Article Recommendations
Short pieces on papers I've admired, written for Faculty Opinions
Wormholes in virtual space: From cognitive maps to cognitive graphs. Warren WH, Rothman DB, Schnapp BH, Ericson JD. Cognition 2017 09; 166:152-163
​
Some of the most interesting, and most fiercely contested, questions in cognition research are those relating to how spatial memories are stored and the mechanisms by which they are processed to allow navigation to desired locations. It has long been proposed that humans, and other animals, have a ‘cognitive map’, often taken to mean that the locations of places are encoded as coordinates within a common, Euclidean reference frame.
Warren and colleagues investigated the nature of humans’ spatial memories using a genuinely innovative study in which participants were asked to navigate around a virtual reality maze containing ‘wormholes’ which instantaneously ‘teleported’ them from one place to another while simultaneously rotating them. The subjects learned to navigate easily and efficiently within the maze without even realising that it was geometrically impossible, but in doing so violated various postulates of Euclidean geometry.
The authors propose that the behaviour of participants in the wormhole maze was best explained by the hypothesis that they organised their knowledge of the maze as a ‘labelled graph’ in which the distances and angles between pairs of locations were stored, but locations were not encoded in a common coordinate system.
While the results are open to multiple interpretations, they present powerful evidence against the idea of a Euclidean map in the brain, and the idea of navigation through ‘impossible’ virtual words is a powerful one with great potential to enrich our understanding of spatial cognition.
​
​
Taking an insect-inspired approach to bird navigation. Pritchard DJ, Healy SD. Learn Behav 2018 03; 46(1):7-22
The need to navigate accurately to particular locations in space, whether to find food, shelter, mates or other resources, is common to animals across a wide range of taxa, body types and brain sizes. Indeed, a powerful argument for studying navigation behaviour is that examining how widely varying species cope with a common challenge could shed light on the evolution of cognitive strategies, yet the approaches of researchers to studying navigation in insects and vertebrates have diverged widely.
The popular hypotheses to explain spatial cognition, as well as the experimental approaches used to test those hypotheses, differ markedly between the two fields, but it is far from clear whether these differences reflect real differences in the cognitive strategies used by the two groups or might, instead, stem from the ways they have been studied.
In this fascinating and thoughtful paper, Pritchard and Healy use hummingbirds as a case study to ask whether taking inspiration from the ideas and approaches current in studies of insect navigation (predominantly ants and bees), might explain some perplexing results seen in birds and provide inspiration for productive future study. They identify two themes, in particular, that are common in insect studies but rarely considered in the bird literature: first, that various strategies by which the animal compares its current view to memorised ones - as opposed to constructing cognitive models of space that are independent of the animal’s current orientation - can explain a surprisingly broad array of results; and second, that the spatial information available to an animal is inextricably linked to its sensory systems and its movements within its environment.
Not only do they convincingly argue that cognitive mechanisms like those proposed in ants and bees might explain the way hummingbirds look for food but they raise the question of whether an insect-inspired approach might spark advances in our understanding of vertebrate navigation as a whole.
​
Schematic representations of local environmental space guide goal-directed navigation. Marchette SA, Ryan J, Epstein RA. Cognition 2017 01; 158:68-80
Many animals, including humans, need to remember the locations of places of interest in a way that allows navigation at different spatial scales. For example, to retrieve your spectacles you need to recall not only in what room you left them, but where in the room they can be found. An intuitively appealing idea might be that the brain stores spatial memories in a ‘zoomable’ map, so that the spatial coordinates of each known location are stored in a single reference frame and, by varying the resolution at which those coordinates are retrieved, we can use the same information to navigate towards something on both global and local scales.
The authors tested this hypothesis in a virtual reality environment by asking human participants to re-find the locations of objects displayed in four virtual ‘museums’. Imperfect human navigators demonstrated a striking pattern of errors in which they frequently went to the ‘right’ place within a room, but in the wrong building entirely, demonstrating that their memories of each object’s location within its museum could not be nested within the larger scale memories of the location of each building.
