Individual Thigmotactic Preference Affects the Fleeing Behavior of the American Cockroach (Blattodea: Blattidae)

Individual Thigmotactic Preference Affects the Fleeing Behavior of the American Cockroach... Positive thigmotactic behavior is associated with the ability to hide from predators and is important to explain aggregation and collective patterns in various animals. For example, this behavior has been observed in woodlice, domiciliary cockroaches, ants, and fish. Lately, research on different species is focused on the importance of animal personality for ecological and evolutionary processes, individual fitness and group cohesion. In fact, it is generally expected to find some degree of interindividual consistent differences for a behavior, unless specific circumstances, like predator attacks, hide the presence of personalities. In this research, we analyzed the individual thigmotactic preference of domiciliary cockroaches (Periplaneta americana (Linnaeus, 1758) (Blattodea: Blattidae)) and how it affected the fleeing behavior of isolated individuals inside a shelter after receiving a light stimulus. We notably highlight how isolated individuals show different consistent preferences regarding their position in the shelter, which is due to the individual thigmotaxis level, before the fleeing behavior. During the fleeing itself, cockroaches nearer to the wall, and therefore with more positive thigmotaxis, showed slower reaction lantencies to the stimulus. We propose that thigmotaxis homogenizes the interindividual differences among individuals and is important to explain the individual and collective fleeing behavior. Key words: thigmotaxis, disturbance, fleeing behavior, animal personality, domiciliary cockroach Landscape heterogeneities can affect the spatial distribution of is a kinetically efficient behavior, as has been shown by theoretical organisms by influencing their movement patterns. For instance, research (Calandre et al. 2014). habitat edges or patch boundaries can help the aggregation or Recently, several authors have studied thigmotaxis in the con- modify the movement patterns of animals with positive thigmotaxis text of individual behavioral differences (Carlson and Langkilde (Jeanson et al. 2003). Positive thigmotactic behavior, or the tendency 2013, Webster and Laland 2015). Behavioral variability and plasti- to remain close to walls or edges, is widespread in different taxa city have evolved, and are maintained, through numerous processes (Grossen and Kelley 1972, Kallai et al. 2007, Sharma et al. 2009) and that ultimately allow animal populations to adapt to different en- can be modulated by several factors. For example, positive thigmo- vironmental conditions (Dall et  al. 2004, Sih et  al. 2004, Carrete taxis decreases in a familiar environment and during exploration, and Tella 2010, Dingemanse et al. 2010, Smith and Blumstein 2010, and increases under stressful situations (Durier and Rivault 2003, Adriaenssens and Johnsson 2013). In this study, we define person- Sharma et al. 2009). In addition, positive thigmotactic behavior is a ality as the presence of behavioral differences between individuals crucial factor to explain and understand the aggregation and move- that are consistent over time for a specific behavior (Dall et al. 2004, ment of several species such as woodlice, domiciliary cockroaches, Bell et  al. 2009, Dingemanse and Wolf 2010, Réale et  al. 2010, ants, or fish (Dussutour et al. 2005, Bell et al. 2007, Devigne et al. Webster and Ward 2011, Adriaenssens and Johnsson 2013, Planas- 2011, Boulay et  al. 2013, Webster and Laland 2015) and can be Sitjà et al. 2015). The presence of these personalities is widespread used as a proxy for behavioral traits such as boldness or shelter use and raises the complexity of group dynamics. For example, the (Carlson and Langkilde 2013, Webster and Laland 2015). One pos- fission–fusion dynamics of groups of bats (Kerth 2010) and birds sible reason why thigmotaxis has been selected for so many species (Aplin et al. 2013) is influenced by the personality of each individual; is that it accelerates a search process in certain conditions, and thus the aggregation dynamics of Periplaneta americana (Linnaeus, © The Author(s) 2018. Published by Oxford University Press on behalf of Entomological Society of America. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/ licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 2 Journal of Insect Science, 2018, Vol. 18, No. 1 1758)  (Blattodea: Blattidae) is influenced by individual and group et  al, 2013). The floor of the arena was covered with white paper personalities (Planas-Sitjà et al. 2015); and individual personality in (120 g/m ), which was changed after each experiment to prevent any stickleback fish strongly affected the group’s structure, movement chemical marking. dynamics, and foraging behavior (Jolles et al. 2017). Regarding flee- A plastic ring (interior diameter: 25 cm, height: 4.5 cm, width: ing behavior, different studies have found the existence of behavioral 5mm), central within the arena, constituted the shelter. Two open- syndromes (Carter et al. 2010, Niemelä et al. 2012, Seltmann et al. ings of 3 × 1.5 cm placed symmetrically opposite to each other were 2012, Cooper Jr and Blumstein 2015) and that some events, like an the only ways out of the shelter. A glass cover was placed on top of increase of risk or stress, might break apart individual personalities the ring to allow the observation of the interior. (Niemelä et al. 2012, Cote et al. 2013, Wright et al. 2017). A light bulb (Philips [Amsterdam, The Netherlands] A55 FR, 100 It is in this context, using a gregarious species, that we performed W) centered on the arena at a height of 50 cm above the glass was the current study aiming to analyze the presence, or absence, of person- switched on at the beginning of each experiment. When the light was ality in the context of thigmotactic behavior, and its effects during flee- turned on, only the inside of the shelter received ~1,700 lx. (luxmeter ing, of an isolated individual. Our biological model was the negatively Testo 545, Testo NV/SA [Lenzkirch, Germany], resolution: 1 lx). The phototactic American cockroach (P.  americana), which forms aggre- rest of the arena, and of the room, was only illuminated by red light gations in dark, warm, and damp places during daylight hours (Bell (0 lx). This arrangement allowed us both to see and to keep the room and Adiyodi 1982, Canonge et al. 2011). P. americana, when suddenly ‘dark’ for the cockroaches (Mote and Goldsmith 1970). illuminated in their resting place, initiate a fleeing response (Okada and Toh 1998, Domenici et al. 2008). Domiciliary cockroaches have a Experimental Procedure very strong tendency to follow a wall, even while fleeing (Camhi and Only adult males (without external damages such as wing damage Johnson 1999, Durier and Rivault 2003). Consequently, knowing that or missing leg segments) were tested to exclude any behavioral var- domiciliary cockroaches express different personalities during their iations related to the ovarian cycle (Paterson and Weaver 1997). sheltering process (Planas-Sitjà et al. 2015), we used their light avoid- Isolated individuals (24 replicates) were randomly taken from the ance behavior to investigate the relationship between thigmotaxis and rearing room and placed the day before the trial in plastic containers the possible behavioral consistency before and while fleeing. (36 × 24 and 14 cm in height). They were kept in the dark until the beginning of their respective trials. Each cockroach was tested three times in total, with a 2-d interval between tests. The cockroaches Materials and Methods had access to dog pellets (Tom & Co., Delhaize Group, Brussels, Rearing Conditions Belgium) and to a piece of cotton soaked in water. The cockroaches were taken from the rearing room of the Université Before the start of each trial, the arena was cleaned with dena- libre de Bruxelles (ULB). P.  americana has been reared in the ULB tured ethanol, and a new sheet of paper was placed on the floor. since 2002 in five Plexiglas vivaria (80 × 40 cm and 100 cm high) in After 10 min, the individual was introduced under light CO narcosis which cardboard tubes are hung from the walls to serve as shelters. into the shelter (the openings were closed), as is standard procedure Groups of around 20 individuals were weekly exchanged between (Jeanson et al. 2003, 2005; Halloy et al. 2007), and remained there vivaria. Each vivarium had around 1,000 individuals, comprising for 30 min. After this waiting period, the openings were opened, and males, females, and larvae from all developmental stages. They were the light was turned on. A  webcam was used to record the cock- provided with dog pellets and water ad libitum. The rearing room roach’s behavior inside the shelter (20 frames/s) and recorded the was kept at 26 ± 1°C with a photoperiod of 12:12 (L:D) h. individual’s response for 5 min or until the cockroach had left the shelter. As noted above, during the night, P. americana are active and Experimental Set-Up leave their shelter to forage (Bell and Adiyodi 1982). For this reason, Our study was carried out using the experimental protocol described all trials were performed during daylight hours corresponding to the in detail in the study by Laurent Salazar et al. (2013). The experi- resting phase of P. americana. ments were performed in a circular arena (Fig. 1) that had an elec- tric fence to prevent the cockroaches from escaping (Laurent Salazar Analysis From the video recordings, we analyzed the following parameters: (1) The position (radius from the center of the shelter), orientation, and contact with the wall of each individual prior to the turning on of the light (t ). To analyze the orientation of the individuals, we calculated the angle between the head–abdomen axis of the cockroach and the axis going from the shelter wall to the center of the arena and passing through the middle of the cockroach. If the angle was comprised between +45° and −45°, we consid- ered that the individual was facing toward the shelter wall (Supp material [online only]). (2) The reaction time (RT) for each individual was the time inter- val between the turning on of the light (t ) and the initiation of its fleeing behavior. We considered an individual to be reacting when it rotated its body or moved forward/backward. When an individual was walking prior to t (0.32 of all trials), it was removed from the analyses. We consider that the immobile/ Fig.  1. Experimental set-up. Individual cockroaches were placed inside the mobile states are different because mobile individuals (already plastic ring. Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Journal of Insect Science, 2018, Vol. 18, No. 1 3 walking) were more prompt to flee than the immobile ones when the wall of the shelter prior to t did not significantly differ between light was turned on. days (Table 1). When we compared the number of times each indi- vidual was moving, and the number of times it was facing toward We used survival curves (or decay curves) to analyze the distribution the shelter wall, against a theoretical random distribution, we found of these parameters. Although the name includes the term ‘survival’, no significant difference (Table 2). However, our results showed that it can be used to analyze any kind of experiment where the results are individuals significantly differed from each other and presented dis- expressed as a time to an end point. For the comparison of survival tinct preferences regarding their position respective to the center of curves, we used a log-rank test (Motulsky 1999)). For the analysis of the shelter (Figs 2a and 3; Table 2). curve fittings and linear regressions, we used Kolmogorov–Smirnov This personality regarding their position has interesting conse- test and F-test, respectively. quences for their thigmotactic preferences. Since individuals touch- To compare distributions, we used chi-square test when the condi- ing the wall did so with their antennae, it was their length that tions allowed it, otherwise we used two-sample Kolmogorov–Smirnov limited the distance where touching was possible. Indeed, the proba- tests. Since our values did not meet the conditions for parametric bility of touching the wall follows a logistic function and shows that tests, we used Kruskal–Wallis and Mann–Whitney tests to compare individuals placed more than 4.6 cm from the wall (7.9 cm from the differences between days and individuals, depending if more than one center) never touch the wall, while those closer than 4.6 cm do comparison was being carried out or if the values were paired. We used the Kendall’s coefficient of concordance (W ) for concord- 1 Touch= ,. R = 08, 10.7(r − adius) ance assessment (Kendall 1938; Kendall and Smith 1939). Kendall’s 1 + e coefficient of concordance (W ) compares the stability of rank posi- two-sample Kolmogorov–Smirnov test: D = 0.18, P = 0.17; Fig. 4). The tions for each group during the trials. The values of W range from high steepness of the fitting (coefficient value = 10.7) clearly shows how 0 (no concordance of ranks) to 1 (complete concordance). We com- touching or not touching the wall is close to an all-or-none response. pared the observed W coefficients with the ‘Kendall random distribu- In other words, all individuals within reach of the wall, will touch it. tion’ (KRD) as explained by Planas-Sitjà et al. (2015). The KRD is the This allowed us to divide the individuals into two groups regarding theoretical distribution of the W coefficients for random rank orders their thigmotactic preference: high thigmotactic level (<4.6 cm from the of the same number of experimental groups and repetitions (e.g., 24 wall) and low thigmotactic level individuals, HTL and LTL, respectively. groups and 3 repetitions and N = 1,000). We performed a Z-test to Across 3 d, we observed 28 LTL individuals and 44 HTL individuals. test the significance of the difference between the observed W coeffi- In addition, we analyzed the influence of the thigmotactic level on cients and the corresponding KRD (Zar 2010, Laurent Salazar et al. cockroach orientation. In the case of the LTL individuals (N = 5/28 2015, Planas-Sitjà et al. 2015). In this case, we assigned the value of were facing toward the wall), there was no significant difference 0 s to the RT of immobile individuals. This was necessary to be able between our observations and theoretical random expectation of to quantify personality traits and allowed us to consider the idea of 0.25 (one-tailed binomial test: P = 0.26). However, the HTL individ- individuals being more prompt to react (with an RT value of 0) when uals (N = 17/44) had a significant tendency of orienting themselves analyzing their reaction personality (e.g., see whether an individual toward the shelter wall (Binomial test: P = 0.032). was always ready to flee, already moving or slow to react). To analyze the influence of the contact with the wall on cock- roach orientation, we used a one-tail binomial test to compare their During the Disturbance observed orientation and a theoretical random expectation of facing The relationship between the thigmotactic level and RT among the toward the wall (P = 0.25). immobile individuals (22 distinct individuals were immobile; during the 3 d, we observed 49 immobile individuals) was clear. Indeed, immobile HTL individuals (31 individuals) differed in their RT from Results the LTL (18 individuals) ones (Mann–Whitney test (mean ± SD): Before the Disturbance RT HTL (27.4 s ± 38) and RT LTL (7s ± 8.2): U = 165, P = 0.024). The distribution of the position of the individuals within the shelter, These results indicate clearly that the thigmotactic level of indi- the number of moving individuals, and the number facing toward viduals during rest has an influence on their fleeing behavior: HTL Table 1. Results of the interday behavioral comparison Day I Day II Day III Comparison between days Spatial distribution Homogenous Chi- Homogenous Chi-square: Homogenous Chi- No difference Chi-square: 2 2 2 2 square: X  = 1.69 X  = 4.91, P = 0.18 square: X  = 0.29, X  = 8.2, P = 0.22 3 , 3 3 6 P = 0.64 P = 0.96 Individuals facing toward the shelter wall 7 8 7 No difference Chi-square: X  = 0.13, P = 0.94 Number of moving individuals 8 6 9 No difference Chi-square: X  = 0.89, P = 0.64 RT:/ λ = 1 RT Exponential distri- Exponential distribution Exponential distri- No difference X  = 1, log- 2 2 2 λ = 0.08 ± 0.02 per s bution R  = 0.83; R  = 0.96; Kolmogorov– bution R  = 0.92; rank test: P = 0.60 Kolmogorov–Smirnov Smirnov test: D = 0.11, Kolmogorov–Smirnov test: D = 0.18, P > 0.99 test: D = 0.13, P > 0.90 P > 0.99 The table shows the daily result for each measure and their comparison between days. Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 4 Journal of Insect Science, 2018, Vol. 18, No. 1 Table 2. Summary of the interindividual differences Presence of Consistent interindi- interindividual difference? vidual difference? Spatial distribution Yes Kruskal–Wallis test: Yes Kendall’s H  = 41.98, P = 0.009 W = 0.58, Z = 3.15, P = 0.0016 Individuals facing No chi-square: X  = 2.6, toward the shelter P = 0.27 wall Number of moving No chi-square: X  = 1.83, individuals P = 0.61 RT No Kruskal–Wallis test: No Kendall’s H  = 20.97, P = 0.34 W = 0.39, Z = 0.7, P = 0.48 The table shows the presence of interindividual difference and its consist- ency for each measure. Only the spatial distribution of individuals showed interindividual difference, which was consistent across days. individuals were four times slower to react than LTL individuals (Fig.  3). Nevertheless, we found no consistency in RT within indi- viduals: individuals’ RT rankings varied each day (Fig. 2b; Table 2). Discussion Fig. 2. (a) Average radius (+SD) (the distance from the center of the shelter to It is well known that numerous species, including domiciliary cock- the cockroach’s body, 0 = center of the shelter). (b) Average RT (+SD) of each roaches like the American cockroach, show positive thigmotaxis, immobile individual. and that this thigmotaxis affects the exploration of the environment as well as the individual and collective behavior (Camhi and Johnson the menace (e.g., kind of predator) or foraging strategy (Hendrie 1999, Jeanson et al. 2003, Okada and Toh 2006, Kallai et al. 2007, et al. 1998). Baba et  al. 2010). In this study, we show that prior to a light dis- The RTs across days were not consistent between individuals, in turbance, individuals of P.  americana displayed clear personalities