How does starvation affect spatial organization within nests in Lasius niger?

Transcription

1 Insect. Soc. (2011) 58: DOI /s Insectes Sociaux RESEARCH ARTICLE How does starvation affect spatial organization within nests in Lasius niger? A.-C. Mailleux G. Sempo S. Depickère C. Detrain J. L. Deneubourg Received: 27 April 2010 / Revised: 24 November 2010 / Accepted: 6 December 2010 / Published online: 30 December 2010 Ó International Union for the Study of Social Insects (IUSSI) 2010 Abstract Spatial distribution of ant workers within the nest is a key element of the colony social organization contributing to the efficiency of task performance and division of labour. Spatial distribution must be efficiently organized when ants are highly starved and have to get food rapidly. By studying ants behaviour within the nest during the beginning of food recruitment, this study demonstrates how the spatial organization is affected by starvation and improves the efficiency and the speed of recruitment as well as the allocation of food. (1) In starved nests, nestmates left the deep part of the nest and crowded near the nest entrance. This modification of the spatial distribution is a local phenomenon concerning only the individuals situated in the first chamber near the nest entrance. These starved individuals have a higher probability of leaving the nest after a contact with recruiters than nestmates situated deeper in the nest. This strongly suggests that nestmates situated near the nest entrance have a low response threshold to the signals emitted by recruiters. Their higher responsiveness speeds up their exit to the foraging area and hence may increase the efficiency of highly starved colonies in exploiting new food opportunities. (2) In starved nests, the trajectory covered by recruiters between contacts with nestmates was nearly twice A.-C. Mailleux (&) Earth and Life Institute, Université Catholique de Louvain, Croix du Sud 4 5, Carnoy, 1348 Louvain-la-Neuve, Belgium anne-catherine.mailleux@uclouvain.be G. Sempo C. Detrain J. L. Deneubourg Service d Ecologie Sociale, Université Libre de Bruxelles, CP231, Avenue F. D. Roosevelt, 50, 1050 Brussels, Belgium S. Depickère IRD, Inlasa, Entomología Médica, c/o Embajada Francia, CP9214 La Paz, Bolivia as small. For recruiters, this represented a gain of time in the allocation of food. As the recruitment process follows snowball dynamics, this gain of time by starved recruiters might also speed up the exploitation of food. This study evidences how the spatial distribution of individuals as a function of their motivational state might have a regulatory function in the recruitment process, which should be generic for many social species. Keywords Lasius niger Spatial organization Nestmates Foraging Starvation Introduction Spatial distribution of ant workers is a key element of the colony social organization contributing to the efficiency of task performance and division of labour (Fresneau and Corbara, 1990; Hölldobler and Wilson, 1990; Seeley, 1986, 1995). Indeed, the spatial distribution of individuals is often directly linked to the task they perform (Anderson and Ratnieks, 1999; e.g. for ants: Depickère et al., 2004, 2008a; Sempo et al., 2006a, b; Wilson, 1962; Wilson and Hölldobler, 2005; for bees: Seeley, 1995). A spatial segregation between specialized workers gives an adaptive advantage to the colony by decreasing, for instance, workers travel time between tasks (Seeley, 1995). Few studies have been designed specifically to show the influence of the food availability on the spatial organization in an ant colony. When the nest is highly starved, the spatial organization must be especially efficient since a quick discovery, exploitation and allocation of the food is essential for the colony survival. Previous works showed that the foraging activity of one colony usually increases with its starvation level (Franks et al., 1990; Mailleux et al., 2006;

