The genetics of learning

The orientation flights of bees are lovely examples of behaviours structured to aid learning. The mushroom bodies of insects have long-been implicated in learning of all types. Here, evidence is presented to the genetic mechanisms related to learning in the MB following the orientation flights of naive bees as well as bees responding to the changes in the layout of the environment.  Interestingly, the mechanisms respond to visual novelty and show deeply conserved mechanisms.

Abstract: “The natural history of adult worker honey bees (Apis mellifera) provides an opportunity to study the molecular basis of learning in an ecological context. Foragers must learn to navigate between the hive and floral locations that may be up to miles away. Young pre-foragers prepare for this task by performing orientation flights near the hive, during which they begin to learn navigational cues such as the appearance of the hive, the position of landmarks, and the movement of the sun. Despite well-described spatial learning and navigation behavior, there is currently limited information on the neural basis of insect spatial learning. We found that Egr, an insect homolog of Egr-1, is rapidly and transiently upregulated in the mushroom bodies in response to orientation. This result is the first example of an Egr-1 homolog acting as a learning-related immediate-early gene in an insect and also demonstrates that honey bee orientation uses a molecular mechanism that is known to be involved in many other forms of learning. This transcriptional response occurred both in na+»ve bees and in foragers induced to re-orient. Further experiments suggest that visual environmental novelty, rather than exercise or memorization of specific visual cues, acts as the stimulus for Egr upregulation. Our results implicate the mushroom bodies in spatial learning and emphasize the deep conservation of Egr-related pathways in experience-dependent plasticity.”

Lutz, C. C., Robinson, G. E. 2013 Activity-dependent gene expression in honey bee mushroom bodies in response to orientation flight.JEB. 216, 2031-2038.

 

 

The many compasses of bees

Abstract: Honeybees have at least three compass mechanisms: a magnetic compass; a celestial or sun compass, based on the daily rotation of the sun and sun-linked skylight patterns; and a backup celestial compass based on a memory of the sun’s movements over time in relation to the landscape. The interactions of these compass systems have yet to be fully elucidated, but the celestial compass is primary in most contexts, the magnetic compass is a backup in certain contexts, and the bees’ memory of the sun’s course in relation to the landscape is a backup system for cloudy days. Here we ask whether bees have any further compass systems, for example a memory of the sun’s movements over time in relation to the magnetic field. To test this, we challenged bees to locate the sun when their known celestial compass systems were unavailable, that is, under overcast skies in unfamiliar landscapes. We measured the bees’ knowledge of the sun’s location by observing their waggle dances, by which foragers indicate the directions toward food sources in relation to the sun’s compass bearing. We found that bees have no celestial compass systems beyond those already known: under overcast skies in unfamiliar landscapes, bees attempt to use their landscape-based backup system to locate the sun, matching the landscapes or skylines at the test sites with those at their natal sites as best they can, even if the matches are poor and yield weak or inconsistent orientation.

Dovey, K. M., Kemfort, J. R., Towne, W. F. 2013 The depth of the honeybee’s backup sun-compass systems. JEB. 216, 2129-2139.

Interaction of visual and substrate cues

The interaction of private and public cues is an interesting aspect of ant navigation. Here, we have another demonstration of the primacy of visual cues during a spatial task, the task being the colony emigration of Temnothorax.

Abstract: “Many ants rely on both visual cues and self-generated chemical signals for navigation, but their relative importance varies across species and context. We evaluated the roles of both modalities during colony emigration by Temnothorax rugatulus. Colonies were induced to move from an old nest in the center of an arena to a new nest at the arena edge. In the midst of the emigration the arena floor was rotated 60°around the old nest entrance, thus displacing any substrate-bound odor cues while leaving visual cues unchanged. This manipulation had no effect on orientation, suggesting little influence of substrate cues on navigation. When this rotation was accompanied by the blocking of most visual cues, the ants became highly disoriented, suggesting that they did not fall back on substrate cues even when deprived of visual information. Finally, when the substrate was left in place but the visual surround was rotated, the ants’ subsequent headings were strongly rotated in the same direction, showing a clear role for visual navigation. Combined with earlier studies, these results suggest that chemical signals deposited by Temnothorax ants serve more for marking of familiar territory than for orientation. The ants instead navigate visually, showing the importance of this modality even for species with small eyes and coarse visual acuity.”

Bowens SR, Glatt DP, Pratt SC (2013) Visual Navigation during Colony Emigration by the Ant Temnothorax rugatulus. PLoS ONE 8(5): e64367.

Do ants need discrete snapshots?

It is well established in the literature that a single view of the world is unique to the location from where the view was perceived/stored. This leads to the idea of snapshot guidance to a single goal. A logical extension is to imagine routes could be composed of multiple snapshots which an ant uses in a sequence – an idea which was strongly suggested by specific experimental evidence. Here we present modelling results which show how evidence that had
suggested the use of discrete snapshots can in fact be produced with a mechanism that does not use discrete snapshots.

