J Comp Physiol A (2015) 201:795–802 DOI 10.1007/s00359-015-1019-2

ORIGINAL PAPER

Homing abilities of the tropical primitively eusocial marginata

Souvik Mandal1 · Raghavendra Gadagkar1

Received: 24 March 2015 / Revised: 30 May 2015 / Accepted: 1 June 2015 / Published online: 7 June 2015 © Springer-Verlag Berlin Heidelberg 2015

Abstract Compared to our extensive knowledge about Introduction the navigation and homing abilities of ants and bees, we know rather little about these phenomena in social wasps. Central place foragers, i.e. those that regularly return to a Here, we report the homing abilities of the tropical primi- particular place such as aggregation, nest or den, have to be tively eusocial wasp Ropalidia marginata and the factors able to find that central place efficiently. The ability to do that affect their homing success. To determine from how far so is referred to as homing. The study of homing behaviour these wasps can return to their nests, we transported for- of central place foragers includes determination of home agers blindfold and released them at gradually increasing range (Burt 1943; Hyane 1949; Potts et al. 2012) or forag- distances from their nests in four cardinal directions. Their ing range (Greenleaf et al. 2007; Osborne et al. 2008) and homing success was determined by checking their pres- also the “maximal homing distance”, i.e. the maximum dis- ence on their nests on three consecutive nights. All foragers tance from which can return back to their central (56 individuals, 115 releases) returned back from an area place (Carbone et al. 2005; Cresswell et al. 2000; Green- of 0.73 0.25 km2 on the day of release (minimal hom- leaf et al. 2007; Rau and Rau 1918; Wehner 1981). Arthro- ± ing area), whereas 83.8 % of the foragers (217 individuals, pods, especially constitute very attractive models 420 releases) returned when we enlarged the area of release for such studies (Cheng 2006) because of their large dis- to 6.22 0.66 km2 around their nests (maximal homing tances of daily travel relative to their body sizes, ease of ± area). Of 66 releases, no wasps returned from beyond the manipulation and experimentation and their relatively sim- maximal homing area. The minimal homing area might be pler neuronal machinery (Wystrach and Graham 2012). familiar to the foragers because they probably routinely Social insects such as ants, bees and wasps are especially forage in this area and the maximal homing area represents suitable because of their central place foraging nature and the maximum distances from which the wasps are capable their exceptional navigational abilities (Dyer 1998; Cress- of returning to their nests, with or without familiarity. well et al. 2000). However, as compared to bees and ants, social wasps have been rather poorly studied in this regard Keywords Homing · Minimal homing area · Maximal (for review, see Zeil 2012). homing area · Social wasp · Ropalidia marginata Ropalidia marginata is a tropical, independent found- ing, primitively eusocial wasp with indeterminate peren- nial nesting cycles, widely distributed in peninsular India. We now have a fair amount of information about its natu- Electronic supplementary material The online version of this article (doi:10.1007/s00359-015-1019-2) contains supplementary ral history, nesting biology and proximate mechanisms of material, which is available to authorized users. social organization. This has permitted us to use R. mar- ginata as a model system to begin to understand the evo- * Raghavendra Gadagkar lutionary forces that promote social life and associated [email protected] altruistic behaviours (Gadagkar 2001). However, lack of 1 Centre for Ecological Sciences, Indian Institute of Science, information about its orientation, navigation and hom- Bangalore 560012, India ing abilities and knowledge of the extent of geographical

