The Problems of a Priori Categorisation of Agonism and Cooperation: Circle‐Tube Interactions in Two Allodapine Bees

Abstract

Circle-tube experiments have been widely used to both examine nestmate recognition in social and solitary insects, as well as to characterise interactions in terms of agonism, cooperation, and avoidance. Despite their use in studies of halictid bees, carpenter bees, adrenid bees, and wasps, they have never been used to explore social interactions in the bee tribe Allodapini, where sociality varies widely in terms of complexity. Here, we investigate behaviours displayed within the highly eusocial allodapine Exoneurella tridentata and the facultatively social Exoneura robusta, using a standardised circle-tube apparatus. We present two broad and important results: (i) Discrimination between nestmates and non-nestmates is much more strongly expressed in the facultatively social species and (ii) principal components analyses do not indicate suites of behaviours that permit clear interpretations as being agonistic, cooperative, or avoidance. Our first set of results suggests that nestmate recognition is not an essential ability for social species. Our second set of results raise important questions about a priori functional categorisations of behaviours and whether these can be used to reliably classify observed behaviours as either avoidance, cooperation, or aggression. Our analyses highlight the risks of interpreting social insect interactions and call for a more cautious approach.Ethology 120 (2014) 551-562

Cryptic castes, social context and colony defence in a social bee, Tetragonula carbonaria

“…Given the cost of these antagonistic encounters, there is a strong functional benefit to ensuring that defensive behaviour is context‐specific. Our results may help explain the surprisingly low levels of aggression towards non‐nestmates in other bees, revealed from trials in isolated arenas (Dew, Gardner & Schwarz, ; Leonhardt, Form, Blüthgen, Schmitt & Feldhaar, ) but see (Breed & Page, ). Social context is also important in honey bees, whose defensive behaviour is usually only expressed at the nest entrance and environmental or social cues may be required to discriminate nestmates (Buchwald & Breed, ; Downs & Ratnieks, ).…”

“…N = 10. [Colour figure can be viewed at wileyonlinelibrary.com] Schwarz, 2014;Leonhardt, Form, Blüthgen, Schmitt & Feldhaar, 2011) but see (Breed & Page, 1991). Social context is also important in honey bees, whose defensive behaviour is usually only expressed at the nest entrance and environmental or social cues may be required to discriminate nestmates (Buchwald & Breed, 2005;Downs & Ratnieks, 1999).…”

Ontogeny of division of labor in a facultatively eusocial sweat bee Megalopta genalis

“…The ethogram was based on previous descriptions of M. genalis behavior (Arneson and Wcislo 2003;Wcislo and Gonzalez 2006). While the reliability of using a priori categorization of behaviors as agonistic or cooperative has been questioned for some allodapine bees (Dew et al 2014), the behaviors we classify avoid these problems because they are based on a long history of reliable results using circle tube experiments in other halictine bees (McConnell-Garner and Kukuk 1997;Pabalan et al 2000;Boesi et al 2009), and we do not attempt to distinguish between cooperative and aggressive behaviors (Table 1). Observers scoring the videos were blind to the caste and identity of each bee, which were identified by white paint marks on their thoraces.…”