Further experiments showed that these ‘schema-preserving’ errors persisted even when the buildings differed in their local geometries and even in their shapes, casting doubt on the proposition that the local positions of objects were stored as Euclidean coordinates even within a local reference frame and suggesting instead a schematic encoding of locations relative to walls, doors, etc.
Intriguingly, when the buildings themselves were removed so that the objects were arranged within a single park-like environment, these errors disappeared, with participants almost always locating the correct grouping of objects, even if they sometimes found the wrong object within a group. This strongly suggests that spatial boundaries play an important role in determining the spatial scales at which location memories are grouped.
​
Male bumblebees perform learning flights on leaving a flower but not when leaving their nest. Robert T, Frasnelli E, Collett TS, Hempel de Ibarra N. J Exp Biol 2017 03 01; 220(Pt 5):930-937
When leaving their nest for the first time, worker bees perform characteristic looping flights, turning back to look at the nest. These learning flights help them to learn the location of the nest, which is essential to a central place forager like a worker bee. Male bumblebees of the species Bombus terrestris, on the other hand, live a very different lifestyle from the female workers: leaving their natal nest for the first time they never look back, metaphorically speaking; instead they spend the rest of their lives fending for themselves and searching for a mate. In this paper, Robert et al. used high speed cameras to investigate the structure of male flights when leaving the nest and demonstrate that the bees literally don’t look back and show no sign of female-like learning flight. After feeding from a rewarding artificial nectar source, however, male bees did perform learning flights, facing back toward the feeder in a similar manner to females. These results suggest that the behaviours bumblebees perform to facilitate learning are closely related to their ecological needs.
Continuous lateral oscillations as a core mechanism for taxis in Drosophila larvae. Wystrach A, Lagogiannis K, Webb B. eLife 2016 10 18; 5
Occam’s razor suggests that, given several competing hypotheses, the simplest should a priori be considered most likely. When it comes to understanding the neural mechanisms that underlie complex behaviours, though, it can often be hard to comprehend what the minimal requirements to reproduce a set of behaviours might be. Even an animal as simple as a Drosophila larva exhibits what appears to be a repertoire of different behavioural states that allow the larva to move up and down chemical, light or temperature gradients. Attempts to understand this behaviour have posited a system requiring some level of ‘decision making’ and switching between behavioural programs. In this study, Wystrach et al. examined high resolution video recordings of moving larvae and identified a regular lateral body oscillation. Both turns and head ‘casting’ behaviour could be explained as modulations in the size of these oscillations. Modelling this system revealed that a remarkable amount of the larva’s behavioural repertoire could be reproduced by a simple oscillating system, receiving direct sensory inputs without complicated processing. Thus, what appears to be a complex set of behaviours involving switches between distinct states might turn out to emerge naturally from a far simpler mechanism.
Prenatal acoustic communication programs offspring for high posthatching temperatures in a songbird. Mariette MM, Buchanan KL. Science 2016 08 19; 353(6301):812-814
Many songbirds call while incubating their eggs and can even communicate with their embryos in this way. In this fascinating study, Mariette and Buchanan discovered that zebra finches produce a particular incubation call only in the later stages of embryonic development and only when the ambient temperature is unusually high. Using a series of thorough experimental manipulations, they show that nestlings that were exposed to these calls before hatching showed different patterns of growth and begging than controls. Birds that had heard the incubation calls and were reared in hot nests grew more slowly than those reared in cooler nests, while control birds showed the opposite trend. The effects continued into adulthood, with birds that grew slowly in hot nests or fast in cool nests achieving greater reproductive success than their less well adapted counterparts. Finally, males that had been exposed to incubation calls and reared in hot nests were more likely themselves to choose hot nests in which to raise their own offspring. This all adds up to a very exciting story in which parent birds are able to communicate to their unhatched offspring regarding the environmental conditions under which they can expect to grow up and the offspring respond by altering their behaviour and developmental trajectory in order to thrive under those conditions.