1 3 796 J Comp Physiol A (2015) 201:795–802 activity of individual wasps remains an important lacuna in Fig. 1 Satellite image (Google map) of the Indian Institute of Sci-▸ making this an ideal model system. We have there- ence campus and its surroundings. Axes are drawn keeping the posi- tions of the nests at the origin in each box. Wasps were released at fore initiated a series of studies in an attempt to correct this different locations along the axes. The polygon drawn with dotted imbalance in our knowledge of this species. In this paper, white line depicts the minimal homing area and the polygon drawn we report the results of experiments designed to assess the with solid white line depicts the maximal homing area of each nest. homing abilities of R. marginata foragers in their natural Contours were drawn using cubic spline interpolation (Kimeldorf and Wahba 1970; Powell 2000) (Colour on the Web only) nesting habitat following experimental displacement (Fabre 1914; Papi 1992; Wehner 1981) and to understand the fac- tors that affect homing success. Releasing sites comprised of woods inside the campus, roofs of buildings and roads inside (with much less motor traffic) and outside (with heavy motor traffic) of the cam- Methods pus. As mentioned above, release sites inside the campus consisted mostly of trees and small buildings and a few Definitions large buildings whereas area release sites outside the cam- pus were comprised mostly of highly dense residential, In this study, we define homing success as the ratio of commercial and industrial areas with large multistoreyed the number of successful returns to the nest and the total buildings and very less vegetation. number of releases from a particular bin (for, e.g. a release site or a group of foragers which spend 0.1 to 0.2 propor- Experimental design and data collection tion of time outside their nest prior to the experimental release). We define ‘minimal homing distance’ (for a par- All individual wasps of each nest were uniquely colour ticular direction) as the farthest distance in that direction marked with quick-drying Testors® enamel paints. Quan- from which all displaced wasps returned to their nest on titative behavioural data were collected for 10 h over two the day of release and the ‘maximal homing distance’ as the days prior to the release experiment, to determine the pro- farthest distance from which at least one displaced wasp portion of time each wasp spent outside of its nest (Gadag- returned to her nest within three days of release. Similarly, kar 2001). Individuals who brought food to their nests at we defined minimal and maximal homing areas as the least once were identified and designated as foragers. Fol- areas circumscribed by the minimal and the maximal hom- lowing this, on sunny days, foragers were removed from ing distances, respectively, in the four cardinal directions. their nests between 0600 and 0630 h (i.e. before their first With these definitions, we have determined the minimal foraging trip of the day) and kept in aerated glass vials. and maximal homing distances and areas for four naturally Starting at 0830 h, when foragers normally start foraging, occurring nests and attempted to understand the factors vials with captured animals were transferred to a polythene affecting homing distance and area. bag that did not block light completely but did not permit a view of the outside, and were transported to the release Study site and nest selection sites on a bicycle. All wasps were released between 0900 and 1030 h. Release sites were determined beforehand with The study was performed between October 2010 and a GPS machine (GARMIN etrex VISTA HCx) with a maxi- August 2012, using four naturally established nests of R. mum 3 m error. ± marginata (namely N3, N4, N5 and N6), which all were Releases were made systematically at 100 m intervals in considered healthy (with brood in all cells and no parasites) four cardinal directions with respect to the nests. At each (Gadagkar 2001) and contained at least 40 females. The release site, a minimum of one and a maximum of four studied nests were found inside electric boxes, located in individuals were released; no more than two individuals Indian Institute of Science, Bangalore, India (N3 located were released on the same day from a site. First, wasps at N 13°01′03.2″, E 077°34′02.4″; N4 at N 13°01′26.0″, were released 100 m away from their nests in four cardi- E 077°33′57.3″; N5 at N 13°01′03.2″, E 077°34′02.4″; nal directions. If at least one amongst them had returned N6 at N 13°01′26.0″, E 077°33′57.3″). The main campus, within three days of release, on the following day, different spread over 1.55 km2 area (calculated using Google Earth), wasps were released from another 100 m away in the same is a mixture of small to medium academic and residential direction. No wasp was released in a same direction on two buildings (the tallest buildings are no more than 20 m in consecutive days. From any release site, if none of the four height), roads with light traffic and woods ranging from released wasps returned within three days after release, shrubs to large trees, summing up to a 89.07 % of green that releasing site was assumed to be out of the maximal cover, thus providing a landscape with no apparent barrier homing distance of the wasps of that nest and no more for wasp traffic (see Fig. 1). wasps were released further on that direction. The previous

1 3 J Comp Physiol A (2015) 201:795–802 797

4km 4km

3km 3km

2km 2km

lkm lkm

0 lkm 2km 3km 0 lkm 2km 3km

4km 4km

3km 3km

2km 2km

lkm lkm

0 lkm 2km 3km 0 lkm 2km 3km

1 3 798 J Comp Physiol A (2015) 201:795–802

Table 1 Minimal and maximal homing distances and areas of the four nests (a more detailed account is in Table S1) Nest Minimal/maximal Total releases Total returns Minimal and maximal Minimal and maximal Homing success (Total homing distances (m) homing areas (km2) returns/total releases) North East South West

