Model selection and model averaging in behavioural ecology: the utility of the IT-AIC framework

Behavioural ecologists often study complex systems in which multiple hypotheses could be proposed to explain observed phenomena. For some systems, simple controlled experiments can be employed to reveal part of the complexity; often, however, observational studies that incorporate a multitude of causal factors may be the only (or preferred) avenue of study. We assess the value of recently advocated approaches to inference in both contexts. Specifically, we examine the use of information theoretic (IT) model selection using Akaike’s information criterion (AIC). We find that, for simple analyses, the advantages of switching to an IT-AIC approach are likely to be slight, especially given recent emphasis on biological rather than statistical significance. By contrast, the model selection approach embodied by IT approaches offers significant advantages when applied to problems of more complex causality. Model averaging is an intuitively appealing extension to model selection. However, we were unable to demonstrate consistent improvements in prediction accuracy when using model averaging with IT-AIC; our equivocal results suggest that more research is needed on its utility. We illustrate our arguments with worked examples from behavioural experiments.     

Feeding and non-feeding aggression can be induced in invasive shore crabs by altering food distribution

Human activity can have a large impact on surrounding ecosystems. For example, humans alter resource distributions for other species, potentially modifying these species competitive dynamics. These changes in local competitive processes are frequently associated with species invasions. Here, we investigate how differences in resource distribution affect competitive behaviour using the highly invasive European shore crab (Carcinus maenas). Using a controlled laboratory experiment in combination with behaviour assays and social network analysis, we show that individuals feeding in habitats with clumped food distributions are more aggressive than individuals feeding in habitats where food is evenly dispersed, and this aggression is present even on days where crabs are not feeding. Additionally, this persistent aggression can be induced, suggesting that individuals of this invasive species possess the flexibility to modify their competitive behaviours in response to differences in food distributions. Furthermore, we show how these individual responses can lead to changes in overall organisation of aggressive interactions within a population. We discuss these results in relation to how human impacts can have long-term effects on competitive behavioural strategies, and how behavioural flexibility can allow invasive species to colonise and persist in highly impacted sites such as urban ecosystems.     

Social networks and models for collective motion in animals

The theory of collective motion and the study of animal social networks have, each individually, received much attention. Currently, most models of collective motion do not consider social network structure. The implications for considering collective motion and social networks together are likely to be important. Social networks could determine how populations move in, split up into and form separate groups (social networks affecting collective motion). Conversely, collective movement could change the structure of social networks by creating social ties that did not exist previously and maintaining existing ties (collective motion affecting social networks). Thus, there is a need to combine the two areas of research and examine the relationship between network structure and collective motion. Here, we review different modelling approaches that combine social network structures and collective motion. Although many of these models have not been developed with ecology in mind, they present a current context in which a biologically relevant theory can be developed. We argue that future models in ecology should take inspiration from empirical observations and consider different mechanisms of how social preferences could be expressed in collectively moving animal groups.     

Experience with chemotactile cues indicating female feeding history impacts male courtship investment in the wolf spider <Emphasis Type="BoldItalic">Schizocosa ocreata</Emphasis>

Male fitness is often determined by the ability of the male to gain access to multiple mates, although in species that exhibit sexual cannibalism, males might increase their likelihood of being cannibalized with each encounter. This risk should create selection for males who are able to perceive potential risks associated with mating encounters. We studied male Schizocosa ocreata wolf spiders to determine whether they use female chemotactile cues (silk, excreta) as indicators of potential risks and how these cues affected subsequent male courtship behaviors. Female treatments included satiated females vs. starved females, as well as a treatment where females had recently cannibalized a male S. ocreata. We performed experiments to assess (1) if males use female chemotactile cues to determine potential risks associated with differing female feeding treatments and alter courtship investment, and 2) how male experience with female chemotactile cues affected courtship investment in subsequent female encounters. At first encounter, males do not vary courtship investment (number of bouts, duration, and vigor) with different female feeding treatments. However, male behaviors during subsequent encounters with female chemotactile cues varied in complex ways, depending on female feeding treatment and male first encounter experience. These data suggest that male experience impacts perception of female chemical cues and offers the male opportunity to avoid sexual cannibalism in high-risk situations.     

Spectral reflectance and communication in the wolf spider, <Emphasis Type="Italic">Schizocosa ocreata</Emphasis> (Hentz): simultaneous crypsis and background contrast in visual signals

Animal color patterns often reflect a compromise between natural selection for crypsis or inconspicuousness to predators and sexual selection for conspicuousness to potential mates. In leaf litter-dwelling wolf spider species like Schizocosa ocreata, body coloration often closely matches the background coloration of a generally brown environment. However, body parts used in communication should exhibit high contrast against background coloration. We used spectral analysis to examine male and female S. ocreata for matching and contrasting coloration against leaf litter. Values were plotted in multivariate color space, based on reflectivity in different frequency ranges. When viewed from above, colors of both males and females overlap with values for dead brown leaf litter and soil, suggesting cryptic coloration when viewed by potential predators. However, when viewed from a lateral perspective, both males and females show color values that are polar opposites of litter backgrounds, suggesting higher contrast when viewed by other spiders. Moreover, male secondary characters used in visual signaling by S. ocreata (tibia brushes) show the highest level of background contrast. These findings suggest that S. ocreata wolf spiders have color patterns that provide both crypsis and background contrast at the same time, depending on receiver viewing perspective.     