2 220 A.-C. Mailleux et al. Toth and Robinson, 2005; Toth et al., 2005, Wallis, 1962). In this context, we studied the behavioural mechanisms at the origin of this intensified activity (Mailleux et al., 2000) and demonstrated that, for Lasius niger, this foraging adjustment is mediated by a higher responsiveness of nestmates to the recruitment trail laid by foragers. The foraging adjustment is also mediated by the food transfer between foragers and recruiters: loaded foragers entering the nest must transfer food to nestmates before returning to the food source. After a short starvation period, nestmates hardly accept a trophallaxis, which forces the forager to contact a higher number of nestmates before being unloaded. Conversely, after a long starvation period, the forager easily finds nestmates to accept its sugar freight. In conclusion, the high mobilization of foragers observed after a long starvation period is essentially due to an increased responsiveness by nestmates to the recruiting signal emitted by successful recruiters and to an increased tendency for nestmates to accept trophallaxis from returning foragers. In addition to these two rules adjusting foraging to starvation, we hypothesize that starvation may induce changes in the spatial organization of foragers within the nest (Depickère et al., 2008b). Here, we study how the spatial location of individuals within the nest is influenced by starvation before and after the beginning of recruitment to a food source. More precisely, the location of interactions between nestmates and recruiters in nests deprived of food for 1, 4 or 8 days is examined. Knowing that spatial organization in the nest could be an important regulatory mechanism of the foraging activity in social insects, our questions are: (1) Is the spatial distribution of nestmates before and after the beginning of a recruitment affected by deprivation? (2) Is the spatial distribution linked to the motivation to leave the nest? Materials and methods Mature queenless and broodless colonies (N = 3) of Lasius niger were collected from the slopes of earth banks in Brussels (Belgium). They numbered around 1,100 workers (range 800 1,400 workers) and were reared in plaster nests. Each colony was subdivided into three connected chambers ( cm) covered by a glass plate and a red plastic sheet and placed in plastic arenas ( cm) whose borders were coated with Fluon (Fig. 1). Nests were regularly moistened and kept at a temperature of 22 ± 3 with a 12-h photoperiod. Ants were fed three times a week with brown sugar solution (0.6 M), cockroaches (Periplaneta americana), and maggots (Calliphora erythrocephala). Nests were chosen to be as homogeneous as possible regarding the activity pattern of nestmates. This selection was made to obtain clear responses to only one factor, the starvation level, and to avoid variability due to other factors that might influence differently the foraging responses. Indeed, preliminary experiments showed that the tested nests presented similar pattern of activities. The three tested colonies were starved for 1, 4 or 8 days, each period of starvation being applied randomly and only once to each colony. After each experiment, the colonies were fed ad libitum during a resting period of 7 days before being randomly reassigned to another starvation period. The maximal starvation period (8 days) that we tested did not induce abnormal mortality (Lenoir, 1979). We studied the spatial location of recruiters and nestmates at the beginning of food recruitment. We focused our observations on the beginning of food exploitation (the first 20 min) since this period greatly influences further collective behaviour that may rapidly snowball. We focused on the ants located in the first chamber of the nest since, during the first 20 min of a recruitment phase, the ants located in the second and third chambers of the nest did not move (no significant change in their position, no change in their departure rates). Spatial distribution of nestmates before the entering of recruiters We studied the influence of food deprivation on the spatial distribution of nestmates. A printed transparent film divided into ten rectangular zones ( cm) was stuck on the glass plate of the first nest chamber (called the studied chamber; see Fig. 1). The rectangular zones were parallel to the front edge of the nest. The zone near the nest entrance was zone 1. We videotaped the ants located in this first nest section. On pictures magnified twice, the number of ants standing in each delimited zone was calculated. This was done when the nest was quiet, before the recruitment to a food source. Fig. 1 Experimental set-up: a transparent film divided into ten rectangular zones ( cm) was stuck on the glass plate of the first nest chamber. The number of ants standing in each delimited zone was calculated. This was done when the nest was quiet, before the recruitment to a food source

