1.1. Small-scale displacement: Foraging and local habitat selection 1.1.1. Process-based models of small-scale movement reveal the importance of connecting movement to fitness consequences
Animals adapt their movement for fitness-seeking activities like foraging, mating, and escaping predators that transpire on fine time and space scales (Fig. 1, section 1). Randomly walking (RW) animals, for example, can increase foraging success under uncertainty (Duffy 2011), but underperform in changing habitats where more advanced Correlated Random Walk (CRW) strategies may dominate (Morales et al. 2004; Duffy 2011). However, assuming CRW can be naïve, especially with complex vertebrate behaviour. Theory can help define movement strategy across taxa, such as classic foraging theory (Optimal Foraging Theory [OFT]; (Pyke et al. 1977; Charnov 1976; Wajnberg et al. 2013)) that argues animals uniformly optimise their space use in response to resources. However, it neglects key behaviour, such as movement capacity and habitat-dependent movement, and movement drivers, such as energy use and predator presence. Recent trait-based approaches, such as IBMs, can integrate these key behaviours and drivers and thus represent more ecologically real scenarios, shifting the OFT paradigm to the ecological stone age (Börger et al. 2008; Holyoak et al. 2008; Mueller & Fagan 2008; Giuggioli & Bartumeus 2010; Bauer & Klaassen 2013; Fagan et al. 2013).
Fronhofer, Hovestadt & Poethke (Fronhofer et al. 2013) claim small-scale movement ought to be based on behaviour and physiology of the individual, not random. Foraging IBMs often involve multiple, interacting traits, i.e. turning angle relative to resources, habitat type (Buchmann et al. 2011; Earl & Zollner 2014; Kanagaraj et al. 2013), or predators in space (Morales et al. 2004; Duffy 2011; Kranstauber et al. 2012; Bode & Delcourt 2013; Wajnberg et al. 2013; Wilson et al. 2013; Latombe et al. 2014), step length (Leising 2001; Morales et al. 2004; Kranstauber et al. 2012), and behavioural states or rules (Railsback et al. 1999; Morales et al. 2004; Giske et al. 2013; Pauli et al. 2013; Ringelman 2014), as well as physiology, i.e. biomass (Roese et al. 1991; Buchmann et al. 2011) and energy reserves, including metabolism and gut capacity (Roese et al. 1991; Guensch et al. 2001; Peck & Daewel 2007; Giske et al. 2013; Pauli et al. 2013; Wood et al. 2013) that together define the movement process. These direct links between individual and environment suggest, for animals, the environment becomes a sandbox for combined traits to persist, dominate, or adapt (Roese et al. 1991; Railsback et al. 1999; Guensch et al. 2001; Mueller et al. 2011; Pauli et al. 2013; Shepard et al. 2013; Wood et al. 2013; Ringelman 2014). As a consequence, we can assess movement under different conditions, such as patchy landscapes and within populations, using a common framework, particularly for conservation (Fahrig 2003). Estimating foraging success in space also involves multiple, cooperating behaviours like individual responses to food and group members (Giske et al. 2013; Bonnell et al. 2013), resource attractiveness (Dumont & Hill 2004; Bode & Delcourt 2013; Kułakowska et al. 2014; Earl & Zollner 2014), and individual to group movement (Dumont & Hill 2004; Couzin et al. 2005; Bonnell et al. 2013). Spatial IBMs also help bridge the gap between model and real movement patterns (Guensch et al. 2001; Bonnell et al. 2013; Mooij & DeAngelis 2003) under environmental change (Roese et al. 1991; Börger et al. 2011) and highlight traits often overlooked in classic foraging models, such as handling time and food size (Roese et al. 1991; Guensch et al. 2001; Pauli et al. 2013).
Process-based approaches, such as foraging IBMs, ought to emphasise not only movement detail at the immediate scale of survival (Roese et al. 1991), but also evolutionary processes such as fitness-seeking (Giske et al. 2013; Ayllón et al. 2016), including risky decision-making (Railsback & Harvey 2002; Laundré et al. 2014). Animals trade off activities like searching and sheltering against energy reserves, where wrong decisions inherently cost energy, or life; however, the movement trade-off against immediate mortality from a predator must consider future risks such as starvation (risk allocation hypothesis; (Lima et al. 1999)). Thus, physiology and behaviour adapted to mediate energy costs of maintenance, growth, and fitness, i.e. reproduction, may impose further, more vital constraints than simply optimising short-term foraging gain, emphasising how the individual energy budget can be pragmatic in predicting fitness outcomes from movement behaviour.