N3 Minimal 40 40 300 300 400 600 0.459 1 Maximal 124 104 1500 1000 1500 1400 5.28 0.84 N4 Minimal 27 27 600 600 800 500 1.036 1 Maximal 92 77 1500 1400 1600 1400 6.38 0.84 N5 Minimal 21 21 300 600 400 400 0.505 1 Maximal 105 86 1500 1500 1500 1400 6.4 0.82 N6 Minimal 26 26 600 500 700 600 1.045 1 Maximal 109 83 1600 1400 1600 1500 6.83 0.83

The calculation of maximal homing areas and distances is based on all the releases performed within maximal homing areas, which includes the releases within minimal homing areas and the areas outside minimal homing areas but within maximal homing areas. Releases outside the maxi- mal homing areas are excluded from the calculation release site from where at least one wasp had returned was 1970; Powell 2000), two imaginary (virtual) boundaries assumed as the maximal homing distance in that direction. were drawn around each nest to demarcate two areas, the Wasps were repeatedly used for the experiment until they minimal homing area and the maximal homing area. The failed to return or disappeared from their nest. For the first minimal homing area was determined by considering release, individuals were randomly picked from the poly- the farthest release points (from each cardinal direction) thene bag for release at each site. Subsequently, individuals from where all released individuals returned on the day were released only in a direction that was different from its of release. Similarly, the maximal homing area was deter- release on the previous day. If an individual was released mined by considering the farthest release points (from 1000 m or more from the nests, it was not released on the each cardinal direction) from where at least a fraction of next day. On each day, a census of all individuals present released individuals returned. All the data analysis was car- on the nest was made between 2100 h and 2200 h, to deter- ried out using the statistical package R (version 3.0.0 for mine which of the released wasps had returned each day. Windows) (R core team 2013). We compared distances and Individuals were considered to have returned successfully, areas using two-factor ANOVA and homing successes (pro- if they were found on their nests during the night census portions) with (non-parametric) Friedman test. within three days of release. Once all identified forager wasps were lost, the experiment was temporarily suspended for that nest. After a minimum of 2 weeks, that nest was Results reused and the experiment continued with fresh quantita- tive observation to identify new foragers. For each nest, the The minimal and maximal homing distances and areas experiment was terminated when no released individuals returned from any of the four cardinal directions. From the To determine the homing areas and homing distances of the four nests put together a total of 249 unique foragers (68, wasps, individuals known to be foragers (having brought 49, 70 and 62, for nests N3, N4, N5 and N6, respectively) food to the nest at least once) were released in four cardi- were released one or more times, amounting to a total of nal directions and their homing success was recorded. The 486 releases. We computed retrospectively that each wasp minimal homing distances in the four cardinal directions, was released 2.32 1.68 times. based on 56 individuals and 115 releases, ranged from ± about 300 to 800 m (mean SD 512.5 150 m) across ± = ± Data analysis the four nests and directions. Similarly the maximal hom- ing distances, based on 217 individuals and 420 releases, Release sites were categorized as (a) sites from where all ranged from about 1000 to 1600 m (mean SD 1456.2 ± = the released individuals returned on the day of release, (b) 5 141.27 m). The minimal homing area was 0.46, 1.04, ± sites from where some or all released individuals returned 0.50 and 1.04 km2, respectively, for the four nests studied within three days of release, but not on day of release (mean SD 0.76 0.32 km2, cv 0.425). Similarly, ± = ± = and, (c) sites from where none of the individuals returned. the maximal homing area was 5.28, 6.38, 6.4 and 6.83 km2, Using cubic spline interpolation (Kimeldorf and Wahba respectively (mean SD 6.22 0.66 km2, cv 0.106). ± = ± = 1 3 J Comp Physiol A (2015) 201:795–802 799

Fig. 2 Homing success of released Ropalidia marginata foragers ent from zero, P < 0.0005, AIC –9.633. Sigmoidal function (solid on each day after release. Inset Homing success (pooled across all line): Y 1/(1 (X/A)B), where =A 1347, P < 0.0005 and B 4.9, days) fitted as a function of distance of release from the nests. Linear P 0.002,= AIC+ –12.986) (Colour= on the Web only) = model (broken line): Y –0.0005X 1.17; slope significantly differ- = = = +

