Moving on with foraging theory: incorporating movement decisions into the functional response of a gregarious shorebird

Optimal foraging strategies for mutually avoiding competitors

Many animals are known to exhibit foraging patterns where the distances they travel in a given direction are drawn from a heavy-tailed Lévy distribution. Previous studies have shown that, under sparse and random resource conditions, solitary non-destructive (with regenerating resources) foragers perform a maximally efficient search with Lévy exponent μ equal to 2, while for destructive foragers, efficiency decreases with μ monotonically and there is no optimal μ. However, in nature, there also exist situations where multiple foragers, displaying avoidance behavior, interact with each other competitively. To understand the effects of such competition, we develop a stochastic agent-based simulation that models competitive foraging among mutually avoiding individuals by incorporating an avoidance zone, or territory, of a certain size around each forager which is not accessible for foraging by other competitors. For non-destructive foraging, our results show that with increasing size of the territory and number of agents the optimal Lévy exponent is still approximately 2 while the overall efficiency of the search decreases. At low values of the Lévy exponent, however, increasing territory size actually increases efficiency. For destructive foraging, we show that certain kinds of avoidance can lead to qualitatively different behavior from solitary foraging, such as the existence of an optimal search with 1<μ<2. Finally, we show that the variance among the efficiencies of the agents increases with increasing Lévy exponent for both solitary and competing foragers, suggesting that reducing variance might be a selective pressure for foragers adopting lower values of μ. Taken together, our results suggest that, for multiple foragers, mutual avoidance and efficiency variance among individuals can lead to optimal Lévy searches with exponents different from those for solitary foragers.

Track and dive-based movement metrics do not predict the number of prey encountered by a marine predator

BACKGROUND

Studying animal movement in the context of the optimal foraging theory has led to the development of simple movement metrics for inferring feeding activity. Yet, the predictive capacity of these metrics in natural environments has been given little attention, raising serious questions of the validity of these metrics. The aim of this study is to test whether simple continuous movement metrics predict feeding intensity in a marine predator, the southern elephant seal (SES; Mirounga leonine), and investigate potential factors influencing the predictive capacity of these metrics.

METHODS

We equipped 21 female SES from the Kerguelen Archipelago with loggers and recorded their movements during post-breeding foraging trips at sea. From accelerometry, we estimated the number of prey encounter events (nPEE) and used it as a reference for feeding intensity. We also extracted several track- and dive-based movement metrics and evaluated how well they explain and predict the variance in nPEE. We conducted our analysis at two temporal scales (dive and day), with two dive profile resolutions (high at 1 Hz and low with five dive segments), and two types of models (linear models and regression trees).

RESULTS

We found that none of the movement metrics predict nPEE with satisfactory power. The vertical transit rates (primarily the ascent rate) during dives had the best predictive performance among all metrics. Dive metrics performed better than track metrics and all metrics performed on average better at the scale of days than the scale of dives. However, the performance of the models at the scale of days showed higher variability among individuals suggesting distinct foraging tactics. Dive-based metrics performed better when computed from high-resolution dive profiles than low-resolution dive profiles. Finally, regression trees produced more accurate predictions than linear models.

CONCLUSIONS

Our study reveals that simple movement metrics do not predict feeding activity in free-ranging marine predators. This could emerge from differences between individuals, temporal scales, and the data resolution used, among many other factors. We conclude that these simple metrics should be avoided or carefully tested a priori with the studied species and the ecological context to account for significant influencing factors.

Incorporating neurological and behavioral mechanisms of sociality into predator-prey models