Caffeinated forage tricks honeybees into increasing foraging and recruitment behaviors. Couvillon MJ, Al Toufailia H, Butterfield TM, Schrell F, Ratnieks FLW, Schürch R. Curr Biol 2015 Nov 02; 25(21):2815-2818
Over the course of a hard working day, many of us will look forward keenly to our next cup of coffee; this fascinating paper suggests that honeybees may feel the same way! Caffeine is a naturally occurring substance in many species of plants and is often found, at low concentrations, in nectar. To test whether caffeinated nectar affects the behaviour of insects that feed on it, Couvillon et al. monitored the foraging activity of honeybees trained to feed on artificial feeders in the field. When the sucrose rewards in the feeder were laced with a realistic concentration of caffeine, not only did individual bees visit the feeder more often, but they were more likely to communicate the feeder’s position to their nest mates using the waggle dance, dancing more enthusiastically and for longer. The end result was that recruitment of new bees to the caffeinated feeder was four times greater than to a control feeder. The authors suggest that the caffeine manipulated the bees’ perception of the nectar quality so that they react as though to nectar with a higher sugar concentration. Plants with caffeine in their nectar may be able to attract more bees, increasing the efficiency of pollination, without having to invest more in producing sweeter nectar. Not only does this study provide an intriguing insight into the ways in which plants can manipulate their insect pollinators, but for those of us who study waggle dances it also suggests a valuable practical method to increase dances.
Honeybees Learn Landscape Features during Exploratory Orientation Flights. Degen J, Kirbach A, Reiter L, Lehmann K ... Singh PK, Manz G, Greggers U, Menzel R. Curr Biol 2016 10 24; 26(20):2800-2804
Insect such as bees show remarkable abilities to navigate their environment, often over large distances, and technological advances are now allowing researchers to address important questions about what information they learn about the landscape and when they learn it (e.g. {1,2}). Honeybee workers taking up foraging for the first time usually perform a series of orientation flights, first circling the nest to learn its location, then flying further out into the landscape. Long-range orientation flights typically cover only a narrow sector of the surrounding environment, with consecutive flights often exploring in different directions, and it is unclear whether they serve the purpose of learning the area for navigational purposes, looking for forage locations or both. In this useful study, the authors used harmonic radar technology first to record exactly where individual bees flew during their orientation flights, and then to look at what they did when displaced either to locations they had just explored or to entirely novel locations. Bees that were displaced to areas they had explored during their orientation flights found their way home faster, taking straighter routes, than those that found themselves in novel locations, demonstrating that they had been learning how to get home during these flights. The authors argue that the bees' performance could not be explained by odour cues and that they take straight paths home from too far away to recognise landmarks associated with the nest, strongly suggesting that the bees memorise visual features of the landscape during orientation flights.
References
1. Ontogeny of orientation flight in the honeybee revealed by harmonic radar.
Capaldi EA, Smith AD, Osborne JL, Fahrbach SE, Farris SM, Reynolds DR, Edwards AS, Martin A, Robinson GE, Poppy GM, Riley JR. Nature. 2000 Feb 3; 403(6769):537-40 DOI: 10.1038/35000564
