where n represents non-intraspecific dimensions, while s represents the dimensions of the intra-specific niche space (Fig. 6).
The intraspecific context is broader than the social settings. Population density and the frequency of other phenotypes of a species may impact the individualized niche even without social interactions (van Benthem and Wittmann 2020). For example, some prey species such as grasshoppers are color polymorphic (Rowell 1972) and some of their predators develop search images to specialize on the most frequent morph in a population (Bond 2007). The expected lifetime reproductive success of an individual with a particular body color may thus depend on the frequency of that color morph in a population -even if all other environmental variables are identical. Rareness of a particular phenotype can be an advantage even when the phenotype in itself coveys no general benefit (Violle et al. 2017). Such processes give rise to frequency-dependent selection, affecting the niche space of individuals, since some We suggest that the difference between the presence and absence of intraspecific niche dimensions represents a particularly interesting aspect of the individualized niche, especially since the social environment can have profound influences on later individual phenotypes (Jäger et al. 2019): How does the niche of an individual change in response to the state of the population as a whole (including density and frequency of other phenotypes)? This offers an interesting perspective on the concept of soft vs. hard selection in evolutionary biology (Wallace 1975;Bell et al. 2021). Hard selection refers to selection that is determined by the phenotype of the focal individual and its environment, while soft selection occurs when selection is density-and frequencydependent. Population density, phenotype frequencies and social interactions are thus important components of the individualized niche. Hutchinson (1957) defined the boundaries of a population's niche by indefinite population persistence and thus non-negative average growth rates in the long term. Population growth rates are determined by the ratio of births to deaths in a population. The equivalent quantities at the level of individuals are survival and reproduction and those can be used for determining the boundaries of individualized niches. However, there are three important considerations, a rather easy and two harder ones, when translating this to the level of individuals.

The question of boundaries
The easy complication is the question of whether niche boundaries are sharp borders or gradual zones of niche fit. In fact, this consideration applies to both individualized and population niches and can be solved by working with continuous values of population growth rates (in the case of populations) or lifetime reproductive success (in the case of individuals). This results in a nuanced view of core and marginal niche space. A minor complication is that population growth rates and individual lifetime reproductive success are often low under most suitable environmental conditions, especially when they are density-dependent, and the focal population is near its local carrying capacity (Engen and Saether 2017). This is less of a problem for the individualized niche if population density is considered as one of the niche dimensions. Nevertheless, even in case of the ecological niche of a population, population size (or population density) can be used to estimate the soft borders of niche boundaries.
The harder problem is which concept of individual lifetime reproductive success should be considered. It might be tempting to use realized lifetime reproductive success, quantified in terms of number of offspring produced. However, realized lifetime reproductive success has a large stochastic component and is often a poor indicator of a particular individual's niche fit. If we use the realized lifetime reproductive success (as e.g. Saltz et al. 2016 seem to do), then we do have a problem with individuals that have thrived throughout live, but have bad luck and do not reproduce by some coincidence (see above for a discussion of risk factors). They would be considered to be out of their niche, because their realized lifetime reproductive success (even inclusive realized fitness) is zero. We therefore define the boundaries of individual niche space in terms of expected lifetime reproductive success, which are functions of the phenotype-environment combination (Fig. 7). Expectations of reproductive success do not necessarily invoke propensities in the sense of stochastic dispositions, but are rather built on statistical summaries that follow the law of large numbers (Drouet and Merlin 2015). Individualized niches are thus identified by mapping lifetime reproductive success on phenotype-environment combinations in the form of multidimensional fitness functions. Since there are no replicates of an individual, there is no empirical solution, neither to decompose individual lifetime reproductive success into a stochastic and a deterministic component, nor to quantify individual lifetime reproductive success across different environments. Resorting on fitness components or fitness proxies might be a viable solution (Patrick and Weimerskirch 2014). However, with fitness components we have to abandon (or at least adjust) the absolute threshold of expected lifetime reproductive success of ≥ 1. Work with fitness components will thus discover mostly gradual (soft) rather than sharp boundaries and this could be done even with relative fitness. Alternatively, we can marginalize across phenotypes (or genotypes) and environments to estimate expected lifetime reproductive success in the form of fitness functions using different individuals as replicates (Fig. 2).
