Behav Ecol Sociobiol (2015) 69:1879–1896 DOI 10.1007/s00265-015-2000-3 ORIGINAL ARTICLE Collective resilience in a disturbed environment: stability of the activity rhythm and group personality in Periplaneta americana 1 1 1 Michel-Olivier Laurent Salazar & Isaac Planas-Sitjà & Jean-Louis Deneubourg & Grégory Sempo1 Received: 14 July 2015 /Revised: 21 August 2015 /Accepted: 24 August 2015 /Published online: 4 September 2015 # Springer-Verlag Berlin Heidelberg 2015 Abstract Studies on the use of aggregation sites, or commu- and induce migration to an undisturbed shelter site, the distur- nal roosts, found that these roosts have high stability but that bance did not influence the group’sglobalactivityrhythm. individuals will migrate to other sites when the original site is disturbed or otherwise no longer suitable. In this study, the Keywords Domiciliary cockroaches . Sleeping aggregation dynamics of groups of Periplaneta americana aggregations . Communal roosting . Crowding . Aggregation were examined without interruption for 5 days and 4 nights. sites . Disturbances . Group personality Our results showed that groups inside an undisturbed aggre- gation site presented stable group personalities and choice in aggregation site. The individuals that were aggregated in a site Introduction that was repeatedly disturbed by a lighting stimulus during their resting period not only exhibited stability in their group Numerous studies dedicated to the study of group living, have personalities but also a resistance to the disturbance, which found that the costs of living in groups, such as parasitism and allowed the initial aggregation site to remain the site of choice increased competition, are outweighed by the benefits, such as for a few days. Regarding the group personalities, we showed the increased probability of finding a mate (Krause and that the groups had distinct personalities with respect to the Ruxton 2002); the correct development of individuals total amount of time spent outside of aggregation sites and at (Lihoreau et al. 2012); waste reuse by group members (Mira the sources during their active period and the time spent shel- 2000; Rychtár et al. 2014); protection from the environment, tered during their resting period. For the reaction to the distur- including from desiccation or waves (Bell and Adiyodi 1982; bances, we found that the groups showed distinct personalities Focardi et al. 1985; Yoder et al. 2005;Benoitetal.2007;van between the fraction that fled and the fraction that migrated to de Koppel et al. 2008; Rojas et al. 2013; Broly et al. 2014); the undisturbed site. Our study also showed that although the possibility to hunt larger prey and better protect a food source disturbances were strong enough to trigger fleeing behaviour from competitors (Krause and Ruxton 2002;Beauchamp 2014); and protection from predators and parasitoids (Vulinec and Miller 1989;Vulinec1990;Alcock1998; Communicated by L. Keller Krause and Ruxton 2002;AggioandDerby2011;Ley- Cooper et al. 2011;KlompmakerandFraaije2012; Electronic supplementary material The online version of this article Beauchamp 2014). (doi:10.1007/s00265-015-2000-3) contains supplementary material, which is available to authorized users. The fact that these benefits generally outweigh the costs is highlighted by the wide variety of group living species. * Michel-Olivier Laurent Salazar Indeed, group living is widespread taxonomically, with many [email protected] species living in groups at various timescales and with varying degrees of interaction. For example, there are animals with 1 Unit of Social Ecology, CP 231, Université libre de Bruxelles, solitary forms and gregarious forms, and the most famous Campus Plaine, Boulevard du Triomphe, Building NO level 5, example is the desert locust Schistocerca gregaria Brussels, Belgium (Forsskål), in which the presence of one form or the other 1880 Behav Ecol Sociobiol (2015) 69:1879–1896 depends on the local population density (e.g. Maeno and sites possibly because using alternative roosting sites farther Tanaka 2011). Other animals spend only some parts of their from foraging areas would prove too costly (e.g. risk of heat life cycles in groups, with some of them spending their gre- stress) (Rogers et al. 2006). Similar results were found for garious phase in a stable site. For instance, numerous studies harvestmen, which keep using unsuitable sites in the absence have focused on the case of typically solitary species that of close alternatives (Grether and Donaldson 2007), indicating gather at specific sites for hibernation or for reproduction that a disturbed aggregation site is not necessarily abandoned (e.g. certain snakes (Gregory 1974; Shine et al. 2001;Ajtić (e.g. Parsons and Eggleston 2006) but may