Consumer-resource population models drive progress in predicting and understanding predation. However, they are often built by averaging the foraging outcomes of individuals to estimate per capita functional responses (functions that describe predation rate). Reliance on per-capita functional responses rests on the assumption that that individuals forage independently without affecting each other. Undermining this assumption, extensive behavioral neuroscience research has made clear that facilitative and antagonistic interactions among conspecifics frequently alter foraging through interference competition and persistent neurophysiological changes. For example, repeated social defeats dysregulates rodent hypothalamic signaling, modulating appetite. In behavioral ecology, similar mechanisms are studied under the concept of dominance hierarchies. Neurological and behavioral changes in response to conspecifics undoubtedly play some sort of role in the foraging of populations, but modern predator-prey theory does not explicitly include them. Here we describe how some modern approaches to population modeling might account for this. Further, we propose that spatial predator-prey models can be modified to describe plastic changes in foraging behavior driven by intraspecific interaction, namely individuals switching between patches or plastic strategies to avoid competition. Extensive neurological and behavioral ecology research suggests that interactions among conspecifics help shape populations' functional responses. Modeling interdependent functional responses woven together by behavioral and neurological mechanisms may thus be indispensable in predicting the outcome of consumer-resource interactions across systems.

Sulfur in lucinid bivalves inhibits intake rates of a molluscivore shorebird

A forager's energy intake rate is usually constrained by a combination of handling time, encounter rate and digestion rate. On top of that, food intake may be constrained when a forager can only process a maximum amount of certain toxic compounds. The latter constraint is well described for herbivores with a limited tolerance to plant secondary metabolites. In sulfidic marine ecosystems, many animals host chemoautotrophic endosymbionts, which store sulfur compounds as an energy resource, potentially making their hosts toxic to predators. The red knot Calidris canutus canutus is a molluscivore shorebird that winters on the mudflats of Banc d'Arguin, where the most abundant bivalve prey Loripes orbiculatus hosts sulfide-oxidizing bacteria. In this system, we studied the potential effect of sulfur on the red knots' intake rates, by offering Loripes with various sulfur content to captive birds. To manipulate toxicity, we starved Loripes for 10 days by removing them from their symbiont's energy source sulfide. As predicted, we found lower sulfur concentrations in starved Loripes. We also included natural variation in sulfur concentrations by offering Loripes collected at two different locations. In both cases lower sulfur levels in Loripes resulted in higher consumption rates in red knots. Over time the red knots increased their intake rates on Loripes, showing their ability to adjust to a higher intake of sulfur.

A computational formulation of the behavior systems account of the temporal organization of motivated behavior

The behavior systems framework suggests that motivated behavior-e.g., seeking food and mates, avoiding predators-consists of sequences of actions organized within nested behavioral states. This framework has bridged behavioral ecology and experimental psychology, providing key insights into critical behavioral processes. In particular, the behavior systems framework entails a particular organization of behavior over time. The present paper examines whether such organization emerges from a generic Markov process, where the current behavioral state determines the probability distribution of subsequent behavioral states. This proposition is developed as a systematic examination of increasingly complex Markov models, seeking a computational formulation that balances adherence to the behavior systems approach, parsimony, and conformity to data. As a result of this exercise, a nonstationary partially hidden Markov model is selected as a computational formulation of the predatory subsystem. It is noted that the temporal distribution of discrete responses may further unveil the structure and parameters of the model but, without proper mathematical modeling, these discrete responses may be misleading. Opportunities for further elaboration of the proposed computational formulation are identified, including developments in its architecture, extensions to defensive and reproductive subsystems, and methodological refinements.

The Influence of Food Density, Flock Size, and Disturbance on the Functional Response of Bewick's Swans (Cygnus columbianus bewickii) in Wintering Habitats

Simple Summary

Changes in environmental conditions cause animals to adjust their behavioral strategies to survive. We investigated foraging behavior in different habitats of wintering Bewick’s swans. We found that the observed feeding rate was not affected by food density but showed a negative relationship with flock size and disturbance time. Handling time had a negative relationship with food density and flock size, but a positive relationship with disturbance. Searching rate was negatively correlated with food density, flock size, and disturbance time. This provides insight into how wintering waterbirds adapt their foraging behavior in complex environments.