A comparison of the coefficients of variation revealed that wasps were released. The rate of decrease in homing suc- the minimal homing areas varied significantly more than cess, as a function of distance was better explained by a the maximal homing areas (Z 1.656, P 0.04) (Zar sigmoidal function (by a factor of 5.34) as compared to a = = 2009). Data on each of the four nests are given in Table 1 linear function. Thus, homing success decreased very grad- and a graphical representation of the minimal and maxi- ually up to about 700 m and decreased very rapidly thereaf- mal homing areas and distances of the four nests is given ter (Fig. 2 inset). in Fig. 1. In the four nests studied, 74.76 % of the wasps Homing success was also influenced by the propor- released outside the minimal homing areas but inside the tion of their time that the released wasps had spent outside maximal homing areas successfully returned to their nests their nest before being subjected to the experimental dis- whereas none of the 66 wasps released beyond the maximal placement, such that wasps which had spent intermediate homing area returned to their nests. proportions of their time outside their nests were the most successful ones whereas those that had spent very low or Factors affecting homing success very high proportions of time outside their nests were less successful (Fig. 3). Considering all wasps released within the maximal hom- ing areas (which includes the minimal homing areas and Factors affecting minimal and maximal homing excluding wasps released beyond the maximal hom- distances ing areas), homing success (proportion of released wasps that successfully returned to their nests) was not different Considering all four nests together, we found that there between nests (considering all directions) and between were no significant differences between the four cardi- directions (considering all nests) (Friedman test, Nest nal directions in either the minimal or maximal hom- χ2 1.275, P 0.735; Direction χ2 4.5, P 0.212). ing distances (two-factor ANOVA, minimal homing dis- = = = = Pooling data from all nests and all directions, we found that tance F 0.684, P 0.584; maximal homing distance 3,9 = = homing success decreased with the distances at which the F 3.604, P 0.059). Similarly considering all four 3,9 = = 1 3 800 J Comp Physiol A (2015) 201:795–802

Fig. 3 Homing success of Ropalidia marginata foragers as a func- each data point represents the ratio of the number of releases in tion of proportion of time spent outside of the nests by the foragers which foragers were successful and the total number of releases from recorded during the quantitative observation on two days prior to a bin with 0.1 interval in the proportion of time the foragers spent experimental release. Foragers with an interval of 0.1 proportion of outside the nest prior to experimental release. Polynomial model: time spent outside of nest are binned for a single data point. Thus, Y –0.0001X2 0.01X 0.6943, R2 0.8098, P < 0.01 = + + = directions together, we found no significant differences them because they routinely forage in this area [generally between nests in the minimal or maximal homing dis- referred to as home range (Burt 1943)]. The minimal hom- tances (two-factor ANOVA, minimal homing distance ing area is derived from the minimal homing distances in F 3.421, P 0.066; maximal homing distance: the four cardinal directions, averaging to about 500 m. If 3,9 = = F 1.899, P 0.200). we thus consider that the wasps routinely fly within about 3,9 = = As in the case of homing success, homing distance was 500 m from their nests (although we might be slightly over- also influenced by the proportion of their time that the estimating this area as some wasps may have searched hard released wasps had spent outside their nest before being even before returning on the day of release), this compares released. Wasps which had spent intermediate proportion very favourably with the value of 300 400 m reported − of their time outside their nests showed the longest hom- for other social wasps (Ugolini and Cannicci 1996). On ing distances whereas those that had spent very low or very the other hand, many bees, mostly social and also some high proportions of time outside their nests showed smaller solitary, seem to forage over much greater distances (Gath- homing distances (Fig. 4). mann and Tscharntke 2002; Zurbuchen et al. 2010). It is worth noting that compared to polylectic species, mon- olectic/oligolectic species have greater foraging distances, Discussion which might be due to their requirement of particular host plants for the latter. It will be interesting to investigate how For the tropical eusocial wasp R. marginata, this study pro- social structure, ecological niche, dietary habits and forag- vides data on the minimal and maximal homing distances ing behaviour mould the evolution of minimal homing area as well as minimal and maximal homing areas obtained in different species of insects. by releasing foragers at gradually increasing distances in Since many of the released wasps did not return from all four cardinal directions and determining the probabil- outside the minimal homing area on the day of release, and ity of their successful return to their nests. We show that not all but only a proportion of the released wasps returned the wasps return rapidly and efficiently when released from outside the minimal homing area [and inside the within about 500 m from their nests and cannot return at maximal homing area (about 6.22 km2)], we suspect that all when released beyond about 1500 m from their nests. this area may not be routinely used by the wasps for for- Since all the released foragers returned from the mini- aging. We also suspect that those that did return from this mal homing area (about 0.73 km2 around their nests) on area may have done so independent of previous familiarity; the day of release, we suspect that this area is familiar to perhaps they were individuals which, instead of depending