3 How does starvation affect spatial organization within ant nests? 221 Spatial distribution of interactions between nestmates and recruiters We studied the spatial location of the interactions between nestmates and a recruiter coming back to the nest from a foraging area. The recruiter had access to one liquid food [a 3-ll droplet of sucrose solution (0.6 M)]. The concentration and volume of these experimental droplets are close to the honeydew droplets produced by aphids. The first recruiter that discovered the food droplet returned to the nest. Once the recruiter left the nest after having contacted some nestmates, it was gently removed. Concomitantly, the removal of the foraging area prevented other ants from reaching the food and limited our study to the behaviour of only one recruiter and to its interactions with nestmates. For all the deprivation periods, all scouts returned to the nest, unloaded their sugar freight and left the nest within 5 min. During 20 min, we video-recorded antennal and trophallactic interactions within the first chamber of the nest (magnification 29). Three different behavioural groups were considered: (a) (b) (c) Re: Recruiter coming back to the nest after having ingested food ad libitum. tc: Nestmate having trophallactic contact with a recruiter. ac: Nestmate having antennal contact with a recruiter (the antennae of the recruiter touching the antennae of the nestmate) For these three behavioural groups, we measured: 1. The total trajectory covered by recruiters in the nest between each contact with nestmates: The recruiter s path in the nest was traced on a transparency stuck on the video screen and then reproduced on a graphic tablet. A data-processing programme measured this intranidal trajectory covered by recruiters. For instance, if a scout has two contacts, the total trajectory is the sum of the trajectory entrance 1st contact? trajectory 1st contact 2nd contact? trajectory 2nd contact entrance. 2. The contact entrance distance was the distance as the crow flies between the entrance and the location of the interactions between Re and tc or ac. This distance was calculated for each nestmate (tc and ac) and each contact. For instance, if the scout had two contacts, we calculated two contact entrance distances. This distance was evaluated with a simple ruler laid on the screen. The behaviour of nestmates was related to the type of contact with the recruiter (trophallactic or antennal contact), to their individual probability of leaving the nest during the first 20 min (the number of ants that left the nest within 20 min/the total number of observed ants), and to the distance from the nest entrance where the contact occurred. Statistical tests Data were analysed using Graph Prism version 5.01 (1998) for Windows (GraphPad Software, Inc., San Diego, CA, USA). Preliminary analyses showed no differences between colonies. The order of the treatments had no effects. Therefore, we pooled the data of all colonies and analysed them together. When data were normally distributed (total number of ants in the studied chamber, trajectory covered by recruiters between contacts), we gave their average values and the standard deviations and compared them with one-way ANOVA tests. When data were not normally distributed (contact entrance distance), we gave their median values and the 25% percentile and compared them with Kruskal Wallis tests. We also compared the slopes and the intercepts or regression lines. All tests were applied under two-tailed hypotheses, and the significant level of P value was set at Results Spatial distribution of nestmates before the entering of recruiters For all deprivation periods, the total number of ants in the studied chamber was about 100: after 1 dd (1 deprivation day): 111 ± 70 (N = 26); 4 dd : 101 ± 40 (N = 21); 8 dd : 99 ± 39 (N = 26) ( X SD, one-way ANOVA, F = 3.0, P [ 0.05, df = 2,70). After 1 dd, ants were homogeneously distributed between the ten zones of the studied chamber. The relative numbers of ants (rel N = number of ants in a zone X/sum of ants in the ten zones) do not differ significantly from each other (one-way ANOVA, F = 1.5, P = 0.16, df = 9,250). After 4 dd or 8 dd, the relative numbers of ants in each zone differed from each other (Fig. 2; one-way ANOVA test, after 4 dd : F = 7.4, P \ , df = 9,200; after 8 dd: F = 18.07, P \ , df = 9,250). For all deprivation periods, the rel N were linearly correlated to the location in the nest (Table 1). The slopes of the three lines were significantly different (F = 68.79, df = 724, P \ ). After 1 dd, the slope of the line was not different from 0 (Table 1). After 4 dd or 8 dd, the rel N decreased linearly with the distance in the nest (Table 1), and the slopes of the two lines were significantly different (F = 10.59, df = 466, P = 0.001). Indeed, as ants crowded against one another near the nest entrance, the rel N was higher near the nest entrance than deeper in the nest.