Abstract

Perceiving how animals adjust their feeding rate under a variety of environmental conditions and understanding the tradeoffs in their foraging strategies are necessary for conservation. The Holling functional response, which describes the relationship of feeding rate and food density to searching rate and handling time, has been applied to a range of waterbirds, especially with regard to Type II functional responses that describe an increasing feeding rate with food density but at a decelerating rate as the curve approaches the asymptote. However, feeding behavior components (feeding rate, searching rate, and handling time) are influenced by factors besides prey density, such as vigilance and flock size. In this study, we aim to elucidate how Bewick’s swans (Cygnus columbianus bewickii) adopt flexible foraging strategies and vary their feeding behavior components in response to disturbance, flock size, and food density. We collected focal sampling data on the foraging behavior of swans that foraged rice grains, foxnut seeds, and tubers in paddy field, foxnut pond, and lake habitats, respectively, in Shengjin and Huangpi lakes during winter from 2016 to 2018. The observed feeding rate was not correlated with food density and displayed a positive relationship with searching rate but negative relationships with handling time, flock size, overall vigilance time, and disturbance time. Handling time was negatively correlated with food density and flock size, yet it increased with disturbance, overall vigilance time, and normal vigilance time. Searching rate was negatively correlated with food density, flock size, and disturbance time. Feeding rate was affected by the combined effects of handling time and searching rate, as well as food density and searching rate. The shape of the observed functional response could not be fitted to Holling’s disc equation. However, the disc equation of the predicted feeding rate of wintering swans was found to be driven by food density. This provides insight into how wintering waterbirds adopt appropriate foraging strategies in response to complicated environmental factors, which has implications for wildlife conservation and habitat management.

A mechanistic theory of personality-dependent movement behaviour based on dynamic energy budgets

Consistent between-individual differences in movement are widely recognised across taxa. In addition, foraging plasticity at the within-individual level suggests a behavioural dependency on the internal energy demand. Because behaviour co-varies with fast-slow life history (LH) strategies in an adaptive context, as theoretically predicted by the pace-of-life syndrome hypothesis, mass/energy fluxes should link behaviour and its plasticity with physiology at both between- and within-individual levels. However, a mechanistic framework driving these links in a fluctuating ecological context is lacking. Focusing on home range behaviour, we propose a novel behavioural-bioenergetics theoretical model to address such complexities at the individual level based on energy balance. We propose explicit mechanistic links between behaviour, physiology/metabolism and LH by merging two well-founded theories, the movement ecology paradigm and the dynamic energetic budget theory. Overall, our behavioural-bioenergetics model integrates the mechanisms explaining how (1) behavioural between- and within-individual variabilities connect with internal state variable dynamics, (2) physiology and behaviour are explicitly interconnected by mass/energy fluxes, and (3) different LHs may arise from both behavioural and physiological variabilities in a given ecological context. Our novel theoretical model reveals encouraging opportunities for empiricists and theoreticians to delve into the eco-evolutionary processes that favour or hinder the development of between-individual differences in behaviour and the evolution of personality-dependent movement syndromes.

Individual variation in functional response parameters is explained by body size but not by behavioural types in a poeciliid fish

Functional responses are per-capita feeding rate models whose parameters often scale with individual body size but the parameters may also be further influenced by behavioural traits consistently differing among individuals, i.e. behavioural types or animal personalities. Behavioural types may intrinsically lead to lower feeding rates when consistently shy, inactive and easily stressed individuals cannot identify or respond to risk-free environments or need less food due to lower metabolic rates linked to behaviour. To test how much variation in functional response parameters is explained by body size and how much by behavioural types, we estimated attack rate and handling time individually for differently sized female least killifish (Heterandria formosa) and repeatedly measured behavioural traits for each individual. We found that individual fish varied substantially in their attack rate and in their handling time. Behavioural traits were stable over time and varied consistently among individuals along two distinct personality axes. The individual variation in functional responses was explained solely by body size, and contrary to our expectations, not additionally by the existing behavioural types in exploration activity and coping style. While behavioural trait-dependent functional responses may offer a route to the understanding of the food web level consequences of behavioural types, our study is so far only the second one on this topic. Importantly, our results indicate in contrast to that previous study that behavioural types do not per se affect individual functional responses assessed in the absence of external biotic stressors.