2. Life-Long Radar Tracking of Bumblebees.
Woodgate JL, Makinson JC, Lim KS, Reynolds AM, Chittka L. PLoS ONE. 2016; 11(8):e0160333 DOI: 10.1371/journal.pone.0160333
Impacts of neonicotinoid use on long-term population changes in wild bees in England. Woodcock BA, Isaac NJ, Bullock JM, Roy DB, Garthwaite DG, Crowe A, Pywell RF. Nat Commun 2016 08 16; 7:12459
The controversy regarding whether neonicotinoid pesticides are responsible for declines in pollinator populations is one of the most important and pressing questions facing both the ecology and agriculture communities today. There has been an EU moratorium on their use in place since 2013 to allow more time for research and a consensus seems to be gradually appearing that neonicotinoids can cause potentially serious, sub-lethal effects in bees and other insect pollinators. However, the short-term nature of most scientific studies had led to a serious lack of information about the impact these relatively subtle deleterious effects may have on insect population structures in the long-term, hampering our ability to make constructive decisions regarding pesticide use. This important and timely paper used long-term observational data of wild bee distributions throughout England to look at whether neonicotinoid usage played a role in population persistence and local extinctions over the last 18 years. The results demonstrate a negative relationship between neonicotinoid usage and the survival of populations on a local scale, across 62 wild bee species. This effect was three times more pronounced among species that are known to forage on oilseed rape (OSR), one of the major crops on which neonicotinoids are used and one of the major sources of forage for bees in the UK. The authors’ analysis suggests that that, while OSR provides important resources for bees of many species, the advantages of wide OSR coverage are outweighed by the negative impact of using neonicotinoid pesticides. While it remains unclear how these effects at a local scale will impact loss of biodiversity at larger spatial scales, it is clear that the sub-lethal effects attributed to neonicotinoid use have the potential to contribute to long-term population declines and this evidence has an important role to play in the ongoing debate over pesticide use.
Why do animals sleep? Studies in vertebrates, particularly mammals, have shown that the consolidation of memory is perhaps the most important function of sleep. Sleep occurs in a wide variety of taxa, but less is known about the extent to which the functions and mechanism of sleep in invertebrates might resemble those in vertebrates. The authors used a proboscis extension reflex (PER) conditioning task to quantify honeybees' memory of a heat stimulus. When bees were exposed during deep sleep to a contextual odour that had been present during training, their memory performance was improved the following day. A clever and logical series of further experiments demonstrated that this performance boost only occurred when the bees experienced the particular odour associated with the learning trials, and only when they were exposed to it in deep sleep, rather than at random points during the night. Past experiments in mammals have suggested that the consolidation of memory during sleep functions by reactivation of the same neural pathways that were active during the initial learning experience. The fact that replaying the learning context (in the form of an odour) boosts memory performance in bees hints that a similar process may occur in insects. This raises the interesting possibility that not only the function of sleep, but at least some of the mechanisms of memory consolidation are evolutionarily conserved from invertebrates to mammals.
Seed coating with a neonicotinoid insecticide negatively affects wild bees. Rundlöf M, Andersson GK, Bommarco R, Fries I ... Klatt BK, Pedersen TR, Yourstone J, Smith HG. Nature 2015 May 07; 521(7550):77-80
This paper provides convincing evidence that neonicotinoid pesticides have a harmful effect on bees, in an extensive field study. The authors looked at a number of indicators of bee health across eight pairs of oil-seed rape fields in southern Sweden, in which one field of each pair was treated with a neonicotinoid-based commercial seed treatment. They present evidence for reduced density of wild bees, impaired colony growth and reproduction in bumblebee colonies, and reduced nesting success by solitary bees in treated fields. This study makes weighty contributions to our understanding in two key areas: first, some of the existing evidence for deleterious effects of neonicotinoids comes from laboratory studies in which the dosage or route of ingestion of neonicotinoids may not be realistic, and the bees have no opportunity to manage their exposure through their behaviour (for example, by foraging elsewhere). This study, by contrast, used a commercially available seed coating on an oil-seed rape crop, at the manufacturer's recommended concentration and on fields that were managed normally, ensuring that we get a true picture of the effects of neonicotinoids as used in agriculture. Second, many prior studies and risk assessments have used honeybees, Apis mellifera, as a model, so that less is known about the effects of neonicotinoids on the wider pollinator community. Here, the authors expanded the range of species examined, looking at the effects of pesticide treatment on bumblebees and solitary bees, in addition to honeybees. They provide evidence that wild bee numbers were reduced in treated fields and that the growth of bumblebee colonies was impaired. However, no treatment effect was seen on the health of honeybee colonies, suggesting that the effects of neonicotinoid use on wild pollinator populations may be greater than previously thought and implying that honeybee studies cannot be reliably extrapolated to other species.