Antoine Wystrach, Michael Mangan, Andrew Philippides and Paul Graham (2013) Snapshots in ants? New interpretations of paradigmatic experiments. J Exp Biol 216:1766-1770

Artificial Compound Eyes

This week’s Nature has an interesting bit of engineering inspired by insect eyes:

Abstract: “In arthropods, evolution has created a remarkably sophisticated class of imaging systems, with a wide-angle field of view, low aberrations, high acuity to motion and an infinite depth of field. A challenge in building digital cameras with the hemispherical, compound apposition layouts of arthropod eyes is that essential design requirements cannot be met with existing planar sensor technologies or conventional optics. Here we present materials, mechanics and integration schemes that afford scalable pathways to working, arthropod-inspired cameras with nearly full hemispherical shapes (about 160 degrees). Their surfaces are densely populated by imaging elements (artificial ommatidia), which are comparable in number (180) to those of the eyes of fire ants (Solenopsis fugax) and bark beetles (Hylastes nigrinus). The devices combine elastomeric compound optical elements with deformable arrays of thin silicon photodetectors into integrated sheets that can be elastically transformed from the planar geometries in which they are fabricated to hemispherical shapes for integration into apposition cameras. Our imaging results and quantitative ray-tracing-based simulations illustrate key features of operation. These general strategies seem to be applicable to other compound eye devices, such as those inspired by moths and lacewings (refracting superposition eyes), lobster and shrimp (reflecting superposition eyes), and houseflies (neural superposition eyes).”

Young Min Song et al. (2013) Digital cameras with designs inspired by the arthropod eye. Nature 497, 95–99

Global and Local Scene Encoding

It is well understood that a key navigational mechanism for insects involves the learning of visual information from panoramic scenes. This leaves us with a basic question of how insects encode visual scenes for navigational. Computational studies have shown us how visual navigation can be achieved with: (i) Raw images; (ii) Sets of local visual features (such as oriented contrast edges), extracted from an image and tagged with retinal position; (iii) Sparse encodings where an entire scene is decomposed to simple parameters which represent a global property of the entire scene (such as centre of mass).

However, the fine details of real ants’ scene encoding remain a mystery. In Lent et al., we present evidence for ants using local and global scene features, when using vision for navigation. The major finding is that as ants aim to a particular point in a scene they learn the ratio of the shape that is in their left and right visual field. This simple ratio can then be used to guide paths to previously unseen shapes. We also see evidence of ants using local features, the interesting thing will be to ask how these mechanisms relate to each other within the dynamic learning process.

David D. Lent, Paul Graham, Thomas S. Collett (2013) Visual Scene Perception in Navigating Wood Ants Current Biology – 22 April 2013 (Vol. 23, Issue 8, pp. 684-690)

A glimpse of directional recruitment in ants

Although many ant species have very neat recruitment systems, in the form of their pheromone trail networks, some ant species – notably desert ants – do not use pheromone trails. Whether these ants have any other form of directional recruitment is a fascinating question. From this study of the foraging ecology of the Australian desert ant Melophorus bagoti, we get a hint that pheromone-independent directional mass recruitment may exist. Nests of M. bagoti were surrounded by feeders with only one of them baited with protein. Ants that had discovered the protein feeder were temporarily captured and then released en masse. Subsequent arrivals at all feeders were recorded.  Occasionally, very great numbers of recruits (a hundredfold more than at other feeders!) were observed at the target feeder specifically. Odour markings and plumes can be discounted, leaving us with a strangely evocative question about how directional information is transmitted from knowledgeable workers to recruits.

Patrick Schultheiss, Sabine S Nooten (2013) Foraging patterns and strategies in an Australian desert ant. Austral Ecology

How do odours and path integration interact?

For Cataglyphis ants, there are two particularly significant odours – are used to pinpoint food or nest. The nest is indicated by a Co2 plume, whereas food is indicated by complex molecules given off by recently dead insects. Before odour following can take place – guidance to the vicinity of food or nest is often driven by path integration. Here, Buehlman et al. show that ants will ignore Co2 plumes when they still have a large home vector, but are nevertheless attracted by food odours when a learnt food vector is still large. The pattern of results obviously reflects an adaptive solution, as you need to make sure you enter your own nest, though you don’t care whether the food you get is from the same patch as you have previously visited. The interesting question now is to ask whether this is driven by cognitive mechanisms are whether the simple sensori-motor ouptut of PI interacts differently with the potential plume following of different odours.

Buehlmann C, Hansson BS, Knaden M. (2013) Flexible weighing of olfactory and vector information in the desert ant, Cataglyphis fortis. Biol Lett 9: 20130070.

The learning of social learning

It has long been known that bees are attracted to flowers where other bees are currently feeding. Indeed, it has also been shown that bees can learn rewarding flower colours through the observation of conspecifics feeding on a particular flower. The question at the heart of this paper (Dawson et al.) is whether such social learning is an example of second order conditioning. In order to assess this, bees were allowed to feed in the presence of model bees – thus making an association between conspecifics and reward. These bees were then allowed to observe model bees on coloured feeders, when tested subsequently they preferred the flower colour that had been indicated by the model. Naive bees (that had not experienced an association between conspecifics and reward) showed no colour preference. By demonstrating that prior experience, of conspecifics paired with reward, is necessary for social learning, Dawson et al. claim it is a function of second order conditioning – rather than an innate process.

Erika H. Dawson, Aurore Avarguès-Weber, Lars Chittka, Ellouise Leadbeater (2013) Learning by Observation Emerges from Simple Associations in an Insect Model. Current Biology – 04 April 2013

 

More electric fields

We know that plants and bees share a close relationship developed through their coevolution. This has been most closely studied in terms of colours and odours. This paper from Clarke et al. shows how another channel might also be used a signal between plant and bee. They showed that floral electric fields, which are impacted by visits from naturally charged bees, can be discriminated by bees, providing extra information about the shape and recent history of a flower. As the authors say “… our study provides a framework for exploring the function and adaptive value of the perception of weak electric fields by bees in nature.”

 Dominic Clarke, Heather Whitney, Gregory Sutton, and Daniel Robert (2013) Detection and Learning of Floral Electric Fields by Bumblebees Science 5 April 2013: 340 (6128), 66-69