Brood parasitism disproportionately increases nest provisioning and helper recruitment in a cooperatively breeding bird

Obligate avian brood parasites lay their eggs in nests of other species (hosts), which raise parasitic young. Parasitic nestlings are likely to influence host’s parental behaviours as they typically beg for food more vigorously than young host for a given hunger level. However, few studies have tested this idea, with conflicting results. These prior studies were largely limited to biparental hosts, but little is known about the effect of brood parasitism on parental behaviours in hosts that breed cooperatively. We followed a multimodel approach to examine the effect of brood parasitism on nest provisioning and helper recruitment in the baywing (Agelaioides badius), a cooperative breeder parasitised by screaming (Molothrus rufoaxillaris) and shiny (Molothrus bonariensis) cowbirds. Multimodel inference results indicated that feeding visits increased with nestling age, cooperative group size and number of cowbird nestlings in the brood. Brood size had little influence on feeding visits, which further suggests that baywings adjusted their provisioning effort in response to cowbird parasitism. In addition, nests parasitised artificially with shiny cowbird eggs or hatchlings recruited more helpers than unmanipulated nests having only host or screaming cowbird young. Our results provide novel evidence that brood parasitism and cooperative breeding interact in determining the levels of nest provisioning.     

Group leadership depends on energetic state in a nomadic collective foraging caterpillar

Group living is a common strategy among animals and has arisen independently in over 300 species of Lepidoptera. Yet, activity synchrony between individuals is necessary to derive the benefits that ensue from an aggregated lifestyle. Which individuals decide which activities to perform and when to perform them is, therefore, a fundamental question. In some species of social caterpillars and sawflies, the role of a potential behavioral polyethism between individuals has been suggested, whereby certain individuals are consistently more likely to initiate and lead a foraging event. However, in these cases, evidence in support of division of labor is lacking. This study was undertaken to determine if certain individuals of Malacosoma disstria are more likely to be consistent group leaders or if transient leaders could be predicted by the differences in energetic states between individuals. The results of this study indicate that unfed caterpillars initiate foraging bouts and are more likely to lead locomotion. There was no size or sex-based bias in those individuals that acted as temporary leaders. Consistent behavioral differences between individuals, if they exist, are therefore not necessary to explain task allocation and synchronization during foraging in this species.     

Stereotypy and variation of the mating call in the Lusitanian toadfish, <Emphasis Type="Italic">Halobatrachus didactylus</Emphasis>

Signal attributes should show different degrees of variability depending on the information to be conveyed. Species identity is usually associated with stereotyped features of a signal, whereas other types of information such as individual quality and motivation are associated with signal plasticity. Lusitanian toadfish males form aggregations during the breeding season and emit a tonal advertisement call (the boatwhistle) to attract mates to their nests. We test the hypothesis that the boatwhistle can convey information both on individual identity and motivation by checking how signal parameters vary with time. We study how the physical (tide level) and social (calling alone or in a chorus) environments and male calling rate affect this advertisement signal and how all these external and internal factors (environment, social and male motivation) blend to modulate the Lusitanian toadfish’s advertisement call. Boatwhistles of each male were very stereotyped in short periods of time (minutes), but intra-male signal variability greatly increased in a longer time scale (days). Nevertheless, significant differences among males could still be found even in a long time scale. Pulse period was the acoustic feature that most contributed to discriminate among males. Tide level and male calling rate modulated boatwhistle characteristics, and there was a differential effect of tide on call attributes depending on male calling rate. Social acoustic environment only affected calling rate. These results suggest that inter-individual differences in call characteristics and call plasticity may mediate both male–male assessment and mate choice.     

Information-theoretic approaches to statistical analysis in behavioural ecology: an introduction

Scientific thinking may require the consideration of multiple hypotheses, which often call for complex statistical models at the level of data analysis. The aim of this introduction is to provide a brief overview on how competing hypotheses are evaluated statistically in behavioural ecological studies and to offer potentially fruitful avenues for future methodological developments. Complex models have traditionally been treated by model selection approaches using threshold-based removal of terms, i.e. stepwise selection. A recently introduced method for model selection applies an information-theoretic (IT) approach, which simultaneously evaluates hypotheses by balancing between model complexity and goodness of fit. The IT method has been increasingly propagated in the field of ecology, while a literature survey shows that its spread in behavioural ecology has been much slower, and model simplification using stepwise selection is still more widespread than IT-based model selection. Why has the use of IT methods in behavioural ecology lagged behind other disciplines? This special issue examines the suitability of the IT method for analysing data with multiple predictors, which researchers encounter in our field. The volume brings together different viewpoints to aid behavioural ecologists in understanding the method, with the hope of enhancing the statistical integration of our discipline.     

Hovering guards of the stingless bee <Emphasis Type="Italic">Tetragonisca angustula</Emphasis> increase colony defensive perimeter as shown by intra- and inter-specific comparisons

Social insects need to defend their nest against robbery, parasitism and predation. The stingless bee Tetragonisca angustula is unique in that it has guards that hover near the nest entrance in addition to guards that stand at the entrance. We tested both the general hypothesis that hovering guards increase the effectiveness with which flying intruders are detected and the specific hypothesis that hovering guards improve the detection of workers of the obligate robber bee, Lestrimellita limao. In an intraspecific study comparing colonies, we found a strong positive relationship between the number of hovering guards and the distance at which a dummy robber bee or L. limao worker, experimentally moved towards the nest entrance, was detected. These results were mirrored in an interspecific study showing that four species of stingless bees with similar population colonies but which lacked hovering guards, detected L. limao only at the nest entrance, in contrast to T. angustula. In addition, we found that a greater number of attacks by guards occurred when dummies were impregnated with citral, a major component of L. limao mandibular gland odour. Our results support the hypothesis that T. angustula hovering guards increase the detection perimeter for flying intruders, especially L. limao.