Expectation-Maximization Binary Clustering for Behavioural Annotation

The growing capacity to process and store animal tracks has spurred the development of new methods to segment animal trajectories into elementary units of movement. Key challenges for movement trajectory segmentation are to (i) minimize the need of supervision, (ii) reduce computational costs, (iii) minimize the need of prior assumptions (e.g. simple parametrizations), and (iv) capture biologically meaningful semantics, useful across a broad range of species. We introduce the Expectation-Maximization binary Clustering (EMbC), a general purpose, unsupervised approach to multivariate data clustering. The EMbC is a variant of the Expectation-Maximization Clustering (EMC), a clustering algorithm based on the maximum likelihood estimation of a Gaussian mixture model. This is an iterative algorithm with a closed form step solution and hence a reasonable computational cost. The method looks for a good compromise between statistical soundness and ease and generality of use (by minimizing prior assumptions and favouring the semantic interpretation of the final clustering). Here we focus on the suitability of the EMbC algorithm for behavioural annotation of movement data. We show and discuss the EMbC outputs in both simulated trajectories and empirical movement trajectories including different species and different tracking methodologies. We use synthetic trajectories to assess the performance of EMbC compared to classic EMC and Hidden Markov Models. Empirical trajectories allow us to explore the robustness of the EMbC to data loss and data inaccuracies, and assess the relationship between EMbC output and expert label assignments. Additionally, we suggest a smoothing procedure to account for temporal correlations among labels, and a proper visualization of the output for movement trajectories. Our algorithm is available as an R-package with a set of complementary functions to ease the analysis.

Mapping species abundance by a spatial zero-inflated Poisson model: a case study in the Wadden Sea, the Netherlands

The objective of the study was to provide a general procedure for mapping species abundance when data are zero-inflated and spatially correlated counts. The bivalve species Macoma balthica was observed on a 500×500 m grid in the Dutch part of the Wadden Sea. In total, 66% of the 3451 counts were zeros. A zero-inflated Poisson mixture model was used to relate counts to environmental covariates. Two models were considered, one with relatively fewer covariates (model "small") than the other (model "large"). The models contained two processes: a Bernoulli (species prevalence) and a Poisson (species intensity, when the Bernoulli process predicts presence). The model was used to make predictions for sites where only environmental data are available. Predicted prevalences and intensities show that the model "small" predicts lower mean prevalence and higher mean intensity, than the model "large". Yet, the product of prevalence and intensity, which might be called the unconditional intensity, is very similar. Cross-validation showed that the model "small" performed slightly better, but the difference was small. The proposed methodology might be generally applicable, but is computer intensive.

Prey type and foraging ecology of Sanderlings Calidris alba in different climate zones: are tropical areas more favourable than temperate sites?

Sanderlings (Calidris alba) are long-distance migratory shorebirds with a non-breeding range that spans temperate and tropical coastal habitats. Breeding in the High Arctic combined with non-breeding seasons in the tropics necessitate long migrations, which are energetically demanding. On an annual basis, the higher energy expenditures during migration might pay off if food availability in the tropics is higher than at temperate latitudes. We compared foraging behaviour of birds at a north temperate and a tropical non-breeding site in the Netherlands and Ghana, respectively. In both cases the birds used similar habitats (open beaches), and experienced similar periods of daylight, which enabled us to compare food abundance and availability, and behavioural time budgets and food intake. During the non-breeding season, Sanderlings in the Netherlands spent 79% of their day foraging; in Ghana birds spent only 38% of the daytime period foraging and the largest proportion of their time resting (58%). The main prey item in the Netherlands was the soft-bodied polychaete Scolelepis squamata, while Sanderlings in Ghana fed almost exclusively on the bivalve Donax pulchellus, which they swallowed whole and crushed internally. Average availability of polychaete worms in the Netherlands was 7.4 g ash free dry mass (AFDM) m⁻², which was one tenth of the 77.1 g AFDM m⁻² estimated for the beach in Ghana. In the tropical environment of Ghana the Sanderlings combined relatively low energy requirements with high prey intake rates (1.64 mg opposed to 0.13 mg AFDM s⁻¹ for Ghana and the Netherlands respectively). Although this may suggest that the Ghana beaches are the most favourable environment, processing the hard-shelled bivalve (D. pulchellus) which is the staple food could be costly. The large amount of daytime spent resting in Ghana may be indicative of the time needed to process the shell fragments, rather than indicate rest.