1.1.2. Foraging IBMs incorporating energetics are useful for predicting movement across scales
Despite their usefulness as foraging models, IBMs are surprisingly unfamiliar with the core underlying principle of feeding and movement, i.e. energetics. Movement capacity is often imposed, e.g. through a maximum number of steps, but not emerging directly from energetic principles (see Table 1 for overview). These limitations probably exist because energetics models can be species specific, data hungry, and parameter heavy (Sibly et al. 2013). However, incorporating energy reserves allow models to conveniently assess basic movement costs sensitive to behaviour (Roese et al. 1991; Guensch et al. 2001; Pauli et al. 2013; Bonnell et al. 2013). For example, combinations of turning angle and step length (Wilson et al. 2013), as well as movement speed (Shepard & Lambertucci 2013; Wilson et al. 2015), can generate diverse movement patterns that accumulate different costs depending on the scale of movement. Movement speed can also include additional detail, such as step frequency and length, when estimating movement costs. Miwa et al. (Miwa et al. 2015) suggest speed shares a stronger relationship with heart rate than step length alone as a proxy for energy costs (overall dynamic body acceleration, (Gleiss et al. 2011)); however, this method still only indirectly approximates energy costs from this relationship using previous data (Miwa et al. 2015). Animals trade-off energy costs against their physiology, motor control, and manoeuvre capabilities, where speed choice can play a key role when negotiating the movement task at hand, e.g. foraging versus escaping predators, and the current environmental conditions, e.g. open habitat versus proximity to refuge, which ultimately determines fitness (Wilson et al. 2015).
Food size, bite size, and intake rate are also key, but often-overlooked, energetics traits influencing movement (Roese et al. 1991; Guensch et al. 2001; Peck & Daewel 2007; Pauli et al. 2013). These detailed measures help estimate important functions such as feeding rates and digestive capacity (Peck & Daewel 2007; Giske et al. 2013) for growing animals (Roese et al. 1991; Peck & Daewel 2007) that in turn contribute to different activity states (Pauli et al. 2013; Louzao et al. 2014) and the overall time budget (Bergman, Carita et al. 2001; Merkle et al. 2014). Ultimately, this detail helps reveal functional movement limits via the ways animals adapt to changing habitat quality (Guensch et al. 2001; Pauli et al. 2013), different temperatures (Peck & Daewel 2007; Ayllón et al. 2016), or to paths of least physical resistance (Shepard et al. 2013; Wilson et al. 2012), as shown for soaring birds and costs of high-altitude migration (Hawkes et al. 2013; Sachs et al. 2013).
Energetics IBMs suggest animals use energy to connect habitat choice with survival probability under different risks in space. Guensch et al. (2001) show stream fish prefer habitats of higher energy gain and low movement costs. Railsback et al. (1999) use body condition to show stream fish trade off energy intake against habitat type and choose patches of low energy gain that maximise survival to escape fast-flowing water currents, while Pauli et al. (2013) use energetics to capture individual growth changes in response to changing resources and habitat type, and different predation and starvation risks. Energy reserves are a useful reference for animals to assess trade-offs in behaviour, such as negotiating predator and foraging risks within a landscape of fear vs. food (Laundré et al. 2014) or complementing perception ranges when resource density changes (Zollner & Lima 2005). As spatial models, energetics IBMs connect individuals with habitat and social information by using energy as a currency against direct risks, such as predators (Railsback et al. 1999; Giske et al. 2013; Pauli et al. 2013) and competing for food (Bode & Delcourt 2013; Bonnell et al. 2013). Energetics models also connect movement with time. Intake rate and bite size are intrinsically linked, suggesting handling time budgets help animals trade off energy costs and gains (Guensch et al. 2001) and shape foraging success (Roese et al. 1991). Energy drives metabolic functions on longer time scales like starvation (Railsback et al. 1999) and reproduction (Railsback et al. 1999; Guensch et al. 2001), as well as intake and growth rates (Guensch et al. 2001) as they vary under temperature and prey constraints (Peck & Daewel 2007). These trade-offs become amenable on longer time scales by affecting size classes within populations, thus influencing evolutionary adaptation to future environmental change (Ayllón et al. 2016). Models, such as IBMs, capable of capturing these vital interacting processes are thus useful frameworks using general mechanisms like energetics to link local actions, such as how animals absorb information and assess risk, with evolutionary ones, such as balancing fitness outcomes.