1 3 J Comp Physiol A (2015) 201:795–802 801

Fig. 4 Distance from which released foragers returned successfully as a function of proportion of time spent outside of the nests by the foragers on previous two days. Solid black dots represent mean return distances (whiskers represent standard deviation) of foragers from a bin of 0.1 proportion of time spent outside of the nests by the foragers on previous two days. The black line represents a polynomial fit (Y –2.21 63X2 2.1956X =0.3805, R2 +0.5491) for +all the data. Data= were binned as described in the legend to Fig. 3

only on familiar landmarks, may have used other means of starvation and/or lose their status on their previous nest; if navigation (for, e.g. ability to perform systematic search). it thus takes too long, they may find it more profitable to Consistent with this interpretation, we find that homing initiate their own nest. Unlike the sterile workers of hon- success on the day of release decreased sharply after about eybees and stingless bees, who have no option of starting 500 m. their own new nests and who have no danger of losing their If time spent outside the nests is a simple proxy for status on their nests, R. marginata foragers are capable of familiarity, we would expect a linear dependence of hom- initiating their own nests (Gadagkar 2001) and are also sus- ing success on the time spent outside. However, unlike ceptible to losing their status on their nests (Bang and Gad- in previous studies (Ugolini and Cannicci 1996), we find agkar 2012). This may account for why these wasps do not that wasps which spent intermediate amounts of time out- return from as large distances as many of the bee workers side their nests (prior to release) had the greatest homing do (Chmurzyn´ski et al. 1998; Goulson and Stout 2001; Pahl success and also returned from greater distances. Their et al. 2011; Roubik and Aluja 1983; Visscher and Seeley greater homing success may be because wasps which are 1982). experienced and most familiar with their surrounding may If the minimal homing areas are those which are rou- actually require less time for their daily foraging activi- tinely used for foraging, then variation in food availability ties. Their ability to return from greater distances may be around the nest (Osborne et al. 2008) due to spatial and sea- because they might also return from places that is within sonal variability and variation in demand for food (How- the maximal homing area but that they have never visited ard and Jeanne 2004) between different nests may lead to before, and they may do so using systematic search, the variation in the minimal homing area from nest to nest. Our ability for which they might acquire during their foraging results that the minimal homing area varies more from nest experience. No wasps may have returned from outside the to nest compared to the maximal homing areas therefore maximal homing area either because they were genuinely support our suspicion that the minimal homing area may be lost or because even the navigational ability of the individ- the area in which the wasps routinely forage. On the other uals may not make it possible for them to return from such hand, the ability to return from outside the minimal hom- a large distance soon enough so as not to face the risk of ing area, probably without familiarity and probably based