4 222 A.-C. Mailleux et al. Fig. 2 Spatial distribution of ants in the nest as a function of the food deprivation, when the nest was quiet, before the recruitment to a food source. The first chamber of the nest was divided into ten rectangular zones perpendicular to the nest entrance (see Fig. 1). Mean and standard deviation of the relative numbers of ants standing in each delimited zone were calculated. N = 26, 21 and 26 for 1, 4 and 8 days of starvation, respectively Therefore, the starvation level did not influence the total number of ants in the first chamber of the nest, but it shaped the spatial distribution of nestmates within the first chamber before the beginning of the food recruitment. Spatial location of the interactions between nestmates and recruiters The values of the contact entrance distances (the distance as the crow flies between the entrance and the location of the nestmates) are detailed in Table 2. Pooling nestmates having either a trophallactic or antennal contacts, the distances to the nest entrance decreased with deprivation. Among nestmates having a contact with a Fig. 3 Proportions of nestmates staying in the nest (open bars) and leaving the nest (filled bars) as a function of the location of the contact with the recruiter, after different starvation levels recruiter, some leave the nest and some stay without moving. The contact entrance distance was shorter for ants leaving the nest than for the ones staying in the nest except after 8 dd (Fig. 3; Table 2). Table 1 Influence of deprivation on the linear correlation between the relative number of ants as a function of the location in the nest (i.e. the spatial distribution of ants, see first paragraph of Results section) Parameters Deprivation period (days) Best-fit values Is slope significantly non-zero? Runs test Slope Y-intercept r F P value Deviation from zero P value Deviation from linearity Spatial distribution ns 0.69 ns \ s 0.19 ns \ s 0.98 ns The three colonies were pooled for the analyses represented in the tables

5 How does starvation affect spatial organization within ant nests? 223 Table 2 The contact entrance distance (cm) (the distance as the crow flies between the entrance and the location of the nestmates) Deprivation period (days) Nestmates leaving the nest Nestmates staying in the nest Mann Whitney (U, P value) Nestmates contacted by recruiters (trophallactic and antennal contacts) Nestmates having a trophallactic contact with recruiter Nestmates having an antennal contact with recruiter ± 1.4 (28) 3.2 ± 1.4 (126) ± 0.0 (40) 3.2 ± 1.4 (37) ± 0.0 (42) 1.4 ± 0.0 (32) KW, P value 7.37, P = , P = 0.01 Dunn, P value 1 versus 8, D = 20.14, P \ versus 8, D = 30.46, P \ versus 8, D = 34.55, P \ ± 1.4 (10) 3.2 ± 2.2 (37) ± 0.0 (17) 3.6 ± 1.4 (24) ± 0.0 (27) 1.4 ± 0.0(21) KW, P value 1.48, P = , P = ± 1.4 (18) 3.2 ± 1.4 (89) ± 0.0 (23) 2.2 ± 1.8 (13) ± 0.0 (15) 2.1 ± 0.0 (11) KW, P value 4.98, P = , P = 0.10 This parameter was compared as a function of the starvation period (Kruskal Wallis tests and Dunn s post hoc tests) and as a function of the behaviour of nestmates that leave the nest or nest (Mann Whitney). The three colonies were pooled for the analyses represented in the tables Considering only nestmates having a trophallactic contact with a recruiter, we observed that the distance was shorter for ants leaving the nest than for the ones staying in the nest except after 8 dd. The distance was not influenced by deprivation. When considering only the nestmates having an antennal contact with a recruiter, we observed that the contact entrance distance was not different for ants leaving the nest than for the ones staying in the nest. The distance was not influenced by deprivation. A global comparison between all distances for nestmates having a trophallactic or an antennal contact showed that these two groups were statistically different (Kruskal Wallis test, KW = 42.94, number of groups = 12, P \ ). Dunn s multiple comparison tests showed that only three groups differed in their contact entrance distances: the nestmates staying in the nest after 1 dd versus the nestmates leaving the nest after 4 dd and after 8 dd (1 dd vs. 4 dd, D = 82.90; 1 dd vs. 8 dd, D = with P \ 0.05 for both). Globally, the contact entrance distance decreased with deprivation and was shorter for nestmates leaving the nest. The trajectory covered by recruiters between contacts decreased with starvation: 