Repeatability is the first step in a broader hypothesis test: a comment on Stuber et al.

In their excellent piece of scholarship, Stuber et al. (2022) capture the history of animal personality and translate it for the disciplines of movement and spatial ecology. The authors effectively demonstrate that for 20 years spatial and movement ecologists have already been asking and answering the question: are spatial behaviors traits? Their striking conclusion is that spatial personalities are more repeatable than other behaviors highlighted in the keystone meta-analysis on personality (Bell et al. 2009). But asking, “does the property we measure, and define as a behavior, constitute personality” is just the first step required to position spatial behaviors within a broader behavioral and evolutionary framework. Thus, demonstrating consistent individual differences is just the beginning of the hypothesis. Nested within the next step in the hypothesis framework (Figure 1) is the question: why are spatial behaviors repeatable? And more precisely emergent from Stuber et al. (2022) is why are spatial behaviors more repeatable than other behaviors (Bell et al. 2009). At least two competing hypotheses merit falsification. Flowchart of the analytical procedure to quantify the potential for an evolutionary response from variation in behavioral phenotypes and environmental variation. Each color represents an individual. (a) Behavioral reaction norms (BRN) quantify individual-level plasticity across an environmental gradient, the slopes of BRN lines, and personality, the behavior in an average environment (the intercept, dashed line). The gold individual displays the highest value for this behavior, and the green individual the lowest. All individuals increase expression of the behavior as the environmental variable increases, but the gold individual has the strongest response. (b) Reaction norms quantify life-history characteristics related to fitness (reproductive success or survival). The green individual has the lowest fitness and the gold the highest. However, for all individuals, fitness declines as the measured environmental gradient increases. (c) By linking panels a and b, a life-history syndrome is quantified, linking individual behavior, either personality or plasticity, to the average value of fitness. The individuals that either have increased plasticity in their behavior (slope of panel a) or a higher value for the behavior in an average environment (intercept panel a), have higher fitness. (d) Behavioral reaction norms quantify adjusted repeatability, the repeatability value after accounting for variation in the environment. (1) Spatial behaviors may be more repeatable because they are subject to external constraints. For example, variance in landscape features, including binned land cover categories (e.g., forest or grassland) and other physiognomic features that restrict (e.g., mountains) or facilitate (e.g., valleys) movements could canalize spatial behavior as pseudo-personalities (Niemelä and Dingemanse 2017). For example, high repeatability and pseudo-personality emerge when individuals experience the same environments throughout their lives and spatial behavior depends on environmental context (Niemelä and Dingemanse 2017). Indeed, spatial behaviors such as territoriality, site fidelity, and recursive movements could be constrained with regard to the environment available to an individual. Such environmental differences highlight the importance of quantifying adjusted repeatability, for example, a repeatability value that controls for environmental context. (2) Spatial behaviors are confounded by pretending variables or exist within behavioral syndromes. Two themes exist in the discussion of spatial personality. First, is spatial behavior an axis of animal personality (e.g., Hertel et al. 2020; Stuber et al. 2022)? Second, is spatial behavior correlated with a traditional personality trait, that is, personality-dependent spatial behavior (sensuSpiegel et al. 2017)? Personality-dependent spatial behavior is the correlation between spatial behavior and, for example, boldness (or other traditional personality measures). Inherent to personality-dependent spatial behavior is the idea that spatial behaviors and non-spatial personality traits covary as a syndrome. However, it remains unclear the extent to which each trait depends on, or is independent, of the other. Enter the behavioral reaction norm (BRN). BRNs are the set of behaviors that an individual expresses across an environmental gradient. Behavioral (Houslay and Wilson 2017) and movement ecologists (Hertel et al. 2020) use BRNs to interrogate how among-individual differences change across environmental contexts (O’Dea et al. 2022). Importantly, BRNs are flexible and can be used to model the change in a spatial personality trait across an environmental gradient and the relationship between spatial personality traits and traditional personality traits, for example, boldness. The individual expression of spatial behavior may vary by environmental context, may be correlated across environmental contexts, and the shape of their variance may also begin to imply which environmental contexts are most likely to induce natural selection (Figure 1a,b). Within the evolutionary framework, repeatability represents the upper bound for heritability (Dochtermann and Schwab 2015), which for behaviors can be transmitted through genetics or learning. Natural selection occurs when among-individual variation in a trait drives the variation in fitness among individuals (Figure 1c), and when those traits are heritable or transmissible (Figure 1d). The two components of adaptive phenotypic evolution—genetic variation and natural selection—are therefore conditional on among-individual variation in the distribution of trait values. As a result, estimating repeatability of a trait has been used to make tentative evolutionary conclusions when formal quantitative genetic analyses are not possible (Dochtermann and Schwab 2015). When a spatial behavior is 1) repeatable and thus a proxy for heritability; 2) individuals live in variable environments in which behavior varies, that is, BRN; and 3) drives fitness variation (Figure 1), we can begin to infer adaptive phenotypic evolution and make testable predictions about the evolution of the behavior. Within the broader context of Stuber et al.’s meta-analysis, repeatability is therefore the first action in our quest to articulate precisely the mechanisms in the evolution of spatial behavior.

Fetched live from OpenAlex and de-inverted. Abstracts are not stored in this database: the inverted indexes are 8.6 GB of the frame’s 9.3 GB of text, and the host has 13 GB free.