1 3 802 J Comp Physiol A (2015) 201:795–802 on other navigational abilities may be a characteristic of Fabre JH (1914) The mason bees. Dodd, Mead and Company, New the species rather than the nest. Consistent with this inter- York Gadagkar R (2001) The social biology of Ropalidia marginata: pretation, we also find that the maximal homing distances toward understanding the evolution of . Harvard Uni- do not vary significantly between the four cardinal direc- versity Press, Cambridge tions and that the maximal homing areas do not differ sig- Gathmann A, Tscharntke T (2002) Foraging ranges of solitary bees. J nificantly between nests. Thus, if the minimal homing area Anim Ecol 71(5):757–764 Goulson D, Stout JC (2001) Homing ability of the bumblebee Bom- is a function of familiarity and maximal homing area is a bus terrestris (: Apidae). Apidologie 32:105–111 function of capability, the difference between these two Greenleaf SS, Williams NM, Winfree R, Kremen C (2007) Bee for- areas would be determined by the area of familiarity which aging ranges and their relationship to body size. Oecologia in turn would depend on resource availability. For, e.g. in 153:589–596 Howard KJ, Jeanne RL (2004) Rates of brood development in a social a very resource-poor environment, the minimal and maxi- wasp: effects of colony size and parasite infection. Insectes Soc mal homing areas may be the same. It thus appears that the 51:179–185 homing abilities of R. marginata are the result of a compro- Hyane DW (1949) Calculation of size of home range. J Mammal mise between the vagaries of resource distribution, intrin- 30:1–18 Kimeldorf GS, Wahba G (1970) A correspondence between Bayes- sic navigational abilities of the wasps and the evolutionary ian estimation on stochastic processes and smoothing by splines. motivation to return home rather than make a new home. Ann Math Stat 41:495–502 Osborne JL, Martin AP, Carreck NL, Swain JL, Knight ME, Goulson Acknowledgments We thank the Department of Science and Tech- D, Hale RJ, Sanderson RA (2008) Bumblebee flight distances in nology, the Department of Biotechnology, and the Ministry of Envi- relation to the forage landscape. J Anim Ecol 77:406–415 ronment and Forests, Government of India for financial assistance. Pahl M, Zhu H, Tautz J, Zhang S (2011) Large scale homing in honey- SM and RG designed the study, SM conducted the study and RG and bees. PLoS One 6(5):e19669. doi:10.1371/journal.pone.0019669 SM cowrote the paper. We thank Sandeep Pulla, for suggesting the Papi F (1992) General aspects. In: Papi F (ed) homing. Chap- use of cubic spline to draw the boundaries of the homing areas and man and Hall, London, pp 5–6 Anjan Nandi for guidance in data analysis and both for their com- Potts JR, Harris S, Giuggioli L (2012) Territorial dynamics and sta- ments on the manuscript. All experiments reported here comply with ble home range formation for central place foragers. PLoS One the current laws of the country in which they were performed. This 7(3):e34033. doi:10.1371/journal.pone.0034033 article does not contain any studies with human participants per- Powell RA (2000) Animal home ranges and territories and home formed by any of the authors. The authors declare that they have no range estimators. In: Boitani L, Fuller T (eds) Research tech- conflict of interest. niques in animal ecology: controversies and consequences. Columbia University Press, New York, pp 65–110 Rau P, Rau N (1918) Wasps studies afield. Princeton University Press, Princeton References Roubik DW, Aluja M (1983) Flight ranges of Melipona and Trigona in tropical forest. J Kans Entomol Soc 56:217–222 Bang A, Gadagkar R (2012) Reproductive queue without overt con- R Core Team (2013) R: a language and environment for statistical flict in the primitively eusocial wasp Ropalidia marginata. Proc computing. R foundation for statistical computing, Vienna, Aus- Nat Acad Sci 109:14494–14499. doi:10.1073/pnas.1212698109 tria. ISBN 3-900051-07-0. http://www.R-project.org/ Burt WH (1943) Territoriality and home range concepts as applied to Ugolini A, Cannicci S (1996) Homing in paper-wasps. In: Turillazzi mammals. J Mammal 24:346–352 S, West-Eberhard MJ (eds) Natural history and evolution of Carbone C, Cowlishaw G, Isaac NJB, Rowcliffe JM (2005) How far paper-wasps. Oxford University Press, Oxford, pp 126–143 do animals go? Determinants of day range in mammals. Am Nat Visscher PK, Seeley TD (1982) Foraging strategy of honeybee colo- 165:290–297 nies in a temperate deciduous forest. Ecology 63:1790–1801 Cheng K (2006) navigation: ants, bees, crabs, spiders find- Wehner R (1981) Spatial vision in . In: Autrum H (ed) ing their way. In: Wasserman EA, Zentall TR (eds) Compara- Handbook of sensory physiology, vol VII/6c. Springer, Berlin, tive cognition: experimental explorations of animal intelligence. pp 287–616 Oxford University Press, Oxford, pp 189–209 Wystrach A, Graham P (2012) What can we learn from studies of Chmurzyn´ski JA, Kieruzel M, Krzysztofiak A, Krzysztofiak L (1998) navigation? Anim Behav 84:13–20 Long-distance homing ability in Dasypoda altercator (Hyme- Zar JH (2009) Biostatistical analysis, 4th edn. Dorling Kindersley noptera, Melittidae). Ethology 104(5):421–429 Publishing Inc., London, pp 141–145 Cresswell JE, Osborne JL, Goulson D (2000) An economic model of Zeil J (2012) Visual homing: an insect perspective. Curr Opin Neuro- the limits to foraging range in central place foragers with numer- biol 22:285–293. doi:10.1016/j.conb.2011.12.008 ical solutions for bumblebees. Ecol Entomol 25:249–255 Zurbuchen A, Landert L, Klaiber J, Müller A, Hein S, Dorn S (2010) Dyer FC (1998) Spatial cognition: lessons from central-place foraging Maximum foraging ranges in solitary bees: only few individuals insects. In: Balda RP, Pepperberg IM, Kamil AC (eds) Animal have the capability to cover long foraging distances. Biol Con- cognition in nature: the convergence of psychology and biology serv 143:669–676 in laboratory and field. Academic Press, California, pp 119–154