One might wonder whether the boundaries of the individualized niche are defined by a lifetime reproductive success of zero or one (Fig. 7). One problem with repro- Fig. 7 Multidimensional fitness function and boundaries for the individualized niche. The graded blue area shows the expected (absolute) lifetime reproductive success kernel. The solid blue line marks what we consider the boundary of the individualized niche at an expected isocline of 1. The dashed black line marks the absolute boundary of where expected fitness drops to zero ductive success expectations is that they might get infinitely small and it might be difficult to tell where they become zero. The condition of (simply) positive reproductive success expectations thus forms a theoretical boundary that is difficult to determine empirically. We argue that while individuals cannot persist indefinitely, they need to leave at least one offspring to perpetuate into future generations. A useful threshold for the boundary of the individualized niches is thus the (long-term) expectation to produce one descendant. We think that this makes a useful benchmark in a gradual view of the individualized niche.

Conclusions
We have started with a discussion of individual differences in behavior. We now want to come back to this and ask whether individualized niches are a mere rebranding of the study of individual differences. In brief, we think there are important differences. First, in our concept it is not the phenotype itself that represents the individualized niche, but the environment that an individual lives in. Not all individual differences in phenotype and behavior are thus relevant to the individualized niche (Trappes et al. 2021). The subset of individual differences, which mediates phenotype-environment matches (Edelaar and Bolnick 2019), is the most relevant to the individualized niche. While the literature on individual differences focuses mainly on survival and fitness consequences of individuals, the individualized niche focuses on the environment and, in particular, relates the phenotype-environment match to individual differences (in line with Roughgarden 1972). Furthermore, in order to estimate individualized niches, the full range of an individual's ecology and life history needs to be studied. This highlights the urge for studies, which incorporate lifetime-long observations of individuals.
We have introduced the field of animal personality and the ecological niche concept and have discussed how they can merged into the concept of an individualized niche. We provide a working definition of individualized niche that builds on Hutchinson's population-level ecological niche. However, there are important intricacies when developing an individualized niche concept. Particularly important are (i) the differentiation between realized and potential niches where the latter is defined by unobservable outcomes, (ii) the dynamic nature of individualized niches with a time-slice, a prospective and a trajectory-based perspective, (iii) the inclusion of intra-specific dimensions in the dimensionality of individualized niches and (iv) the need to define the boundaries of individualized niche space by expected lifetime reproductive success (not realized lifetime reproductive success). We hope that these considerations will help other scientists to further develop the concept of the individualized niche into a practicable tool for empirical studies and conceptual progress.
There are important challenges in applications of the individualized niche concept. One of them is the efficient identification of relevant niche axis. While the niche in itself is highly multidimensional, there are likely a few important niche dimensions that matter the most, when explaining individual differences. Therefore, the challenge for practitioners will be to find ways to reduce the dimensions of individualized niches to those variables, which are important for individuals. The second challenge is the efficient use of statistical models to predict fitness expectations. Nonlinearities and interactions complicate the prediction of fitness expectation (and any marginalization across individuals), so that the functional relationships need to be sufficiently well known. Linear prediction and simple (additive or multiplicative) interactions might be the first approximations in practice, but are likely overly simplified. The third challenge is the efficient use of good proxies of lifetime reproductive success in cases where it cannot be determined directly.
While we see our concept mostly of a metaphorical value, we also think it has practical implications. As a metaphorical concept, it can provide thinking aids for new scientific avenues. Importantly, we provide subcategories of the concept that, we think, may help to distinguish features that are sometimes treated under the term 'individualized niche'. We thus bring structure to the concept. We also provide practical advice on empirical quantification of the individualized niche. The realized and the trajectory-based individualized niches can be quantified quite directly, via repeated observations of the same individuals. The time-slice niche is already often quantified, in many animal personality studies, though a stronger focus on individualized phenotype-environment matches is desirable. The prospective niche can be quantified empirically by focusing on the consequences of developmental switch-points and might even provide fresh perspectives on animal behavior. The potential individualized niche is the most complicated to be measured empirically and requires some grouping of individuals with similar phenotypes, but still provides more detailed perspective of the ecological niches than Hutchinson's population niche. We hope that the individualized niche, in its different flavors, allows a more informative view of what is often treated as the niches of the population. Individuals differ and this often has ecological and evolutionary consequences. The main challenge will be the identification (and quantification) of relevant niche dimensions within the full niche space, which is characterized by high dimensionality.