1.1.3. Energetics IBMs suggest foraging and habitat selection involve fine scale time trade-offs
Energetics IBMs teach us foraging is a complex relationship between costs and gains sensitive to the scale of movement, ranging from basic step length and turning angles (Wilson et al. 2013) to complex social behaviour (Bode & Delcourt 2013; Bonnell et al. 2013). Wilson et al. (Wilson et al. 2013) measured turn costs as individual energy output depending on turn angle across distance travelled, which increases at finer scales; models assuming the environment absorb these costs thus should be treated with caution. By building from the individual level, IBMs help us understand movement constraints in space and time with changing environments (Giske et al. 2013; Pauli et al. 2013) and body size (Roese et al. 1991; Railsback et al. 1999; Peck & Daewel 2007; Fiksen & Carlotti 1998), identify the subtle interface between resource availability and movement in space (Earl & Zollner 2014; Bode & Delcourt 2013), and generate sensible outcomes at the population level (Roese et al. 1991; Ayllón et al. 2016). Individual rates of food intake provide clues on how behaviour reflects energy gain as resources vary across space (Bode & Delcourt 2013). Similarly, interpreting energy gain from physiology, such as stomach capacity, connects individual traits directly to how animals perceive individual risk and thus modify movement decisions (Giske et al. 2013). The type of system also determines how to appropriately interpret energy exchange between individuals and the environment, where energy uptake is either from surfaces or volumes in terrestrial versus aquatic habitats. That is, calculating metabolic rates becomes more uniform in oceans because prey is measured in biomass. This argument also suggests energetics models are sensitive to species and scale; here, static parameters, such as maximum feeding rate, may underestimate intake rates as prey density varies in space and time (Peck & Daewel 2007). The stability of an ecosystem depends on its individual parts, so connecting crucial individual behaviour like foraging to movement should translate across scales, particularly in space (Hosseini 2006). Behaviour and ecology at the population and ecosystem levels thus relies on real estimates of cues driving individual behavior (Buchmann et al. 2011; Buchmann et al. 2012; Buchmann et al. 2013; Earl & Zollner 2014), where basic and general individual mechanisms like energy use are useful and translate to appropriate scales.
Models concerned with predicting only single outcomes, such as growth, may suffer by ignoring cooperating constraints to movement in space, such as trade-offs to predation (Railsback et al. 1999; Fiksen & Carlotti 1998) and risk-taking in landscapes of food vs. fear (Giske et al. 2013; Laundré et al. 2014). Energetics models should also consider metabolic processes other than simply energy intake (Louzao et al. 2014), such as maintenance costs (Kooijman 2010; Sibly et al. 2013) and the consequences of metabolism over longer time scales related to fitness consequences (Ayllón et al. 2016; Malishev et al. in review). Incorporating energetics in IBMs at small scales creates opportunities to connect movement with time, specifically the activity budget of animals, such as food handling rates, as well as providing a mechanism directly linked to survival probability from which animals make their activity and habitat choices. Individual traits like physiology link time budgets directly to energy intake on fine movement scales, which translates simply to behaviour. Despite this straightforward relationship, movement models at this scale commonly make energetics subjective, probably because they are easier to develop and more convenient (Holyoak et al. 2008). Advanced technology, such as GPS units with accelerometers, can simplify how we estimate energy use for different behaviours when collecting data from the field (Gleiss et al. 2011; Louzao et al. 2014; Nams 2014; Wilson et al. 2014; Miwa et al. 2015; Scharf et al. 2016). Using metabolic rules from energetics theory to formulate the individual energy budget provides a general approach to drive movement decisions and estimate movement costs, thus filling in the gaps from field data estimates of energy use as well as escaping the demands and limitations of formulating species-specific models (Grimm et al. 2005; Louzao et al. 2014). Packaging a theory-driven energy budget into an individual-based model (Malishev et al. in review) thus supplies an underlying, unifying mechanism to individual movement across taxa and scales within a general movement framework.