1 3 Homing abilities of the tropical primitively eusocial paper wasp Ropalidia marginata Souvik Mandal and Raghavendra Gadagkar. Centre for Ecological Sciences, Indian Institute of Science, Bangalore, 560012, India. Online supplementary material

Supplementary table S1: Summary of the releases and number of unique forager wasps released from four nests. Abbreviations; R1 = Region inside minimal homing area, R2 = Region outside of minimal homing area but inside of maximal homing area, R3 = Region outside of maximal homing area, HS = Homing Success.

NORTH EAST SOUTH WEST TOTAL No. of unique Nests Regions Releases Returns HS Releases Returns HS Releases Returns HS Releases Returns HS Releases Returns HS wasps released N3 R1 7 7 1 8 8 1 10 10 1 15 15 1 40 40 1 23 R2 27 20 0.74 17 9 0.53 23 20 0.87 17 15 0.88 84 64 0.76 45 R3 4 0 0 8 0 0 4 0 0 4 0 0 20 0 0 0 R1+R2 34 27 0.79 25 17 0.68 33 30 0.91 32 30 0.94 124 104 0.84 68 Total 38 27 33 17 37 30 36 30 144 104 68 N4 R1 7 7 1 6 6 1 9 9 1 5 5 1 27 27 1 9 R2 17 14 0.82 17 12 0.71 14 11 0.79 17 13 0.76 65 50 0.77 27 R3 4 0 0 4 0 0 4 0 0 4 0 0 16 0 0 13 R1+R2 24 21 0.88 23 18 0.78 23 20 0.87 22 18 0.82 92 77 0.84 36 Total 28 21 27 18 27 20 26 18 108 77 49 N5 R1 3 3 1 10 10 1 4 4 1 4 4 1 21 21 1 15 R2 24 17 0.71 18 13 0.72 22 20 0.91 20 15 0.75 84 65 0.77 41 R3 2 0 0 4 0 0 4 0 0 4 0 0 14 0 0 14 R1+R2 27 20 0.74 28 23 0.82 26 24 0.92 24 19 0.79 105 86 0.82 56 Total 29 20 32 23 30 24 28 19 119 86 70 N6 R1 6 6 1 5 5 1 9 9 1 6 6 1 26 26 1 9 R2 24 12 0.5 17 11 0.65 20 17 0.85 22 17 0.77 83 57 0.69 48 R3 4 0 0 4 0 0 4 0 0 4 0 0 16 0 0 5 R1+R2 30 18 0.6 22 16 0.73 29 26 0.9 28 23 0.82 109 83 0.76 57 Total 34 18 26 16 33 26 32 23 125 83 62 All R1 23 23 1 29 29 1 32 32 1 30 30 1 114 114 1 56 Nests R2 92 63 0.68 69 45 0.65 79 68 0.86 76 60 0.79 316 236 0.75 161 R3 14 0 0 20 0 0 16 0 0 16 0 0 66 0 0 32 R1+R2 115 86 0.75 98 74 0.76 111 100 0.9 106 90 0.85 430 350 0.81 217 Total 129 86 118 74 127 100 122 90 496 350 249