contradiction with previous studies (McDermott et al. 2014; Stanley regarding their thigmotactic preference (being in range to touch the et al. 2017) showing consistency in the reaction to a disturbance. It wall, <4.6  cm from the wall, or not) and thus demonstrating that is possible that individuals had different reaction threshold to light thigmotaxis is a quantifiable personality trait in P. americana. disturbances, but the magnitude of the disturbance in our study In addition to this, we observed that high thigmotactic level indi- was such that it overshadowed these differences. Further experi- viduals tended to orient themselves toward the wall. It is possible ments with different light intensities could provide insights on this that inside a shelter an individual cockroach will orient itself auto- matically toward a wall when near to it. However, we are inclined toward the hypothesis that this orientation increases their thigmot- actic level when in their resting phase. Indeed, in experiments where cockroaches explore a bigger arena, they position themselves par- allel to the arena wall, even when touching it (M-O Laurent, Isaac Planas-Sitjà, personal observation). This is different from our current observations. It is possible that the wall has a different influence on a cockroach’s behavior depending on the context. Certainly, in an open and novel space, stressed individuals will tend to be parallel to a wall, move slower, and follow it for protection (Durier and Rivault 2003, Sharma et al. 2009). However, inside a shelter, individuals are arguably less stressed and do not need to position themselves parallel and against the wall. We clearly show how, for immobile individuals, being in contact with the wall (i.e., high individual thigmotactic preference) had an impact on their RTs, with individuals touching the wall being signif- icantly slower to react. It is possible that such individual differences regarding their position could be the result of different survival strat- egies. An individual near a wall that reacts slower would be safer by being more cryptic, and running along the wall later to find shelter if needed (Durier and Rivault 2003, Sharma et al. 2009), while an individual further away from the wall would flee faster and find shel- Fig.  3. Boxplot comparing the RT to a light stimulus between HTL and LTL individuals. ter. These different strategies would be advantageous depending on Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Journal of Insect Science, 2018, Vol. 18, No. 1 5 Fig. 4. Contact with the wall of the shelter before the disturbance as a function of the position of each individual (1 = touching, 0 = not touching the wall). The position of each individual is represented as the distance from the center of the shelter to the cockroach’s body (i.e., radius). hypothesis. In addition, the lack of consistency in our study could were able to quantify different traits for individuals without the be due to the thigmotactic behavior, which was missing in the men- collective context having a potential impact. The quantification of tioned previous works. We have already seen that thigmotaxis can personality on individual thigmotaxis level shed some light on the be influenced by the environment, for example, in domiciliary cock- proximal mechanisms that affect the presence of both individual roaches the presence of a wall modifies their exploratory behavior, and collective personality shown in previous studies on collect- reducing their speed and increasing their number of stops (Durier and ive dynamics under different conditions in P.  americana (Laurent Rivault 2003). Furthermore, it has been shown that being in novel Salazar et al. 2015, Planas-Sitjà et al. 2015). As positive thigmotaxis environments increases the tendency of being positively thigmotactic has been shown to have evolved in many species and to be impor- (Simon et al. 1994, Durier and Rivault 2003, Sharma et al. 2009). tant for their behavior and survival (Grossen and Kelley 1972, Considering these results, we hypothesize that the lack of consist- Kallai et al. 2007, Sharma et al. 2009), we believe that the existence ency regarding the RT can be explained by the fact that thigmotaxis of individual thigmotactic preferences could be extensible to many reduces and homogenizes the interindividual differences between of these species. How these preferences are maintained within a individuals. Indeed, if individuals slow their exploration near walls, group and how they affect the collective behavior are key questions and novel environments increase their preference to be near a wall, for the understanding of the behavioral ecology of gregarious spe- then it is possible that in our experiments (new environment) indi- cies. The presence of personalities within a group has been seen to viduals increased their tendency to be close to the wall. Thus, indi- have an important impact on the performance of the group during viduals that would otherwise display different fleeing thresholds will collective behaviors (Brown and Irving 2014, Cronin 2015). There be positioned close to the wall and show slower RTs when the light is is no doubt that such individual variability also influences collect- turned on, homogenizing the RT. Experiments with a higher framing ive fleeing behavior. Further studies on gregarious insects and their rate (higher than 20 fps) could shed some light on this matter, being individual preferences are indeed needed to acquire a complete view precise enough to detect interindividual differences in RTs despite of their collective behavior and elucidate how these personalities the homogenization caused by the thigmotactic behavior. could be an evolutionary benefit for the collective fleeing response Regarding the global positioning, we observed that the distri- to disturbances. For instance, studies on the positional preferences bution within the shelter prior to the disturbance was homogene- would allow us to create nonrandom groups consisting of individu- ous. In a prior study, Laurent Salazar et al. (2013, Supp Material als with personality regarding their positioning and their responses [online only]) showed that isolated and pairs of individuals were to the stimulus. In such a way, it would be interesting to not only also homogeneously positioned within a shelter, however individ- link these personality traits with collective fleeing dynamics but also uals in larger groups had a distribution significantly different from to other behaviors in different situations, such as collective shelter a theoretical homogeneous distribution, with an increase in the choice, aggregate stability, or food search. Indeed, studies testing number of individuals close to the shelter wall compared with the such groups could help elucidate the influence of group composition theoretical distribution. This observation, albeit in need of further in collective dynamics, especially in research aiming at understand- research, already gives us a glimpse of the difficulties in studying ing the link between individual preferences, individual preference individual preferences in a collective context. By being numerous, amplification, and collective behavior. due to the gregarious behavior, it is possible that their individual preferences for the wall were amplified (Dussutour et  al. 2005), Supplementary Data leading to the heterogeneous distribution in favor to being closer to the wall. Supplementary data are available at Journal of Insect Science online. Previous studies on fleeing responses have been able to concen- trate on the consistency of many behavioral and kinematic traits Acknowledgments (e.g., in vertebrates (Marras et al. 2011; Hitchcock et al. 2015) and invertebrates (Stanley et  al. 2017)). However, to our knowledge, M-OLS and IP-S were funded by a PhD grant from FRIA (Fonds pour la there is an absence of works concerned about personality traits dur- Recherche dans l’Industrie et dans l’Agriculture). JLD is Research Director from the Belgian National Fund for Scientific Research (F.N.R.S). ing fleeing and kinematic traits in insects. In our present study, we Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 6 Journal of Insect Science, 2018, Vol. 18, No. 1 into groups of cockroaches to control self-organized choices. Science. 318: References Cited 1155–1158. Adriaenssens, B., and J. I. Johnsson. 2013. Natural selection, plasticity and the Hendrie, C. A., S. M.  Weiss, D.  Eilam. 1998. Behavioural response of wild emergence of a behavioural syndrome in the wild. Ecol. Lett. 16: 47–55. rodents to the calls of an owl: a comparative study. J. Zool. 245:439–446. Aplin, L. M., Farine, D. R., Morand-Ferron, J., et al. 2013. Individual person- Hitchcock, A. C., T. Chen, E. Connolly, K. Darakananda, J. Jeong, A. Quist, alities predict social behaviour in wild networks of great tits (Parus major). A. Robbins, and D. J. Ellerby. 2015. Trade-offs between performance and Ecol. Lett. 16: 1365–1372. variability in the escape responses of bluegill sunfish (Lepomis macrochi- Baba, Y., A. Tsukada, and C. M. Comer. 2010. Collision avoidance by running rus). Biol. Open. 4: 743–751. insects: antennal guidance in cockroaches. J. Exp. Biol. 213: 2294–302. Jeanson, R., S.  Blanco, R.  Fournier, J.  Deneubourg, V.  Fourcassié, and Bell, W. J., and K. G. Adiyodi. 1982. The American cockroach. Chapman & G. Theraulaz. 2003. A model of animal movements in a bounded space. J. Hall, London, UK. Theor. Biol. 225: 443–451. Bell, W. J., L. M. Roth, and C. A. Nalepa. 2007. Cockroaches: ecology, behav- Jeanson, R., C. Rivault, J.-L. Deneubourg, S. Blanco, R. Fournier, C. Jost, and iour, and natural history. The Johns Hopkins University Press, Baltimore, G.  Theraulaz. 2005. Self-organized aggregation in cockroaches. Anim. MA. Behav. 69: 169–180. Bell, A. M., S. J. Hankison, and K. L. Laskowski. 2009. The repeatability of Jolles, J. W., N. J.  Boogert, V. H.  Sridhar, et  al. 2017. Consistent individual behaviour: a meta-analysis. Anim. Behav. 77: 771–783. differences drive collective behavior and group functioning of schooling Boulay, J., C. Devigne, D. Gosset, and D. Charabidze. 2013. Evidence of active fish. Curr. Biol. 27:2862–2868.e7. aggregation behaviour in Lucilia sericata larvae and possible implication Kallai, J., T.  Makany, A.  Csatho, K.  Karadi, D.  Horvath, B.  Kovacs-Labadi, of a conspecific mark. Anim. Behav. 85: 1191–1197. R. Jarai, L. Nadel, and J. W. Jacobs. 2007. Cognitive and affective aspects Brown, C., and E. Irving. 2014. Individual personality traits influence group of thigmotaxis strategy in humans. Behav. Neurosci. 121: 21–30. exploration in a feral guppy population. Behav. Ecol. 25:95–101. Kendall, M. G. 1938. A new measure of rank correlation. Biometrika 30: Calandre, T., O. Bénichou, and R. Voituriez. 2014. Accelerating search kinetics 81–93. by following boundaries. Phys. Rev. Lett. 112: 26–29. Kendall, M. G., and B. B.  Smith. 1939. The problem of m rankings. Ann. Camhi, J. M., and E. N. Johnson. 1999. High-frequency steering maneuvers Math. Stat. 10: 275–287. mediated by tactile cues: antennal wall-following in the cockroach. J. Exp. Kerth, G. 2010. Group decision-making in fission-fusion societies. Behav Biol. 202: 631–643. Processes 84: 662–663. Canonge, S., J.-L. Deneubourg, and G. Sempo. 2011. Group living enhances Laurent Salazar, M., J.  Deneubourg, and G.  Sempo. 2013. Information cas- individual resources discrimination: the use of public information by cock- cade ruling the fleeing behaviour of a gregarious insect. Anim. Behav. 85: roaches to assess shelter quality. PLoS One 6: e19748. 1271–1285. Carlson, B. E., and T. Langkilde. 2013. Personality traits are expressed in bull- Laurent Salazar, M.-O., I. Planas-Sitjà, J.-L. Deneubourg, and G. Sempo. 2015. frog tadpoles during open-field trials. J. Herpetol. 47: 378–383. Collective resilience in a disturbed environment: stability of the activity Carrete, M., and J. L.  Tella. 2010. Individual consistency in flight initiation rhythm and group personality in Periplaneta americana. Behav. Ecol. distances in burrowing owls: a new hypothesis on disturbance-induced Sociobiol 69: 1879. habitat selection. Biol. Lett. 6: 167–70. Marras, S., S. S. Killen, G. Claireaux, P. Domenici, and D. J. McKenzie. 2011. Carter, A. J., A. W. Goldizen, and S. A. Tromp. 2010. Agamas exhibit behav- Behavioural and kinematic components of the fast-start escape response ioral syndromes: bolder males bask and feed more but may suffer higher in fish: individual variation and temporal repeatability. J. Exp. Biol. 214: predation. Behav. Ecol. 21: 655–661. 3102–3110. Cooper Jr., W. E., and Blumstein, D. T. 2015. Escaping from predators: an inte- McDermott, D. R., M. J. Chips, M. McGuirk, F. Armagost, N. DiRienzo, and J. grative view of escape decisions. Cambridge University Press, UK. N. Pruitt. 2014. Boldness is influenced by sublethal interactions with pred- Cote, J., S.  Fogarty, B.  Tymen, et  al. 2013. Personality-dependent dispersal ators and is associated with successful harem infiltration in Madagascar cancelled under predation risk. Proc. R. Soc. B. Biol. Sci. 280: 20132349. hissing cockroaches. Behav. Ecol. Sociobiol. 68:425–435. Cronin, A. L. 2015. Individual and group personalities characterise consensus Mote, M. I., and T. H. Goldsmith. 1970. Spectral sensitivities of color recep- decision-making in an ant. Ethology 121: 703–713. tors in the compound eye of the cockroach Periplaneta. J. Exp. Zool. 173: Dall, S. R.  X., A. I.  Houston, and J. M.  McNamara. 2004. The behavioural 137–45. ecology of personality: consistent individual differences from an adaptive Motulsky, H. 1999. Analyzing data with GraphPad Prism. GraphPad Software perspective. Ecol. Lett. 7: 734–739. Inc., San Diego, CA, www.graphpad.com. Devigne, C., P. Broly, and J. L. Deneubourg. 2011. Individual preferences and Niemelä, P. T., N.  DiRienzo, and A. V.  Hedrick. 2012. Predator-induced social interactions determine the aggregation of woodlice. PLoS One 6(2): changes in the boldness of naïve field crickets, Gryllus integer, depends on e17389. behavioural type. Anim. Behav. 84: 129–135. Dingemanse, N. J., A. J. N. Kazem, D. Réale, J. Wright. 2010. Behavioural re- Okada, J., and Y. Toh. 1998. Shade response in the escape behaviour of the action norms: animal personality meets individual plasticity. Trends Ecol. cockroach, Periplaneta americana. Zool. Sci. 15: 831–835. Evol. 25:81–89. Okada, J., and Y. Toh. 2006. Active tactile sensing for localization of objects Dingemanse, N. J., and M. Wolf. 2010. Recent models for adaptive personality by the cockroach antenna. J. Comp. Physiol. A. 192: 715–726. differences: a review. Philos. Trans. R. Soc. B. 365: 3947–3958. Paterson, Z. A., and R. J. Weaver. 1997. Characterization and temporal aspects Domenici, P., D. Booth, J. M. Blagburn, and J. P. Bacon. 2008. Cockroaches of haemolymph juvenile hormone esterase in adult cockroach, Periplaneta keep predators guessing by using preferred escape trajectories. Curr. Biol. americana. J. Insect Physiol. 43: 521–532. 18: 1792–1796. Planas-Sitjà, I., J.-L. Deneubourg, C. Gibon, and G. Sempo. 2015. Group per- Durier, V., and C. Rivault. 2003. Exploitation of home range and spatial dis- sonality during collective decision-making: a multi-level approach. Proc. tribution of resources in German cockroaches (Dictyoptera: Blattellidae). R. Soc. B. 282: 20142515. J. Econ. Entomol. 96: 1832–1837. Réale, D., D.  Garant, M. M.  Humphries, P.  Bergeron, V.  Careau, and P.- Dussutour, A., J.-L. Deneubourg, and V. Fourcassié. 2005. Amplification of in- O. Montiglio. 2010. Personality and the emergence of the pace-of-life syn- dividual preferences in a social context: the case of wall-following in ants. drome concept at the population level. Philos. Trans. R. Soc. Lond. B. Biol. Proc. R. Soc. B. 272: 705–714. Grossen, N. E., and M. J.  Kelley. 1972. Species-specific behaviour and ac- Sci. 365: 4051–4063. quisition of avoidance behaviour in rats. J. Comp. Physiol. Psychol. 81: Seltmann, M. W., M.  Öst, K.  Jaatinen, S.  Atkinson, K.  Mashburn, and 307–310. T.  Hollmén. 2012. Stress responsiveness, age and body condition inter- Halloy, J., G. Sempo, G. Caprari, C. Rivault, M. Asadpour, F. Tâche, I. Saïd, actively affect flight initiation distance in breeding female eiders. Anim. V. Durier, S. Canonge, J. M. Amé, et al. 2007. Social integration of robots Behav. 84: 889–896. Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Journal of Insect Science, 2018, Vol. 18, No. 1 7 Sharma, S., S.  Coombs, P.  Patton, and T. B.  De Perera. 2009. The function Stanley, C. R., C.  Mettke-Hofmann, R. F.  Preziosi. 2017. Personality in the of wall-following behaviors in the Mexican blind cavefish and a sighted cockroach Diploptera punctata: evidence for stability across develop- relative, the Mexican tetra (Astyanax). J. Comp. Physiol. A  Neuroethol. mental stages despite age effects on boldness. PLoS One 12:1–23. Sensory, Neural, Behav. Physiol. 195: 225–240. Webster, M. M., and A. J. W. Ward. 2011. Personality and social context. Biol. Sih, A., A. Bell, J. C. Johnson. 2004. Behavioral syndromes: an ecological and Rev. 86: 759–773. evolutionary overview. Trends Ecol. Evol. 19:372–378. Webster, M. M., and K. N.  Laland. 2015. Space-use and sociability are not Sih, A., A. M. Bell, J. C. Johnson, R. E. Ziemba. 2004. Behavioral syndromes: related to public-information use in ninespine sticklebacks. Behav. Ecol. an integrative overview. Q. Rev. Biol. 79:241–277. Sociobiol. 69: 895–907. Simon, P., R. Dupuis, and J. Costentin. 1994. Thigmotaxis as an index of anx- Wright, C. M., J. L. L. Lichtenstein, G. A. Montgomery, et al. 2017. Exposure iety in mice. Influence of dopaminergic transmissions. Behav. Brain Res. to predators reduces collective foraging aggressiveness and eliminates its 61: 59–64. relationship with colony personality composition. Behav Ecol Sociobiol Smith, B. R., and D. T. Blumstein. 2010. Behavioral types as predictors of sur- 71: 126. vival in Trinidadian guppies (Poecilia reticulata). Behav. Ecol. 21: 919–926. Zar, J. H. 2010. Biostatistical analysis. Prentice Hall, Upper Saddle River, NJ, USA. Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Journal of Insect Science Oxford University Press

Individual Thigmotactic Preference Affects the Fleeing Behavior of the American Cockroach (Blattodea: Blattidae)

Free
7 pages

Loading next page...
 
/lp/ou_press/individual-thigmotactic-preference-affects-the-fleeing-behavior-of-the-bASpd1ta09
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of Entomological Society of America.
eISSN
1536-2442
D.O.I.
10.1093/jisesa/iex108
Publisher site
See Article on Publisher Site

Abstract

Positive thigmotactic behavior is associated with the ability to hide from predators and is important to explain aggregation and collective patterns in various animals. For example, this behavior has been observed in woodlice, domiciliary cockroaches, ants, and fish. Lately, research on different species is focused on the importance of animal personality for ecological and evolutionary processes, individual fitness and group cohesion. In fact, it is generally expected to find some degree of interindividual consistent differences for a behavior, unless specific circumstances, like predator attacks, hide the presence of personalities. In this research, we analyzed the individual thigmotactic preference of domiciliary cockroaches (Periplaneta americana (Linnaeus, 1758) (Blattodea: Blattidae)) and how it affected the fleeing behavior of isolated individuals inside a shelter after receiving a light stimulus. We notably highlight how isolated individuals show different consistent preferences regarding their position in the shelter, which is due to the individual thigmotaxis level, before the fleeing behavior. During the fleeing itself, cockroaches nearer to the wall, and therefore with more positive thigmotaxis, showed slower reaction lantencies to the stimulus. We propose that thigmotaxis homogenizes the interindividual differences among individuals and is important to explain the individual and collective fleeing behavior. Key words: thigmotaxis, disturbance, fleeing behavior, animal personality, domiciliary cockroach Landscape heterogeneities can affect the spatial distribution of is a kinetically efficient behavior, as has been shown by theoretical organisms by influencing their movement patterns. For instance, research (Calandre et al. 2014). habitat edges or patch boundaries can help the aggregation or Recently, several authors have studied thigmotaxis in the con- modify the movement patterns of animals with positive thigmotaxis text of individual behavioral differences (Carlson and Langkilde (Jeanson et al. 2003). Positive thigmotactic behavior, or the tendency 2013, Webster and Laland 2015). Behavioral variability and plasti- to remain close to walls or edges, is widespread in different taxa city have evolved, and are maintained, through numerous processes (Grossen and Kelley 1972, Kallai et al. 2007, Sharma et al. 2009) and that ultimately allow animal populations to adapt to different en- can be modulated by several factors. For example, positive thigmo- vironmental conditions (Dall et  al. 2004, Sih et  al. 2004, Carrete taxis decreases in a familiar environment and during exploration, and Tella 2010, Dingemanse et al. 2010, Smith and Blumstein 2010, and increases under stressful situations (Durier and Rivault 2003, Adriaenssens and Johnsson 2013). In this study, we define person- Sharma et al. 2009). In addition, positive thigmotactic behavior is a ality as the presence of behavioral differences between individuals crucial factor to explain and understand the aggregation and move- that are consistent over time for a specific behavior (Dall et al. 2004, ment of several species such as woodlice, domiciliary cockroaches, Bell et  al. 2009, Dingemanse and Wolf 2010, Réale et  al. 2010, ants, or fish (Dussutour et al. 2005, Bell et al. 2007, Devigne et al. Webster and Ward 2011, Adriaenssens and Johnsson 2013, Planas- 2011, Boulay et  al. 2013, Webster and Laland 2015) and can be Sitjà et al. 2015). The presence of these personalities is widespread used as a proxy for behavioral traits such as boldness or shelter use and raises the complexity of group dynamics. For example, the (Carlson and Langkilde 2013, Webster and Laland 2015). One pos- fission–fusion dynamics of groups of bats (Kerth 2010) and birds sible reason why thigmotaxis has been selected for so many species (Aplin et al. 2013) is influenced by the personality of each individual; is that it accelerates a search process in certain conditions, and thus the aggregation dynamics of Periplaneta americana (Linnaeus, © The Author(s) 2018. Published by Oxford University Press on behalf of Entomological Society of America. This is an Open Access article distributed under the terms of the Creative Commons Attribution Non-Commercial License (http://creativecommons.org/ licenses/by-nc/4.0/), which permits non-commercial re-use, distribution, and reproduction in any medium, provided the original work is properly cited. For commercial re-use, please contact journals.permissions@oup.com Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 2 Journal of Insect Science, 2018, Vol. 18, No. 1 1758)  (Blattodea: Blattidae) is influenced by individual and group et  al, 2013). The floor of the arena was covered with white paper personalities (Planas-Sitjà et al. 2015); and individual personality in (120 g/m ), which was changed after each experiment to prevent any stickleback fish strongly affected the group’s structure, movement chemical marking. dynamics, and foraging behavior (Jolles et al. 2017). Regarding flee- A plastic ring (interior diameter: 25 cm, height: 4.5 cm, width: ing behavior, different studies have found the existence of behavioral 5mm), central within the arena, constituted the shelter. Two open- syndromes (Carter et al. 2010, Niemelä et al. 2012, Seltmann et al. ings of 3 × 1.5 cm placed symmetrically opposite to each other were 2012, Cooper Jr and Blumstein 2015) and that some events, like an the only ways out of the shelter. A glass cover was placed on top of increase of risk or stress, might break apart individual personalities the ring to allow the observation of the interior. (Niemelä et al. 2012, Cote et al. 2013, Wright et al. 2017). A light bulb (Philips [Amsterdam, The Netherlands] A55 FR, 100 It is in this context, using a gregarious species, that we performed W) centered on the arena at a height of 50 cm above the glass was the current study aiming to analyze the presence, or absence, of person- switched on at the beginning of each experiment. When the light was ality in the context of thigmotactic behavior, and its effects during flee- turned on, only the inside of the shelter received ~1,700 lx. (luxmeter ing, of an isolated individual. Our biological model was the negatively Testo 545, Testo NV/SA [Lenzkirch, Germany], resolution: 1 lx). The phototactic American cockroach (P.  americana), which forms aggre- rest of the arena, and of the room, was only illuminated by red light gations in dark, warm, and damp places during daylight hours (Bell (0 lx). This arrangement allowed us both to see and to keep the room and Adiyodi 1982, Canonge et al. 2011). P. americana, when suddenly ‘dark’ for the cockroaches (Mote and Goldsmith 1970). illuminated in their resting place, initiate a fleeing response (Okada and Toh 1998, Domenici et al. 2008). Domiciliary cockroaches have a Experimental Procedure very strong tendency to follow a wall, even while fleeing (Camhi and Only adult males (without external damages such as wing damage Johnson 1999, Durier and Rivault 2003). Consequently, knowing that or missing leg segments) were tested to exclude any behavioral var- domiciliary cockroaches express different personalities during their iations related to the ovarian cycle (Paterson and Weaver 1997). sheltering process (Planas-Sitjà et al. 2015), we used their light avoid- Isolated individuals (24 replicates) were randomly taken from the ance behavior to investigate the relationship between thigmotaxis and rearing room and placed the day before the trial in plastic containers the possible behavioral consistency before and while fleeing. (36 × 24 and 14 cm in height). They were kept in the dark until the beginning of their respective trials. Each cockroach was tested three times in total, with a 2-d interval between tests. The cockroaches Materials and Methods had access to dog pellets (Tom & Co., Delhaize Group, Brussels, Rearing Conditions Belgium) and to a piece of cotton soaked in water. The cockroaches were taken from the rearing room of the Université Before the start of each trial, the arena was cleaned with dena- libre de Bruxelles (ULB). P.  americana has been reared in the ULB tured ethanol, and a new sheet of paper was placed on the floor. since 2002 in five Plexiglas vivaria (80 × 40 cm and 100 cm high) in After 10 min, the individual was introduced under light CO narcosis which cardboard tubes are hung from the walls to serve as shelters. into the shelter (the openings were closed), as is standard procedure Groups of around 20 individuals were weekly exchanged between (Jeanson et al. 2003, 2005; Halloy et al. 2007), and remained there vivaria. Each vivarium had around 1,000 individuals, comprising for 30 min. After this waiting period, the openings were opened, and males, females, and larvae from all developmental stages. They were the light was turned on. A  webcam was used to record the cock- provided with dog pellets and water ad libitum. The rearing room roach’s behavior inside the shelter (20 frames/s) and recorded the was kept at 26 ± 1°C with a photoperiod of 12:12 (L:D) h. individual’s response for 5 min or until the cockroach had left the shelter. As noted above, during the night, P. americana are active and Experimental Set-Up leave their shelter to forage (Bell and Adiyodi 1982). For this reason, Our study was carried out using the experimental protocol described all trials were performed during daylight hours corresponding to the in detail in the study by Laurent Salazar et al. (2013). The experi- resting phase of P. americana. ments were performed in a circular arena (Fig. 1) that had an elec- tric fence to prevent the cockroaches from escaping (Laurent Salazar Analysis From the video recordings, we analyzed the following parameters: (1) The position (radius from the center of the shelter), orientation, and contact with the wall of each individual prior to the turning on of the light (t ). To analyze the orientation of the individuals, we calculated the angle between the head–abdomen axis of the cockroach and the axis going from the shelter wall to the center of the arena and passing through the middle of the cockroach. If the angle was comprised between +45° and −45°, we consid- ered that the individual was facing toward the shelter wall (Supp material [online only]). (2) The reaction time (RT) for each individual was the time inter- val between the turning on of the light (t ) and the initiation of its fleeing behavior. We considered an individual to be reacting when it rotated its body or moved forward/backward. When an individual was walking prior to t (0.32 of all trials), it was removed from the analyses. We consider that the immobile/ Fig.  1. Experimental set-up. Individual cockroaches were placed inside the mobile states are different because mobile individuals (already plastic ring. Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Journal of Insect Science, 2018, Vol. 18, No. 1 3 walking) were more prompt to flee than the immobile ones when the wall of the shelter prior to t did not significantly differ between light was turned on. days (Table 1). When we compared the number of times each indi- vidual was moving, and the number of times it was facing toward We used survival curves (or decay curves) to analyze the distribution the shelter wall, against a theoretical random distribution, we found of these parameters. Although the name includes the term ‘survival’, no significant difference (Table 2). However, our results showed that it can be used to analyze any kind of experiment where the results are individuals significantly differed from each other and presented dis- expressed as a time to an end point. For the comparison of survival tinct preferences regarding their position respective to the center of curves, we used a log-rank test (Motulsky 1999)). For the analysis of the shelter (Figs 2a and 3; Table 2). curve fittings and linear regressions, we used Kolmogorov–Smirnov This personality regarding their position has interesting conse- test and F-test, respectively. quences for their thigmotactic preferences. Since individuals touch- To compare distributions, we used chi-square test when the condi- ing the wall did so with their antennae, it was their length that tions allowed it, otherwise we used two-sample Kolmogorov–Smirnov limited the distance where touching was possible. Indeed, the proba- tests. Since our values did not meet the conditions for parametric bility of touching the wall follows a logistic function and shows that tests, we used Kruskal–Wallis and Mann–Whitney tests to compare individuals placed more than 4.6 cm from the wall (7.9 cm from the differences between days and individuals, depending if more than one center) never touch the wall, while those closer than 4.6 cm do comparison was being carried out or if the values were paired. We used the Kendall’s coefficient of concordance (W ) for concord- 1 Touch= ,. R = 08, 10.7(r − adius) ance assessment (Kendall 1938; Kendall and Smith 1939). Kendall’s 1 + e coefficient of concordance (W ) compares the stability of rank posi- two-sample Kolmogorov–Smirnov test: D = 0.18, P = 0.17; Fig. 4). The tions for each group during the trials. The values of W range from high steepness of the fitting (coefficient value = 10.7) clearly shows how 0 (no concordance of ranks) to 1 (complete concordance). We com- touching or not touching the wall is close to an all-or-none response. pared the observed W coefficients with the ‘Kendall random distribu- In other words, all individuals within reach of the wall, will touch it. tion’ (KRD) as explained by Planas-Sitjà et al. (2015). The KRD is the This allowed us to divide the individuals into two groups regarding theoretical distribution of the W coefficients for random rank orders their thigmotactic preference: high thigmotactic level (<4.6 cm from the of the same number of experimental groups and repetitions (e.g., 24 wall) and low thigmotactic level individuals, HTL and LTL, respectively. groups and 3 repetitions and N = 1,000). We performed a Z-test to Across 3 d, we observed 28 LTL individuals and 44 HTL individuals. test the significance of the difference between the observed W coeffi- In addition, we analyzed the influence of the thigmotactic level on cients and the corresponding KRD (Zar 2010, Laurent Salazar et al. cockroach orientation. In the case of the LTL individuals (N = 5/28 2015, Planas-Sitjà et al. 2015). In this case, we assigned the value of were facing toward the wall), there was no significant difference 0 s to the RT of immobile individuals. This was necessary to be able between our observations and theoretical random expectation of to quantify personality traits and allowed us to consider the idea of 0.25 (one-tailed binomial test: P = 0.26). However, the HTL individ- individuals being more prompt to react (with an RT value of 0) when uals (N = 17/44) had a significant tendency of orienting themselves analyzing their reaction personality (e.g., see whether an individual toward the shelter wall (Binomial test: P = 0.032). was always ready to flee, already moving or slow to react). To analyze the influence of the contact with the wall on cock- roach orientation, we used a one-tail binomial test to compare their During the Disturbance observed orientation and a theoretical random expectation of facing The relationship between the thigmotactic level and RT among the toward the wall (P = 0.25). immobile individuals (22 distinct individuals were immobile; during the 3 d, we observed 49 immobile individuals) was clear. Indeed, immobile HTL individuals (31 individuals) differed in their RT from Results the LTL (18 individuals) ones (Mann–Whitney test (mean ± SD): Before the Disturbance RT HTL (27.4 s ± 38) and RT LTL (7s ± 8.2): U = 165, P = 0.024). The distribution of the position of the individuals within the shelter, These results indicate clearly that the thigmotactic level of indi- the number of moving individuals, and the number facing toward viduals during rest has an influence on their fleeing behavior: HTL Table 1. Results of the interday behavioral comparison Day I Day II Day III Comparison between days Spatial distribution Homogenous Chi- Homogenous Chi-square: Homogenous Chi- No difference Chi-square: 2 2 2 2 square: X  = 1.69 X  = 4.91, P = 0.18 square: X  = 0.29, X  = 8.2, P = 0.22 3 , 3 3 6 P = 0.64 P = 0.96 Individuals facing toward the shelter wall 7 8 7 No difference Chi-square: X  = 0.13, P = 0.94 Number of moving individuals 8 6 9 No difference Chi-square: X  = 0.89, P = 0.64 RT:/ λ = 1 RT Exponential distri- Exponential distribution Exponential distri- No difference X  = 1, log- 2 2 2 λ = 0.08 ± 0.02 per s bution R  = 0.83; R  = 0.96; Kolmogorov– bution R  = 0.92; rank test: P = 0.60 Kolmogorov–Smirnov Smirnov test: D = 0.11, Kolmogorov–Smirnov test: D = 0.18, P > 0.99 test: D = 0.13, P > 0.90 P > 0.99 The table shows the daily result for each measure and their comparison between days. Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 4 Journal of Insect Science, 2018, Vol. 18, No. 1 Table 2. Summary of the interindividual differences Presence of Consistent interindi- interindividual difference? vidual difference? Spatial distribution Yes Kruskal–Wallis test: Yes Kendall’s H  = 41.98, P = 0.009 W = 0.58, Z = 3.15, P = 0.0016 Individuals facing No chi-square: X  = 2.6, toward the shelter P = 0.27 wall Number of moving No chi-square: X  = 1.83, individuals P = 0.61 RT No Kruskal–Wallis test: No Kendall’s H  = 20.97, P = 0.34 W = 0.39, Z = 0.7, P = 0.48 The table shows the presence of interindividual difference and its consist- ency for each measure. Only the spatial distribution of individuals showed interindividual difference, which was consistent across days. individuals were four times slower to react than LTL individuals (Fig.  3). Nevertheless, we found no consistency in RT within indi- viduals: individuals’ RT rankings varied each day (Fig. 2b; Table 2). Discussion Fig. 2. (a) Average radius (+SD) (the distance from the center of the shelter to It is well known that numerous species, including domiciliary cock- the cockroach’s body, 0 = center of the shelter). (b) Average RT (+SD) of each roaches like the American cockroach, show positive thigmotaxis, immobile individual. and that this thigmotaxis affects the exploration of the environment as well as the individual and collective behavior (Camhi and Johnson the menace (e.g., kind of predator) or foraging strategy (Hendrie 1999, Jeanson et al. 2003, Okada and Toh 2006, Kallai et al. 2007, et al. 1998). Baba et  al. 2010). In this study, we show that prior to a light dis- The RTs across days were not consistent between individuals, in turbance, individuals of P.  americana displayed clear personalities contradiction with previous studies (McDermott et al. 2014; Stanley regarding their thigmotactic preference (being in range to touch the et al. 2017) showing consistency in the reaction to a disturbance. It wall, <4.6  cm from the wall, or not) and thus demonstrating that is possible that individuals had different reaction threshold to light thigmotaxis is a quantifiable personality trait in P. americana. disturbances, but the magnitude of the disturbance in our study In addition to this, we observed that high thigmotactic level indi- was such that it overshadowed these differences. Further experi- viduals tended to orient themselves toward the wall. It is possible ments with different light intensities could provide insights on this that inside a shelter an individual cockroach will orient itself auto- matically toward a wall when near to it. However, we are inclined toward the hypothesis that this orientation increases their thigmot- actic level when in their resting phase. Indeed, in experiments where cockroaches explore a bigger arena, they position themselves par- allel to the arena wall, even when touching it (M-O Laurent, Isaac Planas-Sitjà, personal observation). This is different from our current observations. It is possible that the wall has a different influence on a cockroach’s behavior depending on the context. Certainly, in an open and novel space, stressed individuals will tend to be parallel to a wall, move slower, and follow it for protection (Durier and Rivault 2003, Sharma et al. 2009). However, inside a shelter, individuals are arguably less stressed and do not need to position themselves parallel and against the wall. We clearly show how, for immobile individuals, being in contact with the wall (i.e., high individual thigmotactic preference) had an impact on their RTs, with individuals touching the wall being signif- icantly slower to react. It is possible that such individual differences regarding their position could be the result of different survival strat- egies. An individual near a wall that reacts slower would be safer by being more cryptic, and running along the wall later to find shelter if needed (Durier and Rivault 2003, Sharma et al. 2009), while an individual further away from the wall would flee faster and find shel- Fig.  3. Boxplot comparing the RT to a light stimulus between HTL and LTL individuals. ter. These different strategies would be advantageous depending on Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Journal of Insect Science, 2018, Vol. 18, No. 1 5 Fig. 4. Contact with the wall of the shelter before the disturbance as a function of the position of each individual (1 = touching, 0 = not touching the wall). The position of each individual is represented as the distance from the center of the shelter to the cockroach’s body (i.e., radius). hypothesis. In addition, the lack of consistency in our study could were able to quantify different traits for individuals without the be due to the thigmotactic behavior, which was missing in the men- collective context having a potential impact. The quantification of tioned previous works. We have already seen that thigmotaxis can personality on individual thigmotaxis level shed some light on the be influenced by the environment, for example, in domiciliary cock- proximal mechanisms that affect the presence of both individual roaches the presence of a wall modifies their exploratory behavior, and collective personality shown in previous studies on collect- reducing their speed and increasing their number of stops (Durier and ive dynamics under different conditions in P.  americana (Laurent Rivault 2003). Furthermore, it has been shown that being in novel Salazar et al. 2015, Planas-Sitjà et al. 2015). As positive thigmotaxis environments increases the tendency of being positively thigmotactic has been shown to have evolved in many species and to be impor- (Simon et al. 1994, Durier and Rivault 2003, Sharma et al. 2009). tant for their behavior and survival (Grossen and Kelley 1972, Considering these results, we hypothesize that the lack of consist- Kallai et al. 2007, Sharma et al. 2009), we believe that the existence ency regarding the RT can be explained by the fact that thigmotaxis of individual thigmotactic preferences could be extensible to many reduces and homogenizes the interindividual differences between of these species. How these preferences are maintained within a individuals. Indeed, if individuals slow their exploration near walls, group and how they affect the collective behavior are key questions and novel environments increase their preference to be near a wall, for the understanding of the behavioral ecology of gregarious spe- then it is possible that in our experiments (new environment) indi- cies. The presence of personalities within a group has been seen to viduals increased their tendency to be close to the wall. Thus, indi- have an important impact on the performance of the group during viduals that would otherwise display different fleeing thresholds will collective behaviors (Brown and Irving 2014, Cronin 2015). There be positioned close to the wall and show slower RTs when the light is is no doubt that such individual variability also influences collect- turned on, homogenizing the RT. Experiments with a higher framing ive fleeing behavior. Further studies on gregarious insects and their rate (higher than 20 fps) could shed some light on this matter, being individual preferences are indeed needed to acquire a complete view precise enough to detect interindividual differences in RTs despite of their collective behavior and elucidate how these personalities the homogenization caused by the thigmotactic behavior. could be an evolutionary benefit for the collective fleeing response Regarding the global positioning, we observed that the distri- to disturbances. For instance, studies on the positional preferences bution within the shelter prior to the disturbance was homogene- would allow us to create nonrandom groups consisting of individu- ous. In a prior study, Laurent Salazar et al. (2013, Supp Material als with personality regarding their positioning and their responses [online only]) showed that isolated and pairs of individuals were to the stimulus. In such a way, it would be interesting to not only also homogeneously positioned within a shelter, however individ- link these personality traits with collective fleeing dynamics but also uals in larger groups had a distribution significantly different from to other behaviors in different situations, such as collective shelter a theoretical homogeneous distribution, with an increase in the choice, aggregate stability, or food search. Indeed, studies testing number of individuals close to the shelter wall compared with the such groups could help elucidate the influence of group composition theoretical distribution. This observation, albeit in need of further in collective dynamics, especially in research aiming at understand- research, already gives us a glimpse of the difficulties in studying ing the link between individual preferences, individual preference individual preferences in a collective context. By being numerous, amplification, and collective behavior. due to the gregarious behavior, it is possible that their individual preferences for the wall were amplified (Dussutour et  al. 2005), Supplementary Data leading to the heterogeneous distribution in favor to being closer to the wall. Supplementary data are available at Journal of Insect Science online. Previous studies on fleeing responses have been able to concen- trate on the consistency of many behavioral and kinematic traits Acknowledgments (e.g., in vertebrates (Marras et al. 2011; Hitchcock et al. 2015) and invertebrates (Stanley et  al. 2017)). However, to our knowledge, M-OLS and IP-S were funded by a PhD grant from FRIA (Fonds pour la there is an absence of works concerned about personality traits dur- Recherche dans l’Industrie et dans l’Agriculture). JLD is Research Director from the Belgian National Fund for Scientific Research (F.N.R.S). ing fleeing and kinematic traits in insects. In our present study, we Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 6 Journal of Insect Science, 2018, Vol. 18, No. 1 into groups of cockroaches to control self-organized choices. Science. 318: References Cited 1155–1158. Adriaenssens, B., and J. I. Johnsson. 2013. Natural selection, plasticity and the Hendrie, C. A., S. M.  Weiss, D.  Eilam. 1998. Behavioural response of wild emergence of a behavioural syndrome in the wild. Ecol. Lett. 16: 47–55. rodents to the calls of an owl: a comparative study. J. Zool. 245:439–446. Aplin, L. M., Farine, D. R., Morand-Ferron, J., et al. 2013. Individual person- Hitchcock, A. C., T. Chen, E. Connolly, K. Darakananda, J. Jeong, A. Quist, alities predict social behaviour in wild networks of great tits (Parus major). A. Robbins, and D. J. Ellerby. 2015. Trade-offs between performance and Ecol. Lett. 16: 1365–1372. variability in the escape responses of bluegill sunfish (Lepomis macrochi- Baba, Y., A. Tsukada, and C. M. Comer. 2010. Collision avoidance by running rus). Biol. Open. 4: 743–751. insects: antennal guidance in cockroaches. J. Exp. Biol. 213: 2294–302. Jeanson, R., S.  Blanco, R.  Fournier, J.  Deneubourg, V.  Fourcassié, and Bell, W. J., and K. G. Adiyodi. 1982. The American cockroach. Chapman & G. Theraulaz. 2003. A model of animal movements in a bounded space. J. Hall, London, UK. Theor. Biol. 225: 443–451. Bell, W. J., L. M. Roth, and C. A. Nalepa. 2007. Cockroaches: ecology, behav- Jeanson, R., C. Rivault, J.-L. Deneubourg, S. Blanco, R. Fournier, C. Jost, and iour, and natural history. The Johns Hopkins University Press, Baltimore, G.  Theraulaz. 2005. Self-organized aggregation in cockroaches. Anim. MA. Behav. 69: 169–180. Bell, A. M., S. J. Hankison, and K. L. Laskowski. 2009. The repeatability of Jolles, J. W., N. J.  Boogert, V. H.  Sridhar, et  al. 2017. Consistent individual behaviour: a meta-analysis. Anim. Behav. 77: 771–783. differences drive collective behavior and group functioning of schooling Boulay, J., C. Devigne, D. Gosset, and D. Charabidze. 2013. Evidence of active fish. Curr. Biol. 27:2862–2868.e7. aggregation behaviour in Lucilia sericata larvae and possible implication Kallai, J., T.  Makany, A.  Csatho, K.  Karadi, D.  Horvath, B.  Kovacs-Labadi, of a conspecific mark. Anim. Behav. 85: 1191–1197. R. Jarai, L. Nadel, and J. W. Jacobs. 2007. Cognitive and affective aspects Brown, C., and E. Irving. 2014. Individual personality traits influence group of thigmotaxis strategy in humans. Behav. Neurosci. 121: 21–30. exploration in a feral guppy population. Behav. Ecol. 25:95–101. Kendall, M. G. 1938. A new measure of rank correlation. Biometrika 30: Calandre, T., O. Bénichou, and R. Voituriez. 2014. Accelerating search kinetics 81–93. by following boundaries. Phys. Rev. Lett. 112: 26–29. Kendall, M. G., and B. B.  Smith. 1939. The problem of m rankings. Ann. Camhi, J. M., and E. N. Johnson. 1999. High-frequency steering maneuvers Math. Stat. 10: 275–287. mediated by tactile cues: antennal wall-following in the cockroach. J. Exp. Kerth, G. 2010. Group decision-making in fission-fusion societies. Behav Biol. 202: 631–643. Processes 84: 662–663. Canonge, S., J.-L. Deneubourg, and G. Sempo. 2011. Group living enhances Laurent Salazar, M., J.  Deneubourg, and G.  Sempo. 2013. Information cas- individual resources discrimination: the use of public information by cock- cade ruling the fleeing behaviour of a gregarious insect. Anim. Behav. 85: roaches to assess shelter quality. PLoS One 6: e19748. 1271–1285. Carlson, B. E., and T. Langkilde. 2013. Personality traits are expressed in bull- Laurent Salazar, M.-O., I. Planas-Sitjà, J.-L. Deneubourg, and G. Sempo. 2015. frog tadpoles during open-field trials. J. Herpetol. 47: 378–383. Collective resilience in a disturbed environment: stability of the activity Carrete, M., and J. L.  Tella. 2010. Individual consistency in flight initiation rhythm and group personality in Periplaneta americana. Behav. Ecol. distances in burrowing owls: a new hypothesis on disturbance-induced Sociobiol 69: 1879. habitat selection. Biol. Lett. 6: 167–70. Marras, S., S. S. Killen, G. Claireaux, P. Domenici, and D. J. McKenzie. 2011. Carter, A. J., A. W. Goldizen, and S. A. Tromp. 2010. Agamas exhibit behav- Behavioural and kinematic components of the fast-start escape response ioral syndromes: bolder males bask and feed more but may suffer higher in fish: individual variation and temporal repeatability. J. Exp. Biol. 214: predation. Behav. Ecol. 21: 655–661. 3102–3110. Cooper Jr., W. E., and Blumstein, D. T. 2015. Escaping from predators: an inte- McDermott, D. R., M. J. Chips, M. McGuirk, F. Armagost, N. DiRienzo, and J. grative view of escape decisions. Cambridge University Press, UK. N. Pruitt. 2014. Boldness is influenced by sublethal interactions with pred- Cote, J., S.  Fogarty, B.  Tymen, et  al. 2013. Personality-dependent dispersal ators and is associated with successful harem infiltration in Madagascar cancelled under predation risk. Proc. R. Soc. B. Biol. Sci. 280: 20132349. hissing cockroaches. Behav. Ecol. Sociobiol. 68:425–435. Cronin, A. L. 2015. Individual and group personalities characterise consensus Mote, M. I., and T. H. Goldsmith. 1970. Spectral sensitivities of color recep- decision-making in an ant. Ethology 121: 703–713. tors in the compound eye of the cockroach Periplaneta. J. Exp. Zool. 173: Dall, S. R.  X., A. I.  Houston, and J. M.  McNamara. 2004. The behavioural 137–45. ecology of personality: consistent individual differences from an adaptive Motulsky, H. 1999. Analyzing data with GraphPad Prism. GraphPad Software perspective. Ecol. Lett. 7: 734–739. Inc., San Diego, CA, www.graphpad.com. Devigne, C., P. Broly, and J. L. Deneubourg. 2011. Individual preferences and Niemelä, P. T., N.  DiRienzo, and A. V.  Hedrick. 2012. Predator-induced social interactions determine the aggregation of woodlice. PLoS One 6(2): changes in the boldness of naïve field crickets, Gryllus integer, depends on e17389. behavioural type. Anim. Behav. 84: 129–135. Dingemanse, N. J., A. J. N. Kazem, D. Réale, J. Wright. 2010. Behavioural re- Okada, J., and Y. Toh. 1998. Shade response in the escape behaviour of the action norms: animal personality meets individual plasticity. Trends Ecol. cockroach, Periplaneta americana. Zool. Sci. 15: 831–835. Evol. 25:81–89. Okada, J., and Y. Toh. 2006. Active tactile sensing for localization of objects Dingemanse, N. J., and M. Wolf. 2010. Recent models for adaptive personality by the cockroach antenna. J. Comp. Physiol. A. 192: 715–726. differences: a review. Philos. Trans. R. Soc. B. 365: 3947–3958. Paterson, Z. A., and R. J. Weaver. 1997. Characterization and temporal aspects Domenici, P., D. Booth, J. M. Blagburn, and J. P. Bacon. 2008. Cockroaches of haemolymph juvenile hormone esterase in adult cockroach, Periplaneta keep predators guessing by using preferred escape trajectories. Curr. Biol. americana. J. Insect Physiol. 43: 521–532. 18: 1792–1796. Planas-Sitjà, I., J.-L. Deneubourg, C. Gibon, and G. Sempo. 2015. Group per- Durier, V., and C. Rivault. 2003. Exploitation of home range and spatial dis- sonality during collective decision-making: a multi-level approach. Proc. tribution of resources in German cockroaches (Dictyoptera: Blattellidae). R. Soc. B. 282: 20142515. J. Econ. Entomol. 96: 1832–1837. Réale, D., D.  Garant, M. M.  Humphries, P.  Bergeron, V.  Careau, and P.- Dussutour, A., J.-L. Deneubourg, and V. Fourcassié. 2005. Amplification of in- O. Montiglio. 2010. Personality and the emergence of the pace-of-life syn- dividual preferences in a social context: the case of wall-following in ants. drome concept at the population level. Philos. Trans. R. Soc. Lond. B. Biol. Proc. R. Soc. B. 272: 705–714. Grossen, N. E., and M. J.  Kelley. 1972. Species-specific behaviour and ac- Sci. 365: 4051–4063. quisition of avoidance behaviour in rats. J. Comp. Physiol. Psychol. 81: Seltmann, M. W., M.  Öst, K.  Jaatinen, S.  Atkinson, K.  Mashburn, and 307–310. T.  Hollmén. 2012. Stress responsiveness, age and body condition inter- Halloy, J., G. Sempo, G. Caprari, C. Rivault, M. Asadpour, F. Tâche, I. Saïd, actively affect flight initiation distance in breeding female eiders. Anim. V. Durier, S. Canonge, J. M. Amé, et al. 2007. Social integration of robots Behav. 84: 889–896. Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018 Journal of Insect Science, 2018, Vol. 18, No. 1 7 Sharma, S., S.  Coombs, P.  Patton, and T. B.  De Perera. 2009. The function Stanley, C. R., C.  Mettke-Hofmann, R. F.  Preziosi. 2017. Personality in the of wall-following behaviors in the Mexican blind cavefish and a sighted cockroach Diploptera punctata: evidence for stability across develop- relative, the Mexican tetra (Astyanax). J. Comp. Physiol. A  Neuroethol. mental stages despite age effects on boldness. PLoS One 12:1–23. Sensory, Neural, Behav. Physiol. 195: 225–240. Webster, M. M., and A. J. W. Ward. 2011. Personality and social context. Biol. Sih, A., A. Bell, J. C. Johnson. 2004. Behavioral syndromes: an ecological and Rev. 86: 759–773. evolutionary overview. Trends Ecol. Evol. 19:372–378. Webster, M. M., and K. N.  Laland. 2015. Space-use and sociability are not Sih, A., A. M. Bell, J. C. Johnson, R. E. Ziemba. 2004. Behavioral syndromes: related to public-information use in ninespine sticklebacks. Behav. Ecol. an integrative overview. Q. Rev. Biol. 79:241–277. Sociobiol. 69: 895–907. Simon, P., R. Dupuis, and J. Costentin. 1994. Thigmotaxis as an index of anx- Wright, C. M., J. L. L. Lichtenstein, G. A. Montgomery, et al. 2017. Exposure iety in mice. Influence of dopaminergic transmissions. Behav. Brain Res. to predators reduces collective foraging aggressiveness and eliminates its 61: 59–64. relationship with colony personality composition. Behav Ecol Sociobiol Smith, B. R., and D. T. Blumstein. 2010. Behavioral types as predictors of sur- 71: 126. vival in Trinidadian guppies (Poecilia reticulata). Behav. Ecol. 21: 919–926. Zar, J. H. 2010. Biostatistical analysis. Prentice Hall, Upper Saddle River, NJ, USA. Downloaded from https://academic.oup.com/jinsectscience/article-abstract/18/1/9/4830252 by Ed 'DeepDyve' Gillespie user on 16 March 2018

Journal

Journal of Insect ScienceOxford University Press

Published: Jan 1, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off