continue to be et al. 2013;Zappalortietal.2014), insects (Thornhill 1980; used. Although the aggregation sites themselves can be stable, Raak-Van Den Berg et al. 2013;DublonandSumpter2014; individuals can move between different aggregation sites. Durieux et al. 2014; Turner 2015) and birds (Gibson et al. Studies on the aggregation behaviour of different species 1991; Höglund and Alatalo 2014)). Other studies have fo- have shown that site fidelity is variable, with some indi- cused on the many species that form aggregations only at viduals joining aggregates at different aggregation sites certain times of the day, while inactive, and disaggregate into from 1 day to another (e.g. Heinrich and Vogt 1980; small groups or solitary individuals during their periods of Lewis 1995;Kerth2010). activity. At the beginning of the inactive period, the indi- The focus of the current study is to analyse the stability of viduals aggregate in specific sites whose use is stable shelter site choice by groups of individuals that present soli- through time, with sites used for consecutive days to years tary and gregarious periods that are closely linked to a day/ (Lewis 1995;Silvaetal.2011;Laughlinetal.2014). This night cycle under calm and disturbed conditions. Since these type of aggregation, for example, is exhibited by birds aggregation/disaggregation processes are carried out by (Beauchamp 1999), bats (Lewis 1995;Kashimaetal. groups of individuals, we analyse if it is possible to dis- 2013), gregarious caterpillars (Fitzgerald and Costa tinguish different group behaviours. This pattern of soli- 1999;Spechtetal.2007), lobsters (Aggio and Derby tary and gregarious periods is notably exhibited by domi- 2011;KlompmakerandFraaije2012), molluscs (Focardi ciliary cockroaches which are model organisms for many et al. 1985), harvestmen (Machado et al. 2000)anddomi- studies in biology, including research on aggregation (Bell ciliary cockroaches (Lihoreau et al. 2012). and Adiyodi 1982;Lihoreauetal.2012). Domiciliary If these aggregation sites are typically used as long as the cockroaches, e.g. Blattella germanica (Linnaeus) and sites are suitable, they will stop being used when disturbances Periplaneta americana (Linnaeus), are nocturnal foragers appear, either from predators or from anthropomorphic origin and form resting aggregates inside shelters during the day (e.g. Parsons and Eggleston 2006; Grether and Donaldson (Lihoreau et al. 2012). 2007). Concerning this last point, several authors have devot- Studies performed on the aggregation and on the activity ed their studies to understanding the effects of anthropogenic rhythms of domiciliary cockroaches have been mostly carried disturbances on the behaviour of species that roost near urban out with isolated or small groups of individuals for periods or anthropogenic landscapes (Carrete and Tella 2011; less than 24 h (e.g. Harker 1956;LiptonandSutherland1970; Shamoun-Baranes et al. 2011;Chudzinskaetal.2013; Roberts 1982; Rivault 1985;Belletal.2007; Canonge et al. Pirotta et al. 2015). It was notably shown that living in such 2011). These studies have shown that the collective decision environments can result in a differentiation between popula- making guiding the aggregation process is mediated by cutic- tions. For instance, bird and mammal populations living in ular hydrocarbons (Saïd et al. 2005; Sempo et al. 2006)andby their natural environment differed significantly from the urban the amplification that results from the inter-attractions among populations in their personalities, circadian rhythms and sed- individuals in a situation in which each individual has limited entary levels (Ditchkoff et al. 2006;Parteckeetal.2006; environmental information (Parrish and Edelstein-Keshet Atwell et al. 2012; Evans et al. 2012; Dominoni et al. 2014). 1999;Halloyetal.2007; Sumpter 2010). In addition, individ- Despite the obvious effects of urban environments on wild- ual survival, growth and likelihood of response to environ- life, it remains unclear why and how some species adapt to mental heterogeneities and even fleeing behaviour are linked urban settings (Carrete and Tella 2011). Indeed, anthropogenic to the size of the aggregate in domiciliary cockroaches perturbations affect foraging activity (Chudzinska et al. 2013; (Wharton et al. 1968;Sempoetal.2009;Lihoreauetal. Pirotta et al. 2015), flight distance (Carrete and Tella 2011; 2012). Moreover, regarding the behaviour of P. americana Price et al. 2014) and alarm or vigilance behaviour (Wright when confronted with a disturbance, Laurent Salazar et al. et al. 2010; Price et al. 2014). For example, studies on the great (2013) analysed the effect of group size on the group-fleeing knot (Calidris tenuirostris (Horsfield)) and the red knot