Communication in foraging

General insights

What is the value of communication in groups? The answer to this question is likely to vary, but we have identified some common factors that are important to consider when evaluating the group- and individual-level benefits and costs of particular types of information and particular methods of communicating it. In particular, not only the costs of producing a signal need to be considered, but opportunity costs for senders (Dornhaus et al. 2006) and receivers (Dechaume-Moncharmont et al. 2005). The relative costs for senders and receivers are affected by whether information is ‘pushed’ by senders or ‘pulled’ by receivers (such as when it is deposited on a central place, e.g. a ‘blackboard’; (Dornhaus and Chittka 2004a; Dornhaus and Chittka 2005). Communication in groups also may lead to reduced levels of innovation (Lanan et al. 2012; Donaldson-Matasci and Dornhaus 2014). Benefits of communication depend on temporal dynamics of information (Dechaume-Moncharmont et al. 2005), how easy it is for individuals to personally acquire information (Dornhaus 2002; Dornhaus and Chittka 2004b; Donaldson-Matasci and Dornhaus 2012), whether it is used in multiple contexts (Cao and Dornhaus 2012), and expected quality of personal information vs. socially learned information (Donaldson-Matasci and Dornhaus 2012; Dunlap et al. submitted). Benefits may become visible at the level of the group rather than that of individuals (Donaldson-Matasci and Dornhaus 2012). There can also be benefits that depend on group size (Lanan et al. 2012), particularly when group size affects the pool of available information (Donaldson-Matasci et al. 2013).

Model systems/Methods

Social insects use a variety of communication strategies, particularly in foraging (rev. in (Dornhaus and Chittka 2004a; Dornhaus and Powell 2009; Dornhaus 2012; Dornhaus et al. 2012; Dornhaus 2014). Much of our understanding of why particular strategies evolved comes from modeling studies of bee (Dechaume-Moncharmont et al. 2005; Dornhaus et al. 2006a; Dornhaus et al. 2006b; Dornhaus 2012); rev. in (Raine et al. 2006) and ant foraging (Lanan et al. 2012); we have also studied social insects intensively in the field (Dornhaus 2002; Dornhaus and Chittka 2004b; Thom and Dornhaus 2007; Donaldson-Matasci and Dornhaus 2012; Bengston and Dornhaus 2013; Donaldson-Matasci et al. 2013; Donaldson-Matasci and Dornhaus 2014; Rivera et al. 2015), where they may forage over vast areas and hundreds of resources, as well as in the lab (Dornhaus and Chittka 1999; Dornhaus and Chittka 2001; Chittka et al. 2003; Dornhaus et al. 2003; Dornhaus and Cameron 2003; Dornhaus and Chittka 2005; Granero et al. 2005; Jones and Dornhaus 2011; Lanan et al. 2011; Leonard et al. 2011b; Cao and Dornhaus 2012; Kaczorowski et al. 2012; Leonard et al. 2013; Westling et al. 2014; Muth et al. 2015), where resources and information availability and reliability can be precisely controlled.

Specific examples: the bee 'dances' and the value of information communicated

Communication can be costly, not only in terms of the time and energy invested in the signal itself, but also in terms of successful foragers who return earlier than necessary from food sources to inform others or to update their own information (Dornhaus et al. 2006a), and in terms of potential foragers who wait at the nest to be available as recruits (Dechaume-Moncharmont et al. 2005). What this implies is that recruitment systems will only evolve if they confer sufficient benefits - but what are these benefits?

In honey bees (Apis mellifera), it seems that benefits of communicating the locations of profitable resources are more subtle than we might expect (Dornhaus and Chittka 2004b). Benefits lie not in a reduction in search time for recruits, but rather primarily in the more adaptive allocation of foragers to the best resources out of many available ones (Donaldson-Matasci and Dornhaus 2012). Somewhat counterintuitively, we show that in honey bees in the field, the distance foragers have to fly to find resources, and the quality of these resources, have no effect on communication benefits (Donaldson-Matasci and Dornhaus 2012), despite the fact that if resources are farther away, bees cannot find them quickly without the dance communication (Dornhaus 2002), and simulations predict that communicating locations of resources improves group success particularly if resources are sparse and low quality (Dornhaus et al. 2006b). We have also shown that larger colonies benefit more from this communication system in the field, probably because they are able to integrate information from more scouts, and thus able to arrive more quickly at short-lived resources (Donaldson-Matasci et al. 2013). In contrast, in a model that does not limit the duration that resources can be exploited, the benefit of communication was not affected by group size (Dornhaus et al. 2006b; Dornhaus 2012).

Bumble bees (Bombus terrestris and B. transversalis) on the other hand do not communicate locations of resources (Dornhaus and Chittka 1999). However, they employ both a signal (a pheromone signal) and a cue (nectar influx to honeypots, (Dornhaus and Chittka 2001) to socially transmit information about when foraging is profitable, and which flower scents to look for (Dornhaus and Chittka 1999; Dornhaus and Cameron 2003). A pheromone signal similar to this is likely also used by honey bees in addition to their dance (Thom and Dornhaus 2007). In bumble bees, the pheromone used is produced in tergal glands (Dornhaus et al. 2003), and its main active component appears to be eucalyptol (Granero et al. 2005). The fact that bees monitor nectar influx to honeypots means that honeypots function as a kind of ‘blackboard’, where information is deposited in a central place (Dornhaus and Chittka 2005). Bumble bees thus have two means of communicating the presence of profitable food sources, where one may be called a 'push-information' pathway (since bees can potentially perceive the signal wherever they are in the nest) and the other a 'pull-information' pathway (since bees have to actively check honeypots to pick up cues) (Dornhaus and Chittka 2004a).



(Dornhaus and Chittka 2004a; Raine et al. 2006; Dornhaus and Powell 2009; Dornhaus 2014)