1 dd : 8.9 ± 8.1 cm (N = 27); 4 dd : 5.7 ± 4.7 cm (N = 25); 8 dd : 4.0 ± 2.1 cm (N = 20) ( X SD, one-way ANOVA, F = 4.9, P = 0.01, df = 2, 73; Bonferri s multiple comparison test: 1 versus 4 dd : mean difference = 3.23, t = 2.05, P [ 0.05; 1 versus 8 dd : mean difference = 4.86, t = 3.06, P \ 0.05; 4 versus 8 dd : mean difference = 1.63, t = 1.01, P [ 0.05). This results, first, from the increase in the relative number of nestmates observed near the nest entrance after a long starvation period (see first paragraph of Results section) and second, from the higher ratio of nestmates accepting trophallactic exchanges after a long starvation period (Mailleux et al., 2010). Discussion When colonies undergo a deprivation period, the spatial organization of individuals inside the nest is modified. By studying ants behaviour within the nest during the beginning of food recruitment, this study demonstrates how the spatial organization within the nest improves foraging after a long starvation. The influence of starvation on ant colonies is usually studied in terms of behavioural changes at the level of recruiters by paying attention to their modulation of recruitment signals as well as to their food intake outside the nest (Hangartner, 1969; Josens and Roces, 2000; Mailleux et al., 2000, 2006). Nevertheless, food deprivation also acts on the behaviour of ants inside the nest. Indeed, starvation can elicit a spatial redistribution of workers that leave the depth of the first chamber of the nest and tend to aggregate themselves near the nest entrance. These nestmates located near the nest entrance have a higher probability of leaving the nest after a contact with recruiters than nestmates located deeper in

6 224 A.-C. Mailleux et al. the nest. This strongly suggests that nestmates located near the nest entrance are individuals with lower response thresholds to the tactile and/or chemical signals emitted by recruiting ants as suggested by de Biseau and Pasteels (2000). Nestmates higher responsiveness speeded up their exit to the foraging area and hence may increase the efficiency of highly starved colonies in exploiting new food opportunities. Ant spatial distribution within the nest based on threshold distribution has been poorly documented. This is the first work showing that starved nestmates modify their spatial location inside the nest. This behavioural adjustment implies a positive change in the pool of ants that can potentially be recruited by foragers. It is noticeable that, whatever the starvation level, the total number of ants in the first chamber does not change, and ants from the second and third chamber do not move into the first one. Therefore, the modification of the spatial distribution is a local phenomenon concerning only the individuals situated in a chamber near the nest entrance. In starved nests, the trajectory covered by recruiters between contacts with nestmates was about twice smaller. This result demonstrates the reorganization of nestmates that move closer to the nest entrance and display a higher propensity to receive food (Mailleux et al., 2010). This might represent an important gain of time for recruiters and consequently for the whole colony. Such changes might be important to take into account when studying foraging dynamics. Indeed, previous works showed that, in Leptothorax unifasciatus, ants are spatially organized in the nest, and individuals located at the periphery are most likely to leave the nest (Sendova-Franks and Franks, 1995). In Apis mellifera, the undertakers charged to remove dead bees from the hive are preferentially closer the entrance (Trumbo et al., 1997). In Myrmica sabuleti, the recruitment dynamics are adjusted as a function of the features of the source discovered: the mechanisms allowing this regulation could also involve the spatial relationship of the foragers within the nest and the trail response threshold distribution among foragers (de Biseau, 1994; de Biseau and Pasteels, 2000). Then, in the light of our results and knowing that collective foraging is mainly based on amplifying processes, a small time gained by foragers to unload near the nest entrance might play a key role by strongly shortening the beginning of the recruiting phase. This spatial reorganization and its regulatory function in the recruitment process seem to have a generic value applicable to many species of social insects. Acknowledgments We thank F. Saffre for writing the programme measuring the intranidal distance. This work was supported by the IRSIB (Institut d Encouragement de la Recherche Scientifique et de l Innovation de Bruxelles) for A. C. Mailleux and by the National Fund for Scientific Research (FNRS, Belgium) for G. Sempo, C. Detrain and J. L. Deneubourg. This is publication BRC175 of the Earth and Life Institute at UCL. References Anderson C. and Ratnieks F.L.W Task partitioning in insect societies. I. Effect of colony size on queuing delay and colony ergonomic efficiency. Am. Nat. 154: de Biseau J.C Mécanismes et valeurs adaptative du recrutement alimentaire chez Myrmica sabuleti (Hymenoptera, Formicidae): Approche expérimentale et théorique. Doctoral Thesis, Université Libre de Bruxelles, pp 188 de Biseau J.C. and Pasteels J.M Response thresholds to recruitment signals and the regulation of foraging intensity in the ant Myrmica sabuleti (Hymenoptera, Formicidae). Behav. Proc. 48: Depickère S., Fresneau D. and Deneubourg J.L Dynamics of aggregation in Lasius niger (Formicidae): influence of polyethism. Insect. Soc. 51: Depickère S., Fresneau D. and Deneubourg J.L. 2008a. Effect of social and environmental factors on ant aggregation: A general response? J. Insect Physiol. 54: Depickère S., Fresneau D., Deneubourg J.L. and Detrain C. 2008b. Spatial organization in ants nests: does starvation modify the aggregative behaviour of Lasius niger species? Insect. Soc. 55: Franks N.R., Bryant S., Griffiths R. and Hemerik L Synchronization of the behaviour within nests of the ant Leptothorax acervorum (Fabricius). I. Discovering the phenomenon and its relation to the level of deprivation. Bull. Math. Biol. 52: Fresneau D. and Corbara B Spatial organization in the nest during colony ontogeny in the Ponerine ant Pachycondyla (Neoponera) apicalis. In: Social Insects in the Environment (G.K. Veeresh, B. Mallik and C.A. Viraktamath, Eds), Oxford and IBH Publ. Co. Pvt. Ltd., New Delhi. pp 387 Hangartner W Structure and variability of the individual odor trail in Solenopsis geminata. Z. Vergl. Physiol. 62: Hölldobler B. and Wilson E.O The Ants. Cambridge, Massachussets: The Belknap Press of Harvard University Press, 732 pp Josens R.B. and Roces F Foraging in the ant Camponotus mus: nectar-intake and crop filling depend on colony deprivation. J. Insect Physiol. 46: Lenoir A Le comportement alimentaire et la division du travail chez la fourmi Lasius niger. Bull. Biol. Fr. Belg. 113: Mailleux A.C., Detrain C. and Deneubourg J.L How do the ants assess food volume? Anim. Behav. 59: Mailleux A.C., Detrain C. and Deneubourg J.L Starvation drives a threshold triggering communication. J. Exp. Biol. 209: Mailleux A.C., Buffin A., Detrain C. and Deneubourg J.L Recruiter or recruit: who boots the recruitment in starved nests in mass foraging ants? Anim. Behav. 79: Seeley T.D Social foraging by honeybees: How colonies allocate foragers among patches of flowers. Behav. Ecol. Sociobiol. 19: Seeley T.D The Wisdom of the Hive. Cambridge, MA: Harvard University Press. 295 pp Sempo G., Depickère S. and Detrain C. 2006a. How brood influences caste aggregation patterns in the dimorphic ant species Pheidole pallidula. Behav. Ecol. 53: Sempo G., Depickère S. and Detrain C. 2006b. Spatial organization in a dimorphic ant: caste specificity of clustering patterns and area marking. Insect. Soc. 17: Sendova-Franks A.B. and Franks N.R Spatial relationships within nests of the ant Leptothorax unifasciatus (Latr.) and their implications for the division of labour. Anim. Behav. 50: Toth A.L. and Robinson G.E Worker nutrition and division of labour in honeybees. Anim. Behav. 69:

7 How does starvation affect spatial organization within ant nests? 225 Toth A.L., Kantarovitch S., Meisel A.F. and Robinson G.E Nutritional status influences socially regulated foraging ontogeny in honey bees. J. Exp. Biol. 208: Trumbo S.T., Huang Z.Y. and Robinson G.E Division of labor between undertaker specialists and other middle-aged workers in honey bee colonies. Behav. Ecol. Sociobiol. 41: Wallis D.I The relation between hunger, activity and worker function in an ant colony. Proc. Zool. Soc. London 139: Wilson E.O Chemical communication among workers of the fire ant Solenopsis saevissima (Fr. Smith) 1. The organisation of mass-foraging. Anim. Behav. 10: Wilson E.O. and Hölldobler B Eusociality: Origin and consequences. Proc. Natl. Acad. Sci. USA 102: