Insect Navigation: Neural Basis to Behavior
Summary and Keywords
Navigation is the ability of animals to move through their environment in a planned manner. Different from directed but reflex-driven movements, it involves the comparison of the animal’s current heading with its intended heading (i.e., the goal direction). When the two angles don’t match, a compensatory steering movement must be initiated. This basic scenario can be described as an elementary navigational decision. Many elementary decisions chained together in specific ways form a coherent navigational strategy. With respect to navigational goals, there are four main forms of navigation: explorative navigation (exploring the environment for food, mates, shelter, etc.); homing (returning to a nest); straight-line orientation (getting away from a central place in a straight line); and long-distance migration (seasonal long-range movements to a location such as an overwintering place). The homing behavior of ants and bees has been examined in the most detail. These insects use several strategies to return to their nest after foraging, including path integration, route following, and, potentially, even exploit internal maps. Independent of the strategy used, insects can use global sensory information (e.g., skylight cues), local cues (e.g., visual panorama), and idiothetic (i.e., internal, self-generated) cues to obtain information about their current and intended headings.
How are these processes controlled by the insect brain? While many unanswered questions remain, much progress has been made in recent years in understanding the neural basis of insect navigation. Neural pathways encoding polarized light information (a global navigational cue) target a brain region called the central complex, which is also involved in movement control and steering. Being thus placed at the interface of sensory information processing and motor control, this region has received much attention recently and emerged as the navigational “heart” of the insect brain. It houses an ordered array of head-direction cells that use a wide range of sensory information to encode the current heading of the animal. At the same time, it receives information about the movement speed of the animal and thus is suited to compute the home vector for path integration. With the help of neurons following highly stereotypical projection patterns, the central complex theoretically can perform the comparison of current and intended heading that underlies most navigation processes. Examining the detailed neural circuits responsible for head-direction coding, intended heading representation, and steering initiation in this brain area will likely lead to a solid understanding of the neural basis of insect navigation in the years to come.
With few exceptions, animals purposefully relocate their bodies between different places in their environment, a process that we define as navigation. This ability is what distinguishes them from other macroscopic organisms on Earth. They navigate their surroundings to find food, to avoid being eaten, to locate reproductive partners, and to escape unfavorable environmental conditions. What does it ultimately take to navigate? Animals have to know where they want to go (i.e., the navigational goal) and where they currently are in relation to that goal. This means that in each moment in time, they have to know their desired heading and their current heading. By comparing the two, they can generate steering commands to compensate for any mismatch. This comparison distinguishes all forms of navigational behaviors from short-term, reflex-driven behavior, such as escape jumps, flight stabilization movements, and obstacle avoidance. Even though these latter forms of movement also can be directed, they do not require matching a current body orientation with a long-term-goal direction.
To acquire information about goals and directions, an animal needs adequate sensory organs and a brain equipped with the processing power to extract meaningful information from their continuous output. Then, this information is used to initiate steering by sending signals to the muscles required to set the body in motion. But navigational decisions are driven by more than just immediate sensory input; at each moment in time, the animal’s intended heading will be defined by behavioral context, motivational context, and previous experience, all of which will lead to a reevaluation of sensory signals in order to make the next steering decision.
While this is true for essentially all animals, insects provide uniquely broad access to these questions at many levels. They are easily manipulated for behavioral experiments, and with their small, tractable brains combined with a range of complex navigation behaviors carried out in vastly different sensory environments, insects also have proved to be ideal for the study of the neural basis of navigation.
Before addressing how insect brains transform sensory information into navigation behavior, we must understand what sensory information these animals use, as well as what behavioral strategies they employ.
Available Sensory Information
Sensory information is key to keeping track of one’s own heading, as well as to identifying any desired heading. Which information do insects use? Although the specific answer to this question depends on the sensory environment, activity period, and behavioral strategy of each species, navigation behavior can exploit global or local sensory cues in principle.
Global cues provide an overall reference frame, which can aid all forms of navigation and is independent of the movements of the animal within its environment. Visual celestial cues, as well as the Earth’s magnetic field, are the main sources of global information (Figure 1). During the day, the disk of the Sun and several features of scattered sunlight in the sky are the most prominent cues. Scattering of sunlight produces the pattern of polarized skylight, the spectral gradient between solar and antisolar sky hemisphere, and the skylight intensity gradient, all of which contain directional information even if the Sun is not directly visible (Brines & Gould, 1982; Heinze, 2015a). During twilight conditions, when the Sun is just below the horizon, these cues are in fact the most reliable sources of compass information. For all Sun-derived cues, the animal must compensate for the daily apparent movement of the Sun across the sky, at least for behaviors during which it is necessary to maintain a consistent reference frame for extended periods of time.
The situation during the night is more complex. In principle, all scattered-light cues are present around the Moon in the same way as around the Sun, albeit several orders of magnitude dimmer. However, whereas the Sun follows the same apparent trajectory across the sky every day, the apparent movements of the Moon involve not only daily changes in position due to the Earth’s rotation, but also monthly changes in its trajectory due to the revolution of the Moon around the Earth. Moreover, the phases of the Moon lead to drastic variations in brightness, severely affecting the visibility not only of the lunar disk, but also of all Moon-derived scattered light cues.
Similar to the movement of the Sun, the stars follow a reliable daily trajectory. However, most stars are very dim, and it remains to be shown whether insects are able to resolve individual stars in order to use them for navigation. The final visual cue that is available to nocturnal animals is the bright streak of the galactic disk, the Milky Way (Figure 1C). While this elongated cue moves across the sky alongside all other stars, it delivers useful compass information only when it intersects with the horizon at steep angles.
A global cue that is independent of the Earth’s rotation is its magnetic field (Figure 1F). Four magnetic-field parameters can be used for navigation: the inclination of the field lines (i.e., the angle between the field lines and the horizon); the declination of the field lines (i.e., the angular difference between geomagnetic North and geographic North); the polarity of the field lines (i.e., the direction of North/South); and the strength of the field. These cues provide a reliable directional and positional global reference (Lohmann, Lohmann, & Putman, 2007). Even though there is growing evidence that insects, like birds, fish, and turtles, can use the magnetic field, the mechanism for magnetic field sensing remains unknown in any species (Clites & Pierce, 2017; Gould, 2010; Reppert, Guerra, & Merlin, 2016). Interestingly, the most plausible hypothesis for insects, the radical-pair mechanism mediated by cryptochrome proteins (Liedvogel & Mouritsen, 2010; Ritz et al., 2000; Hore & Mouritsen, 2016), implies that animals detect the Earth’s magnetic field using light-dependent processes and in fact might perceive it as a visual feature of the environment.
Local sensory cues can be external (allothetic; Figure 2A–E)—that is, independent of the animal’s movement—or self-generated (idiothetic; Figure 2F–H). Visual and olfactory landmarks, as well as stable winds, are important external local cues useful for navigation. The entire panorama (i.e., the sum of all visual objects along the horizon against the uniformity of the sky) is another powerful input for navigation systems. Even at low resolution, it provides the viewer with a unique signature of a particular location within an animal’s habitat (Ardin, Peng, Mangan, Lagogiannis, & Webb, 2016; Zeil, 2012). Finally, goal-related sensory information is vital during explorative behavior. Food odors, flower colors, female pheromones, or courtship songs all provide the navigator with multisensory, directional information essential to determining its desired heading.
At the same time, self-generated cues deliver information about the animal’s heading and relative position in space by continuously providing readouts of body rotations, leg movements, and forward speed (Figure 2F–H). Rotational optic flow across the retina (when moving through the environment), as well as proprioceptive signals (e.g., from leg joints), are the most prominent examples that insects use to keep track of body orientation. While also used for flight stabilization reflexes (Krapp & Hengstenberg, 1996), these cues provide important input to angular path integrators—that is, they allow an insect to integrate its body rotations over time in order to generate an estimate of head direction without the immediate need to calibrate that signal to an external reference (Green et al., 2017; Turner-Evans et al., 2017). Similarly, step counting and translational optic flow sensing have been shown to serve as distance measures (odometers) in ants and bees, respectively (Srinivasan, 2014; Wittlinger, Wehner, & Wolf, 2006). These angular and odometer signals combined can, at least over very short distances, comprise all the input necessary to establish an insect’s angle and position in space during and after a navigational bout, independent of concrete external cues, but at the cost of significant error accumulation and growing uncertainty over time (Cheung, 2014).
Overall, insects use many sources of sensory information for navigation. The question of which cues are best suited for any particular behavior depends on the behavior’s spatial scale, its goal, the environment that it is performed in, whether it is carried out during the day or night, and the sensory abilities of each species.
Navigational Goals and Strategies
After the sensory basis for navigation is outlined, the question arises of what this information is used for—that is, which behavioral strategies use the described cues? Essentially, four different modes of navigation can be distinguished, as discussed in the next sections.
In its simplest form, navigation comprises moving away from a point in space as fast as possible. This is achieved most efficiently by moving in a straight line. The “goal” of navigation in this case is merely to hold a steady direction (Figure 3A). Importantly, and different from momentary escape behavior, this direction, once it is chosen, is maintained for an extended period of time until the motivational state of the animal changes and the navigational bout is replaced by another behavior. A prime example of this straight-line navigation is performed by ball-rolling dung beetles (el Jundi et al., 2016). These insects locate a dung pat and carve out a ball of dung. To avoid the fierce competition and the danger of having its food stolen by other beetles, a beetle rolls the dung ball away from the dung pile as fast as possible (i.e., in a straight line). The direction is chosen randomly, as no single direction has an a priori advantage over any other one. However, once on its way, the beetle will maintain its heading for a significant amount of time and reembark on the same trajectory if disturbed. A long series of behavioral experiments have shown that these animals use exclusively global celestial cues for achieving this behavior, including the Sun (Dacke, el Jundi, Smolka, Byrne, & Baird, 2014); the skylight polarization pattern (Dacke, Nilsson, Scholtz, Byrne, & Warrant, 2003); the Moon (el Jundi et al., 2015b); the spectral gradient (el Jundi, Foster, Byrne, Baird, & Dacke, 2015a); the intensity gradient (el Jundi, Smolka, Baird, Byrne, & Dacke, 2014b); and the Milky Way (Dacke, Baird, Byrne, Scholtz, & Warrant, 2013; Foster et al., 2017).
Maybe the most intuitive mode of navigation can be observed before the beetle reaches the dung pile. It will fly across its habitat in search of food (Warrant & Dacke, 2016). This presumably occurs only in the correct motivational and environmental context (i.e., if the beetle is hungry and the weather conditions are favorable). This search for food is a behavior that all insects exhibit and is one of the fundamental features of animal behavior in general. Whether for a male silk-moth locating a female, a desert ant during its hunt for insect carcasses in a featureless environment, or a fly searching for a piece of fruit in the undergrowth of a forest, the search strategy involves maintaining a steady heading for some time until either a food-related sensory signal is perceived or an internal driving force changes the motivational state. In both cases, the animal will turn and either pursue the food item or begin the search anew. During these straight movement bouts, the desired heading can be chosen randomly (as in the dung beetle) or correspond to the source of the goal-related sensory input.
In principle, all animals have to perform these behaviors. Species will differ only in the spatial range of movements, the sensory information used, and the types of information that signal a goal. The key is to maintain an internally selected direction that is tethered to some feature of the external world. This feature can identify the goal itself or be a feature of the world in the line of sight, when searching for a goal-related sensory signal (e.g., a tree kept stable in the frontal visual field). Fixation of a prominent vertical stripe during flight and walking behavior in Drosophila are typical examples of how features of the environment are used as transient targets, allowing a steady heading to be maintained (Neuser, Triphan, Mronz, Poeck, & Strauss, 2008; Reiser & Dickinson, 2010; Strauss & Pichler, 1998).
As opposed to dung beetles, which dig a new home each time they have escaped to a favorable spot in the savannah with their fresh dung ball, or butterflies pursuing one flower after another, some insects have a fixed nest to return to after foraging. In particular, hymenopteran insects (i.e., bees, ants and wasps) are well-studied central-place foragers (Webb & Wystrach, 2016). While food-related stimuli are the driving force during their outbound foraging trip, the home location determines the desired heading during the homeward journey. Radar tracking and other behavioral observations have shown that many species (e.g., desert ants, bumblebees, and honeybees) leave the nest for convoluted foraging trips but can return home in straight lines (Srinivasan, 2015). These insects use an integrated version of their outbound journey as a home vector (path integration; Figure 3C). Essentially, they keep track of all directional changes occurring during the foraging trip using global celestial cues (Evangelista, Kraft, Dacke, Labhart, & Srinivasan, 2014), and combine these with the distance covered by integrating the flight speed over time, measured by translational optic flow (Srinivasan, 2014). Other species (e.g., bull ants or desert ants living in visually rich habitats) follow fixed, memorized routes on both the outbound and inbound trips and use image-matching to identify maximally familiar views of the panorama acquired during the outbound trips (Graham & Cheng, 2009; Zeil, 2012) (Figure 3D). Yet others (e.g., honeybees) use landmarks (identified objects extracted from the visual panorama; Figure 3E) (Collett, 2010; Degen et al., 2016), or possibly even internal maps, to locate food sources and the nest (Cheeseman et al., 2014b; Menzel et al., 2005). While the latter strategy is highly controversial (Cheeseman et al., 2014a; Cheung et al., 2014), if it indeed exists in insects, it could employ a combination of stored vectors associated with landmarks in the habitat for finding novel, optimal paths to the nest (“true navigation”; i.e., using a map and a compass).
The desired heading during homing corresponds to a direction indicated by a global cue during path integration (e.g., flying toward the Sun), or to the optimal orientation of a panoramic image/landmark to which the current visual input is compared to during route following. Home is indicated by the end of the used vector, or when the current view optimally matches the last stored image of the sequence along a homing route. The precise location and identity of the nest entrance are then verified by a combination of multisensory cues, many of which are acquired during orientation flights/walks performed when leaving the nest for the first time (Collett, de Ibarra, Riabinina, & Philippides, 2013b; Stürzl, Zeil, Boeddeker, & Hemmi, 2016; Zeil, 2012).
Whereas homing is a challenging navigation behavior, particularly in complex environments, some insects follow an even more astounding strategy: long-distance migration (Figure 3B). Once per year, these insects fly thousands of kilometers across unfamiliar terrain to escape unfavorable conditions in their breeding grounds (Homberg, 2015; Merlin, Heinze, & Reppert, 2012; Warrant et al., 2016). This mass migration takes these species to specific regions of the planet with conditions that ensure their survival, from where they return to their point of origin several months later. The target regions can be broad geographical areas or be as specific as a set of selected mountain caves (Heinze & Warrant, 2016; Warrant et al., 2016). The complete journey is covered either by the same individuals (e.g., Bogong moths) or over the course of several generations (e.g., monarch butterflies).
The trip can be broken into two consecutive parts: the long-distance flight and the homing-in on the target site. For the first segment, migratory insects must possess a reliable compass that precisely encodes the current flight bearing. This heading has to be continuously compared to their desired migratory direction over the course of many days or weeks. In essence, long-range navigation can be viewed as a form of straight-line orientation, but at a much larger scale, with the chosen direction being identical across all individuals of a species.
As no individual insect performs the migration more than once, the desired heading has to be genetically fixed and thus can be expected to be hardwired into the brain, yet in a way that can be reversed between seasons to ensure the completion of the full migration cycle (Guerra & Reppert, 2013). The second segment of migration, the location of the precise target site, is likely more similar to exploratory navigation and, even though virtually nothing is known about it, can be expected to utilize many cues from multiple sensory modalities. In many ways, the transition between the two stages seems similar to starting a search behavior at the end of a path integration vector. In both cases, the motivational state of the insect changes, switching from the straight, long navigation segment to a systematic search behavior.
Which cues are used during long-distance migration? In principle, as the desired heading has to be stable over long periods of time during migration, the compass cues that identify the heading have to be as stable as possible. Thus, global reference frames are used during migration (Merlin et al., 2012). During the day, the Sun and its surrounding pattern of polarized light are ideal cues that are used as reference, once their daily motion across the sky is accounted for. Most clearly, this has been demonstrated with behavioral experiments in monarch butterflies using flight simulators in which the animals can freely choose their flight direction while viewing the open sky (Merlin et al., 2012; Mouritsen & Frost, 2002). Whereas the Sun serves as the main decisive factor in choosing the flight direction during the monarch butterfly migration (Stalleicken et al., 2005), polarized light also can serve as a guiding cue when the Sun is not visible (Reppert, Zhu, & White, 2004).
In addition, indoor experiments have revealed that monarch butterflies can use magnetic fields to choose their flight direction (Guerra, Gegear, & Reppert, 2014). The cues that are used by this species during migration in a natural setting remain to be determined. Similarly, indoor experiments with locusts in a flight tunnel revealed that these insects can use polarized light to adjust their flight bearing (Mappes & Homberg, 2004), but whether they use this cue during their migrations is unclear (Homberg, 2015).
Interestingly, the antennae appear to play a major role in enabling monarch butterflies to choose the correct flight direction over the course of the day (Merlin, Gegear, & Reppert, 2009). Manipulations of the antennae, either by clipping or painting, has revealed that circadian clocks located in the antennae are essential to the monarch butterfly’s ability to compensate for the apparent movement of the Sun across the sky during the course of the day (Guerra, Merlin, Gegear, & Reppert, 2012; Merlin et al., 2009). In addition, the maintenance of this process depends on light-entrainment of the antennal clocks. The molecules involved in this task are likely specific blue-light receptors called cryptochromes, which, interestingly, also are proposed to be likely sensors of magnetic fields (Reppert et al., 2016). In line with this idea, the antennae of the monarch also were required for using magnetic fields to choose flight directions (Guerra et al., 2014).
Neural Principles of Insect Navigation
How does the insect brain drive the outlined navigation behaviors? Their small size and accessibility, together with recent advances in functional imaging and genetic access to identified neurons, have allowed substantial progress in understanding the neural mechanisms underlying spatial orientation. The main focus so far has been threefold: first, illuminating the neural mechanisms at the basis of processing navigation-relevant sensory information; second, the mechanisms of central integration of navigation-relevant information and the resulting encoding of heading direction; third, the generation of steering commands for navigation.
Almost every stimulus can be used for navigation, but very few cues are used exclusively for this purpose. While any object in the environment is of potential interest as a target for explorative navigation, global sensory cues are likely used only for navigation. Much research on the neural basis of navigation, thus, has focused on the processing principles of these specialized cues.
One exquisite example is the neural circuit underlying polarized-light processing (Heinze, 2014; Homberg, Heinze, Pfeiffer, Kinoshita, & el Jundi, 2011) (Figure 4). Neurons involved in processing polarized light respond to a polarization filter that is rotated above the animal’s head with sinusoidal modulations of their action potential frequency (Heinze, 2014; Labhart, 1988). They possess a preferred angle of polarization—that is, a specific orientation of the electric field (E)–vector, at which their neural activity is highest. At the orthogonal E-vector angle, the neuron is least active.
If a polarization-sensitive neuron (POL neuron) is excited at its preferred E-vector but inhibited at the antipreferred E-vector, it is referred to as exhibiting polarization opponency. This feature is typical for many POL neurons and hints at how polarized-light information is analyzed by the visual periphery. Here, photoreceptors of a specialized part of the compound eye (the dorsal rim area, or DRA) are tuned to specific E-vectors (for mechanisms, see, e.g., Heinze, 2014) and are arranged in pairs with orthogonal tuning within each ommatidium (Homberg & Paech, 2002; Labhart & Meyer, 1999; Stalleicken, Labhart, & Mouritsen, 2006; Weir et al., 2016). Their output is thought to be antagonistically combined in second-order neurons. This arrangement ensures that the POL channel is insensitive to changes in brightness of the incoming light (Labhart, 2016). In addition, both receptor types typically share identical photopigments and are thus sensitive to the same wavelengths of light (in most insects in the ultraviolet range, and blue light in crickets and locusts) (Heinze, 2014; Labhart & Meyer, 1999). This feature ensures that the spectral composition of light does not interfere with the detection of polarized light. Thus, in principle, the POL pathway of insects selectively encodes the angle of polarized light presented to the animal (for a detailed discussion of polarized-light detection versus E-vector perception, see Labhart, 2016).
The first processing stage of visual information in the insect brain, the optic lobes, contains numerous POL neurons, which were investigated first in crickets (Labhart, 1988; Labhart & Meyer, 2002), and more recently in locusts (el Jundi, Pfeiffer, & Homberg, 2011). Individual examples of similar neurons also have been recorded in cockroaches (Loesel & Homberg, 2001) and ants (Labhart, 2000). The best-examined cell type interconnects the second optic neuropil, the medulla, between both brain hemispheres.
In crickets, these neurons (POL1-neurons) have very large receptive fields—that is, they pool input from many photoreceptors and are tuned to one of three E-vector angles, set apart by about 60 degrees (Labhart, Petzold, & Helbling, 2001). This arrangement thus provides three analyzer channels, which, when their relative activity is combined, allows for unambiguously reconstructing any E-vector presented to the animal (Labhart, 2016). Because of the large receptive fields of these cells (with diameters of more than 60 degrees), this population-encoded E-vector corresponds to the average E-vector present in the sky viewed by these neurons and might be used to feed into a head-direction system that uses celestial E-vectors.
The general significance of the three-angle code of the cricket POL1 neurons is unclear, however, for two reasons: first, locust POL neurons of the optic lobe do not share this feature (Heinze, 2014; el Jundi et al., 2011); and second, the downstream target cells of POL1 cells are not known. Rather, transmedulla neurons that innervate the medulla-DRA and provide output to a region of the central brain called the anterior optic tubercle (AOTU) could perform the task of sending information to the central brain (Heinze, 2014). Their likely role in polarized-light processing is supported by the finding that these cells innervate a layer of the medulla that is shared by all POL neurons of the optic lobe (el Jundi et al., 2011). However, owing to their small fiber diameter, these cells have not been functionally characterized to date, and the connections to the remaining members of the optic lobe POL network remain to be illuminated.
POL neurons in the AOTU have been investigated thoroughly in locusts (Homberg et al., 2011; el Jundi & Homberg, 2012; Pfeiffer & Homberg, 2007; Pfeiffer, Kinoshita, & Homberg, 2005). The four known AOTU cell types that respond to E-vectors all innervate the lower subunit of this brain region, a finding that was confirmed in monarch butterflies as well (Heinze & Reppert, 2011). Considering the more than 300 million years of separate evolutionary history between those two species, the neurons passing through this brain area appear to constitute a general pathway for processing and relaying polarized-light information before it reaches higher integration centers of the insect brain.
The target of the AOTU projection neurons (TULAL1) are the lateral complexes of the central brain (LX). Here, a region called the bulb contains extremely large synapses that link the AOTU projection neurons with the input neurons (TL neurons) of the central complex (CX) (Heinze, Florman, Asokaraj, el Jundi, & Reppert, 2013; Held et al., 2016; Träger, Wagner, Bausenwein, & Homberg, 2008). TL neurons in locusts mirror the skylight polarization pattern with the E-vector tuning directions across their wide receptive fields, suggesting a role as matched filters of the natural sky that are maximally activated by specific orientations of the animal relative to the sky (Bech, Homberg, & Pfeiffer, 2014). These neurons have their output terminals in the CX, the final target of compass information. This region is a conglomerate of four neuropils: the upper and lower divisions of the central body (CBU, CBL), the protocerebral bridge (PB), and the paired noduli (NO) (Homberg, 2008; Pfeiffer & Homberg, 2014) (Figure 4C–F).
At least two parallel pathways starting in the optic lobe and ending in distinct layers of the CBL have been reported in several species (Heinze, 2014; Omoto et al., 2017). Two of these pathways carry information about polarized light in both locusts and monarch butterflies, a conservation that suggests functional significance of this parallel processing. Polarization-sensitive neurons at the input stage of the CX also have been reported in crickets (Sakura, Lambrinos, & Labhart, 2008), dung beetles (el Jundi et al., 2015b), and solitary bees (Stone et al., 2017), demonstrating that celestial compass information is represented in this integration center of the brain across many species. This suggests that the CX plays a key role in processing navigation-relevant information.
Other Celestial Cues
The role that the CX plays in processing navigation-relevant visual information is supported by the finding that POL neurons in this region and upstream of it not only respond to polarized light, but also are tuned to specific azimuths of bright, unpolarized objects moving around the animal (el Jundi, Pfeiffer, Heinze, & Homberg, 2014a). These experiments were performed in locusts (Pfeiffer & Homberg, 2007), monarch butterflies (Heinze & Reppert, 2011), and dung beetles (el Jundi et al., 2015b), using a bright light-emitting diode (LED) on a circular path around the animals (at a constant elevation) to elicit neural responses. Elegant behavioral experiments in dung beetles have shown that the used LEDs are indeed interpreted as a celestial body when these animals decide on a direction to roll a dung ball (el Jundi et al., 2015b). In general, POL neurons combine both the polarization information of the blue sky and the solar (or lunar) azimuth and thus encode an integrated signal representing body orientation relative to a global reference.
Where are polarized light information and unpolarized azimuth information first combined along the POL pathway? Already in the optic lobes of locusts, POL neurons respond to the azimuth of a light spot moving around the animals (el Jundi et al., 2011). Neurons suited to perform this integration are the aforementioned transmedulla neurons. They innervate the DRA of the medulla but also have a single, long dendrite that passes through the medulla from top to bottom; that is, in the retinotopic arrangement of the medulla, this dendrite could sample information from ommatidia that point to one specific azimuth, independent of the elevation of the light source (el Jundi et al., 2014a). Unfortunately, no functional data exist from these cells, and this tempting hypothesis remains to be proved in the future.
All neurons downstream of the optic lobe (i.e., POL-neurons of the AOTU, the bulbs, and the central complex) respond to both sets of stimuli in locusts, monarch butterflies, and dung beetles in highly similar ways (el Jundi et al., 2014a, 2015b). In most cases, the azimuth response is independent of the spectral composition of the stimulus. The one exception are neurons of the locust AOTU, in which the tuning to ultraviolet light is opposite to the tuning to green light (Kinoshita, Pfeiffer, & Homberg, 2007; Pfeiffer & Homberg, 2007). This spatial opponency theoretically enables the locust to exploit the spectral gradient of the sky and thus disambiguate the axial symmetry of the polarization pattern. As the ratio of green to ultraviolet light differs in the solar and antisolar sky hemispheres, these neurons would be differentially activated when the animal faces in opposite directions, even if the E-vector information that they receive is identical.
There is one problem when information on solar azimuth and E-vector is used at the same time to derive a compass bearing: Only when the animal looks exactly at the zenith the two stimuli have a fixed relation to one another throughout the day (a 90-degree difference angle) (Pfeiffer & Homberg, 2007). However, no DRAs of insects studied so far do that, but they receive information from lower elevations as well (Heinze, 2014; Labhart & Meyer, 1999). In these regions of the receptive field, the difference angle between the solar azimuth and the E-vector orientation depends on the elevation of the Sun (as the E-vectors are arranged in concentric circles around the Sun). Thus, the two sensory cues would provide different estimates of solar azimuth over the course of the day if a fixed, 90-degree relation is assumed.
Locusts and butterflies solve this problem by adjusting the E-vector tuning of POL neurons in the AOTU in a daytime-dependent manner (Heinze & Reppert, 2011; Pfeiffer & Homberg, 2007). They change the values in exactly the way that is needed to compensate for the changing difference angle between the solar azimuth and the E-vector angle in the center of the DRA receptive field. This form of time compensation is called elevation compensation, and it ensures that higher-order brain regions receive a consistent heading estimate independent of the source of information. This underlines that this pathway indeed carries a head-direction signal, not explicit information about the polarization pattern of the sky or other polarized-light features of the environment (Labhart, 2016). Similarly, in two species of dung beetles, neurons in this pathway switch between encoding a celestial body (Sun or Moon) to encoding the skylight polarization pattern, thereby matching the cues used to drive orientation behavior in a specific navigational context (el Jundi et al., 2015b). Independent of the species, insects thus appear to utilize the most behaviorally relevant and reliable sensory information to relay body-orientation signals to central decision-making centers.
Local Visual Cues and Landmarks
Insects that rely on landmarks and the visual panorama for navigation are mostly hymenopteran insects (ants, bees, and wasps). The brains of these insects exhibit a prominent neural pathway from the optic lobe to the mushroom body that is much less pronounced or absent in other insects, as well as in ants that do not perform visual navigation (Fahrbach, 2006; Kühn-Bühlmann & Wehner, 2006; Paulk & Gronenberg, 2008). This suggests that visual information that has to be memorized to learn route landmarks is sent to the mushroom body for storage and later retrieval. Indeed, a model of the ant mushroom body has shown that even with conservative estimates of neural numbers, this structure could store hundreds of panoramic images needed for route navigation (Ardin et al., 2016). Which information is sent to the mushroom body (i.e., which aspects of the visual scene are extracted and stored) remains to be shown.
Local visual cues are also important during navigation strategies other than route following and landmark orientation. During explorative navigation, features of the environment are used as temporary navigation targets without the need to memorize them, such as is shown by stripe fixation in flies (Neuser et al., 2008; Reiser & Dickinson, 2010; Strauss & Heisenberg, 1993). Where does this information converge with the outlined compass pathways? The answer is not yet known, but anatomical and physiological data have demonstrated a visual, color-sensitive pathway through the upper division of the AOTU in parallel to the compass pathway, which ends in the lateral accessory lobes, regions closely associated with the CX (Mota, Gronenberg, Giurfa, & Sandoz, 2013; Pfeiffer & Kinoshita, 2012; Pfeiffer et al., 2005; Zeller et al., 2015). In addition, most output from the optic lobe is directed toward the lateral protocerebrum, either the anterior parts or the posterior parts (Otsuna & Ito, 2006; Paulk, Phillips-Portillo, Dacks, Fellous, & Gronenberg, 2008; Strausfeld & Okamura, 2007; Wu et al., 2016). Across insects, these regions house more or less pronounced optic glomeruli (Strausfeld & Okamura, 2007), and neurons respond to both motion and color stimuli (Aptekar, Keleş, Lu, Zolotova, & Frye, 2015; Okamura & Strausfeld, 2007; Paulk et al., 2008). While some large neurons that descend to motor control centers receive input from these regions and likely drive direct behavioral responses (e.g., escape reflexes) (Strausfeld & Okamura, 2007), the majority of output cells from these areas are unknown and likely project to largely uncharted territories of the insect brain. How this information is processed, and if or where it converges with compass cues, is unknown.
One possible integration site is suggested by anatomical and functional data: the CBU of the CX. Input neurons of this region could constitute an “object pathway” that carries information about behaviorally relevant sensory information from many brain regions to specific layers of the CBU. Neurons in flies that innervate this region (called the fan-shaped body in flies) are involved in encoding behaviorally relevant visual objects (Liu et al., 2006) and, given that similar cells exist across insects (Heinze et al., 2013; Homberg, 1985; el Jundi et al., 2009; Phillips-Portillo & Strausfeld, 2012), it seems possible that this is their general function.
Another suggestive finding pointing in this direction is that the variability in the morphology of these cells is comparably high across insect species, which is in contrast to the remaining parts of the CX. It is in line with the idea that these cells carry information about behaviorally relevant aspects of the environment. As these will differ between species, high variability is expected to account for the species-specific navigational targets.
The Earth’s Magnetic Field
How the Earth’s magnetic field is perceived and processed in the brain is one of the biggest enigmas in neuroethology. There is virtually nothing known about mechanisms. Several hypotheses exist about how magnetic fields could be perceived by animals, but solid data about potential receptors are lacking (Clites & Pierce, 2017). Arguably, the most likely mechanism is the radical pair theory, in which photosensitive molecules are activated by light but yield different reaction products depending on their orientation with respect to the magnetic field (Ritz, Adem, & Schulten, 2000). A promising candidate is the blue light receptor cryptochrome (Hore & Mouritsen, 2016; Liedvogel & Mouritsen, 2010). If such sensors are arranged in a fixed, three-dimensional array, the orientation of the field lines can be extracted. As this process is light dependent and the photoreceptors of the eyes contain tightly packed, three-dimensional membrane stacks, the eyes have been suggested as possible locations of magnetosensors (Ritz, Ahmad, Mouritsen, Wiltschko, & Wiltschko, 2010). This would imply that animals, including insects, perceive the magnetic field as a visual feature of the world and might be able to use existing visual pathways to process this information.
However, the magnetosensor in theory could be located anywhere in the body, and data from monarch butterflies show that in this species, the antennae are required for mediating responses to magnetic fields in an indoor flight simulator (Guerra et al., 2014). This effect was also light dependent and suggests the antennae as possible alternative locations of a cryptochrome-based magnetosensor. It can be speculated that the sensory-processing pathway might be using existing pathways for olfactory or pheromone processing (see the discussion later in this article about steering responses in the context of pheromone tracking in moths). Overall, understanding this aspect of sensing navigation-relevant information is still in its infancy. Insects, however, with their easily manipulated behavior and miniature brains, could provide the long-sought model needed to solve the questions of how magnetic fields are used for navigation (Guerra & Reppert, 2015; Warrant et al., 2016).
All sensory information relevant for navigation must converge in higher regions of the brain for two purposes: first, robust encoding of the current heading of the animal; and second, encoding of the desired heading based on sensory information that is combined with previous experience and internal state. Both must be compared to produce compensatory motor-steering commands in case the two directions do not match. In recent years, the CX has emerged as the likely site of navigational control; thus, it will be the focus of the next sections.
Encoding of the Current Heading
Using neurons responding to polarized light as a beacon to identify navigation-relevant brain regions, the polarization vision pathway, originating in the optic lobe, revealed the neuropils of the CX in the central brain as the final destination of this information (Homberg et al., 2011). All CX compartments are tightly interconnected by highly ordered arrays of neurons. Columnar neurons, innervating small regions (slices) of one or several compartments, produce a regular columnar organization of the CX neuropils that consists of 16 to 18 vertical slices (columns). These are intersected by horizontal layers in the CBU and the CBL, giving the CX its characteristic, highly ordered neuroarchitecture (Homberg, 2008; Pfeiffer & Homberg, 2014).
Within the CX, a multitude of POL-neurons were identified and shown to be functionally conserved in locusts (Homberg et al., 2011), butterflies (Heinze & Reppert, 2011), beetles (el Jundi et al., 2015b), and bees (Stone et al., 2017) (Figure 5A). Together, they define a neural network that is proposed to transform purely sensory compass signals into premotor commands suited to guide navigation (Heinze, 2014; Homberg et al., 2011). The two parallel pathways from the bulbs reach distinct layers of the CBL via different types of tangential neurons (TL2 and TL3; homologous to Drosophila ring neurons) (Vitzthum, Müller, & Homberg, 2002). The functional significance of this arrangement is not yet resolved, but several differences in the physiology of the tangential neurons in either pathway suggest nonredundant roles. In locusts, the cells in one pathway receive input from only one eye, while the other one is binocular (Heinze et al., 2009; Vitzthum et al., 2002). Similarly, in Drosophila, one pathway has wide receptive fields, while the other receives more focused input, albeit not with respect to polarized light but in response to motion stimuli (Omoto et al., 2017).
After this input stage, columnar neurons of the CBL (CL1a cells, homologous to Drosophila E-PG cells) likely receive compass information from TL-neurons and pass it to the PB. Importantly, in the PB, each POL-neuron’s tuning correlates with its anatomical position within this structure (Heinze & Homberg, 2007). As the polarization angle of skylight directly relates to the Sun’s azimuth, this arrangement essentially resembles an array of head-direction cells within a sky-based reference frame (Figure 5B).
Recently, the proposed function of PB-neurons as head-direction cells has been directly confirmed in Drosophila (Seelig & Jayaraman, 2015). Here, functional imaging was performed in a set of columnar neurons homologous to CL1a-neurons in other insects (Green et al., 2017; Kim, Rouault, Druckmann, & Jayaraman, 2017; Seelig & Jayaraman, 2015; Turner-Evans et al., 2017) (E-PG-neurons). These cells transfer information from single slices of the CBL (the ellipsoid body in flies) to single slices of the PB. When monitoring the activity of the entire population of these cells while the fly was walking on an air-suspended ball inside a virtual reality arena, a single bump of activity was revealed within this neuron population, in line with predictions from POL-neurons (Figure 5C). When the fly turned right, this bump moved leftward, and vice versa. Across the entire population, 360 degrees of the fly’s horizon were represented (Seelig & Jayaraman, 2015). This means that the position of the activity bump with respect to the width of the CX predicts the angular orientation of the fly; that is, it serves as an internal compass (Heinze, 2015).
Importantly, this head-direction signal did not depend on which specific visual signals were presented in the arena, and it was even present in complete darkness (Seelig & Jayaraman, 2015). The cells thus integrate self-generated proprioceptive cues with external visual cues into a coherent heading signal. The correlation between the anatomical position of the neural signal and the body orientation with respect to external cues was largely fixed within single experiments, but it was random when compared across individual flies. Thus, the phase of the head-direction code appears to be reset in each new environment. This is highly reminiscent of head-direction cells in vertebrates (Varga, Kathman, Martin, Guo, & Ritzmann, 2017) but unlike the POL-neuron-based direction map found in migratory locusts, which is consistent across individuals (Heinze & Homberg, 2007).
A second difference between the locust POL-neuron map and the Drosophila head-direction map is that POL-neurons cover only 180 degrees of the horizon in each hemisphere, while the Drosophila cells map 360 degrees. It should be noted that the cell types in which the POL-mapping was revealed (TB1, CPU1) are likely postsynaptic to the CL1a-neurons (E-PG-neurons) (Heinze, 2014; Heinze & Homberg, 2007). In fact, CL1a-cells are the only set of columnar cells in locusts, in which a correlation between anatomical position and POL-tuning was not found (Heinze & Homberg, 2009). Whether this is because there are several sets of physiologically different yet anatomically identical types of CL1a cells in locusts, whether these cells show no ordered E-vector mapping because they were examined across many locust individuals, or whether they indeed differ fundamentally from their Drosophila counterparts will have to be resolved in the future. Resolving these interspecies disparities likely will illuminate fundamentally conserved principles of how the CX-circuitry computes the head-direction signal.
Recent work, again in Drosophila, has already brought progress in this respect. Functional imaging in E-PG-neurons in combination with another set of CX-columnar cells, the P-EN-neurons [CL2 in other insects (Heinze & Homberg, 2008, 2009; Müller, Homberg, & Kühn, 1997)], has established a ring-attractor network as the basis for the head-direction signal (Green et al., 2017; Kim et al., 2017; Turner-Evans et al., 2017). P-EN-neurons also encode head direction, but their activity is either enhanced or decreased in amplitude depending on the direction of the fly’s angular movements (Green et al., 2017; Turner-Evans et al., 2017).
Due to a one-slice offset in the recursive, excitatory connections between both cell types, an unbalance in the P-EN-neuron activity generated by angular movements shifts the E-PG-neuron activity toward a new position in the attractor network, in line with the turning direction of the fly (Figure 5D). This circuit thus provides a mechanism of how rotational movements are continuously translated into an updated head-direction signal (Heinze, 2017a). The sensory information used to achieve this effect is not yet fully described, but rotational optic flow is encoded by P-EN-cells, while proprioceptive input likely exists, as visual cues are not necessary for the functionality of this circuit (Green et al., 2017; Turner-Evans et al., 2017).
The recurrent excitatory connection between P-EN and E-PG neurons is in line with the experimentally determined features of the CX-ring attractor circuit, as it provides local excitation. However, in addition to this local excitation, the ring-attractor network requires global inhibition (Green et al., 2017; Turner-Evans et al., 2017). Several models of the CX, in both Drosophila (Green et al., 2017; Kakaria & de Bivort, 2017; Turner-Evans et al., 2017) and bees (Stone et al., 2017), have suggested intrinsic cells of the PB (TB1-cells in insects other than flies) as likely source of this inhibition. When modeling a ring-attractor in the bee PB, only TB1-neurons were required for a functional attractor network, even though the homologous cells to E-PG and P-EN cells also exist in bees. This suggests that the overall attractor network is built in ways that ensure a robust generation of a sustained activity bump by complementary circuit motifs.
The outlined compass circuit of the CX is highly conserved. Work in the cockroach CX has uncovered head-direction cells that are very similar to their fly counterparts, albeit without revealing their anatomical identity (Varga & Ritzmann, 2016; Varga et al., 2017). In dung beetles (el Jundi et al., 2015b), monarch butterflies (Heinze & Reppert, 2011), and bees (Stone et al., 2017), not only is visual compass information transmitted to the CX by neurons homologous to those in flies (ring-neurons) and locusts (TL-neurons), but all other components of the compass network (CL1, TB1, CPU1) are present as well, down to the level of neuronal subtypes. This suggests that the delineated head-direction circuit is the basis for encoding body orientation across insects.
Encoding of the Desired Heading
How is information about body orientation used to guide the next steering decision? To compensate for mismatches between desired and current heading, the desired heading also has to be represented in the CX. Little is known about how this is achieved neurally. In addition, goals differ among navigational strategies (Figure 3), and strategies continuously compete with each other (Collett, 2012) and switch depending on motivational state (Knaden & Wehner, 2006), making clear predictions about an insect’s momentary navigational goal challenging to determine. Nevertheless, in some behaviors, goal directions are clearly defined.
When insects that use path integration return to their nest after an exploratory foraging trip, their goal is the nest, encoded by the home vector (i.e., an integrated memory of all directional changes and distances covered during the outbound trip). In bees, distances are measured by integrating an optic flow–based speed signal over time (Srinivasan, 2014, 2015), while sky compass cues serve as directional information (Evangelista et al., 2014). Recent work studying the sweat bee Megalopta genalis showed that neurons selectively activated by translational optic flow target the CX-noduli and are suited to encode the forward speed of the animal (Stone et al., 2017). Although these neurons also could be involved in other functions, these recordings show that bee CX receives both optic flow–based speed and compass information (Figure 5E).
A set of columnar neurons innervating the noduli and the PB [CPU4-neurons (Heinze & Homberg, 2008, 2009; Heinze et al., 2013), postsynaptic to the proposed speed neurons (Stone et al., 2017)] has been suggested to integrate both cues and theoretically is suited to generate a distributed memory of the home vector across the columns of the CX (Figure 5F). In each column, these neurons could accumulate neural activity proportional to flight speed whenever the bee flies in a specific compass direction (Stone et al., 2017). This directionally gated distance memory would yield a population-coded activity bump that directly represents the bee’s navigational goal. Indeed, when this hypothetical circuit was implemented as a biologically constrained computational model, it yielded a home-vector representation in the modeled CPU4-population that could be used to steer the bee back home (Stone et al., 2017), similar to purely theoretical models (Goldschmidt, Manoonpong, & Dasgupta, 2017; Haferlach, Wessnitzer, Mangan, & Webb, 2007).
Walking insects, such as ants, use a step-counter instead of optic flow to measure distances (Wittlinger, Wehner, & Wolf, 2006). Whether neurons homologous to the bee’s proposed speed neurons are activated by these inputs remains to be shown. If this is the case, a common mechanism for integrating speed and direction in path-integrating insects seems likely, and the activity of CPU4-neurons generally might represent the goal direction during this behavior.
The memory encoding the home vector in the described model is based on ongoing activity. Indeed, indications that short-term working memory is required to recall a navigational goal also have been found in Drosophila in different behavioral paradigms (Neuser et al., 2008; Ofstad, Zuker, & Reiser, 2011), most recently including idiothetic path integration (Kim & Dickinson, 2017). Similarly, dung beetles maintain a steady heading when rolling a dung-ball (Figure 3A) and return to their initially chosen direction after disturbance, which also requires short-term memory (el Jundi et al., 2016). Generally, activity-based, short-term memory of navigation directions could be used to compensate for course deviations during any directed movement. Evidence for recovery of directional information after presenting distracting sensory input has been found in CX-neurons of locusts (Bockhorst & Homberg, 2017). While the substrate of this memory is not known in any of these examples, CPU4-neurons are potential candidate cells in all cases.
Migratory insects have to maintain a steady heading as well, albeit not for seconds or minutes, but for weeks or months (Reppert et al., 2016; Warrant et al., 2016). While there are no data revealing the neural substrate of migratory headings in any insect, Stone et al. (2017) have speculated that the mechanism suited to encode the home vector during path integration also might represent migratory heading. By genetically fixing synaptic weights, the output of CPU4-neurons could be biased in a way that generates a hard-wired activity bump across these cells that encodes a stable goal direction (Heinze, 2017b). If relying on a sun compass, this goal direction would have to be shifted by output from the circadian clock to compensate for the daily movements of the Sun across the sky. A model explaining time-compensated migratory headings in monarch butterflies was proposed recently (Shlizerman, Phillips-Portillo, Forger, & Reppert, 2016), although concrete neural substrates remain unresolved. Both models combined provide a framework, based on which the neural implementation of migratory headings can be unraveled in the years to come.
Not only acute sensory information is responsible for driving behavior, but memorized features of the environment are key to many navigation decisions as well, such as for image-matching strategies in ants when navigating along fixed foraging routes or during landmark-based navigation (Collett, Chittka, & Collett, 2013a; Menzel, de Marco, & Greggers, 2006). Such memories are more stable than home vectors (Ziegler & Wehner, 1997) and require synaptic remodeling (Groh, Lu, Meinertzhagen, & Rössler, 2012; Stieb, Muenz, Wehner, & Rössler, 2010).
The major site for such long-term information storage in the insect brain is the mushroom body (Menzel, 2014; Mizunami, Weibrecht, & Strausfeld, 1998). It has been estimated through modeling of mushroom-body circuitry that several hundred panoramic images could be stored in an ant mushroom body (Ardin et al., 2016), thus comprising sufficient capacity for navigating along familiar routes. The output of the mushroom body in this model would indicate whether the current view matches one of the stored images. However, no prominent connection between the mushroom body and the CX has been found in any insect to date. It remains an open question, therefore, how mushroom body output could be integrated with representations of current and intended headings in the CX to elicit motor-steering responses.
Generation of Motor-Steering Commands
After comparing current and intended headings, the insect brain has to initiate steering movements to compensate for mismatches between the two directions. Besides its function in sensory coding, the CX long has been established as a center for locomotion (reviewed in Strausfeld, 1999). Two lines of evidence originally led to this insight. First, mutants of Drosophila that had developmental defects leading to structural damage of the CX (e.g., no bridge, ellipsoid body open, tay bridge, or ocelliless) were impaired in leg coordination, step-length regulation, and walking-speed regulation; showed abnormal turning behavior; and had general problems with directing actions in space (Martin, Raabe, & Heisenberg, 1999; Strauss & Heisenberg, 1993; Strauss, Hanesch, Kinkelin, Wolf, & Heisenberg, 1992; Triphan, Poeck, Neuser, & Strauss, 2010). Second, from a comparative approach, when carefully examining the anatomy of the CX across many species, there appears to be a correlation between the degree of dexterity of a species and the degree of separation of neighboring CX-columns (Strausfeld, 1999, 2009, 2012).
This finding suggests that the CX is instrumental in coordinating leg movements in space. The more precise those movements have to be, the more segregated the CX-columns seem to become. While the columnar neuroarchitecture has been shown to be associated with head-direction encoding, with each column representing a specific head direction (Heinze & Homberg, 2007; Turner-Evans et al., 2017), this layout is also suited to direct limb movements in the represented directions and the precision of these movements might depend on the amount of lateral overlap between neighboring columns, as suggested by Strausfeld (2009, 2012).
More recently, work with cockroaches has directly revealed that neuronal activity in the CX predicts the animal’s imminent movements (Bender, Pollack, & Ritzmann, 2010; Guo & Ritzmann, 2013; Martin, Guo, Mu, Harley, & Ritzmann, 2015). To achieve this finding, neurons of the cockroach CX were recorded with extracellular electrodes during tethered walking on an air-suspended ball or while freely exploring an experimental arena. This allowed correlating neural activity to the movements of the animal (i.e., to ongoing or future locomotor behavior). This work revealed that many cells show peak activity just before movements are initiated. These neurons were grouped into several functional classes, each of which predicted a specific combination of forward speed and angular velocity. The cells, therefore, essentially generate a map of future positions of the animal (Martin et al., 2015) (Figure 5G).
Unfortunately, the nature of the experiments did not allow for deducting the anatomical cell type of these neurons. However, as recordings were performed from the CBU, they might correspond to neurons described as part of the POL-network in other insects (most likely CPU1/2 cells, given their large fiber diameters and their proposed role as main output cells of the CX). That the neural activity correlated with movement indeed causes the observed behavioral effects was demonstrated directly by observing the animal’s movements while injecting a current via recording electrodes (Martin et al., 2015). The induced movements were well correlated with the movements predicted from recordings at that electrode position. Finally, leg coordination reflexes were examined directly during current injections into the CX that would lead to the turning of the animal. These reflexes were modulated in ways that closely resembled the modulation of the same reflexes during natural turning (Martin et al., 2015). Activity of CX-neurons, therefore, directly modulates thoracic reflexes that underlie turning movements.
Whereas the anatomical identity of these cockroach steering cells is unknown, a model of the bee’s CX suggests concrete neurons for this function (Stone et al., 2017). As mentioned in the section “Central Integration,” during path integration, CPU4-columnar neurons have been proposed to serve as a memory substrate for the home vector, while the PB compass cells likely represent the bee’s current heading. A second type of columnar cell (CPU1) is suited to receive input from the candidate memory cells and the direction cells and thus is ideally suited to comparing the bee’s current heading with its goal direction. Across insects, CPU1-cells converge in the lateral accessory lobes (LAL) (Heinze & Homberg, 2007, 2008; Heinze et al., 2013; Lin et al., 2013; Wolff, Iyer, & Rubin, 2015), regions known to be involved in motor control (reviewed in Namiki & Kanzaki, 2016a, 2016b). Indeed, when tested in simulations, this circuit yielded correct steering decisions during path integration (Stone et al., 2017). This also outlines how the CX could use its head-direction signal to control steering during navigation behaviors in general.
How do LALs initiate steering? Work studying the silkmoth (Bombyx mori) has investigated this region and its involvement in steering for more than two decades. Neurons in the LALs show a characteristic behavior of alternating periods of persistently high and low firing rates. The transition between these states is triggered by short sensory stimuli, most prominently pheromone pulses. This behavior gave these neurons the name flip-flop neurons (Iwano et al., 2010; Namiki, Iwabuchi, Pansopha Kono, & Kanzaki, 2014; Olberg, 1983). Importantly, this alternating activity is exactly what would be expected from a steering network when considering the behavioral context: When moths follow an odor plume (in this case the female’s pheromone trail), the animals locate the plume, orient themselves upwind, and then perform a zig-zagging flight pattern that crosses the plume many times (Namiki & Kanzaki, 2016b). Each time the male crosses the plume, it has to reverse direction in order to continue flying toward the odor source.
Interestingly, many types of neurons with flip-flop activity interconnect the LALs between both brain hemispheres, some of which are GABAergic (Iwano et al., 2010). This suggests that they are part of a mechanism that inhibits the contralateral LAL, whenever the ipsilateral LAL is active and vice versa. If one assumes that descending pathways originating in either LAL control steering toward one direction, this mutually exclusive, alternating activation would be perfectly suited to drive steering during odor-plume tracking (Namiki & Kanzaki, 2016b). Indeed, motor neurons in the subesophageal ganglion innervating the neck muscles (responsible for turning the head) also show flip-flop activity and are likely activated by neurons descending from the LAL (Mishima & Kanzaki, 1998). As head turning is associated with body turns, the flip-flop activity in the LAL neurons is likely premotor activity responsible for steering during zig-zagging behavior.
As the activity transitioning between high and low states in flip-flop neurons of the LAL also can be induced by light pulses (Olberg, 1983), this circuit in the LAL is likely not limited to the control of pheromone-tracking behavior, but it might be in control of all steering decisions in the context of navigation behaviors. Indeed, alternating between rightward and leftward steering during navigation is a common theme across many insect behaviors (Namiki & Kanzaki, 2016b).
Despite all the progress that has been made over recent years to unravel the principles underlying insect navigation, major challenges remain. Much of this article has focused on the neural processes underlying visually driven behaviors, but as outlined at the beginning, many sensory signals are involved in guiding insects. These include olfactory signals, proprioceptive information, wind, sound, magnetic fields, and tactile information. Having established the CX of the insect brain as the main hub for navigational control, the question arises of how these nonvisual inputs are integrated with the visual compass of the CX.
Findings in Drosophila show that the head-direction circuit in that species also functions in darkness (Seelig & Jayaraman, 2015), and this suggests that nonvisual idiothetic cues are integrated to form the activity bump encoding the current body orientation. Proprioceptive signals are most likely the source of this information, but it is unknown how these are transferred to the fly’s CX. Some evidence from locusts suggests that CBL-tangential neurons are not only encoding visual information, but also signal passive movements of the wings (Homberg, 1994). Whether these are the same cells involved in compass encoding or cells operating in parallel remains to be shown.
Via extracellular recordings, mechanosensory input from the antenna was shown to be present in cells of the cockroach CX, indeed partially overlapping with visual neurons (Ritzmann, Ridgel, & Pollack, 2008). Other mechanical stimuli like wind and sound are detected via different means in different species, and therefore, distinct pathways leading to the brain that account for these differences must exist in each of them. One region involved in the processing of mechanosensory information (e.g., wind, sounds, and touch) from the antennae is the antennal mechanosensory and motor center (AMMC) in the deutocerebrum of the brain (Ito et al., 2014; Kamikouchi, Shimada, & Ito, 2006). Yet it is unclear how this region connects to navigation-relevant brain areas.
Even more difficult is the situation regarding olfactory information, as no direct connection exists between the CX and the well-described system for processing olfactory signals (i.e., antennal lobe, mushroom body, and lateral horn). Pheromone-processing pathways involving the superior medial protocerebrum have been proposed as indirectly linking the antennal lobe to the LAL (Kanzaki, Soo, Seki, & Wada, 2003) and providing potential convergence sites downstream of the CX. Overall, how species integrate many different sensory inputs into a coherent representation of heading, as well as into a clearly defined goal direction, remains a major open question in the field.
Another main challenge is to develop a thorough understanding of the representation of navigational goals in the brain. The concrete neurons in the CX predicted to encode the home vector during path integration in bees are potentially also suited to encode the desired heading of insects in other behavioral contexts (Stone et al., 2017; Heinze, 2017b). However, proving that this is indeed the case is not trivial; it will require novel technological developments to maintain the internal driving force responsible for the selection of a goal in a particular behavioral context—all while having experimental access to CX-neurons. In addition, these long-term intended headings have to be suppressed when the immediate behavior is driven by escape reflexes or flight stabilization maneuvers. Where and how are these behavioral control modules integrated?
Finally, how are the various navigational strategies generated and combined? Behavioral work with ants and bees has delivered a rich pool of data that now can be integrated with what we have learned from neural control of behavior. Are different strategies controlled by the same neural circuits? Are they mutually exclusive, or are they operating at the same time, with the most reliable output driving steering movements? Even spectacular navigational strategies like long-range navigation cause only minor changes in overall brain anatomy compared to nonmigratory relatives (de Vries et al., 2017), suggesting that similar circuits are required for different strategies and that the differences are to be found on the level of synapses, dendritic trees, and other details of neural circuitry.
Aptekar, J. W., Keleş, M. F., Lu, P. M., Zolotova, N. M., & Frye, M. A. (2015). Neurons forming optic glomeruli compute figure-ground discriminations in Drosophila. Journal of Neuroscience, 35, 7587–7599.Find this resource:
Ardin, P., Peng, F., Mangan, M., Lagogiannis, K., & Webb, B. (2016). Using an insect mushroom body circuit to encode route memory in complex natural environments. PLoS Computational Biology, 12, e1004683.Find this resource:
Bech, M., Homberg, U., & Pfeiffer, K. (2014). Receptive fields of locust brain neurons are matched to polarization patterns of the sky. Current Biology, 24, 2124–2129.Find this resource:
Bender, J. A., Pollack, A. J., & Ritzmann, R. E. (2010). Neural activity in the central complex of the insect brain is linked to locomotor changes. Current Biology, 20, 921–926.Find this resource:
Bockhorst, T., & Homberg, U. (2017). Interaction of compass sensing and object-motion detection in the locust central complex. Journal of Neurophysiology, 118, 496–506.Find this resource:
Brines, M. L., & Gould, J. L. (1982). Skylight polarization patterns and animal orientation. Journal of Experimental Biology, 96, 69–91.Find this resource:
Chapman, J. W., Reynolds, D. R., Mouritsen, H., Hill, J. K., Riley, J. R., Sivell, D., . . . Woiwod, I. P. (2008). Wind selection and drift compensation optimize migratory pathways in a high-flying moth. Current Biology, 18, 514–518.Find this resource:
Cheeseman, J. F., Millar, C. D., Greggers, U., Lehmann, K., Pawley, M. D. M., Gallistel, C. R., . . . & Menzel, R. (2014a). Reply to Cheung et al.: The cognitive map hypothesis remains the best interpretation of the data in honeybee navigation. Proceedings of the National Academy of Sciences of the United States of America, 111(42), E4398.Find this resource:
Cheeseman, J. F., Millar, C. D., Greggers, U., Lehmann, K., Pawley, M. D. M., Gallistel, C. R., . . . Menzel, R. (2014b). Way-finding in displaced clock-shifted bees proves bees use a cognitive map. Proceedings of the National Academy of Sciences of the United States of America, 111(24), 8949–8954.Find this resource:
Cheung, A. (2014). Animal path integration: A model of positional uncertainty along tortuous paths. Journal of Theoretical Biology, 341, 17–33.Find this resource:
Cheung, A., Collett, M., Collett, T. S., Dewar, A., Dyer, F., Graham, P., . . . Zeil, J. (2014). Still no convincing evidence for cognitive map use by honeybees. Proceedings of the National Academy of Sciences of the United States of America, 111(42), E4396–E4397.Find this resource:
Clites, B. L., & Pierce, J. T. (2017). Identifying cellular and molecular mechanisms for magnetosensation. Annual Review of Neuroscience, 40, 231–250.Find this resource:
Collett, M. (2010). How desert ants use a visual landmark for guidance along a habitual route. Proceedings of the National Academy of Sciences of the United States of America, 107(25), 11638–11643.Find this resource:
Collett, M. (2012). How navigational guidance systems are combined in a desert ant. Current Biology, 22, 927–932.Find this resource:
Collett, M., Chittka, L., & Collett, T. S. (2013a). Spatial memory in insect navigation. Current Biology, 23, R789–R800.Find this resource:
Collett, T. S., & Collett, M. (2002). Memory use in insect visual navigation. Nature Reviews Neuroscience, 3, 542–552.Find this resource:
Collett, T. S., de Ibarra, N. H., Riabinina, O., & Philippides, A. (2013b). Coordinating compass-based and nest-based flight directions during bumblebee learning and return flights. Journal of Experimental Biology, 216, 1105–1113.Find this resource:
Dacke, M., Baird, E., Byrne, M. J., Scholtz, C. H., & Warrant, E. J. (2013). Dung beetles use the Milky Way for orientation. Current Biology, 23, 298–300.Find this resource:
Dacke, M., el Jundi, B., Smolka, J., Byrne, M. J., & Baird, E. (2014). The role of the sun in the celestial compass of dung beetles. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 369(1636), 20130036.Find this resource:
Dacke, M., Nilsson, D.-E., Scholtz, C. H., Byrne, M. J., & Warrant, E. J. (2003). Animal behaviour: Insect orientation to polarized moonlight. Nature, 424, 33.Find this resource:
de Vries, L., Pfeiffer, K., Trebels, B., Adden, A. K., Green, K., Warrant, E. J., & Heinze, S. (2017). Comparison of navigation-related brain regions in migratory versus non-migratory noctuid moths. Frontiers of Behavioral Neuroscience, 11, 158.Find this resource:
Degen, J., Kirbach, A., Reiter, L., Lehmann, K., Norton, P., Storms, M., . . . Menzel, R. (2016). Honeybees learn landscape features during exploratory orientation flights. Current Biology, 26, 2800–2804.Find this resource:
Evangelista, C., Kraft, P., Dacke, M., Labhart, T., & Srinivasan, M. V. (2014). Honeybee navigation: Critically examining the role of the polarization compass. Philosophical Transactions of the Royal Society of London B: Biological Sciences, 369(1636), 20130037.Find this resource:
Fahrbach, S. E. (2006). Structure of the mushroom bodies of the insect brain. Annual Review of Entomology, 51, 209–232.Find this resource:
Foster, J. J., el Jundi, B., Smolka, J., Khaldy, L., Nilsson, D.-E., Byrne, M. J., & Dacke, M. (2017). Stellar performance: Mechanisms underlying Milky Way orientation in dung beetles. Philosophical Transactions of the Royal Society of London B, 372(1717), 20160079.Find this resource:
Goldschmidt, D., Manoonpong, P., & Dasgupta, S. (2017). A neurocomputational model of goal-directed navigation in insect-inspired artificial agents. Frontiers in Neurorobotics, 11, e1004683–17.Find this resource:
Gould, J. L. (2010). Magnetoreception. Current Biology, 20, R431–R435.Find this resource:
Graham, P., & Cheng, K. (2009). Ants use the panoramic skyline as a visual cue during navigation. Current Biology, 19, R935–R937.Find this resource:
Green, J., Adachi, A., Shah, K. K., Hirokawa, J. D., Magani, P. S., & Maimon, G. (2017). A neural circuit architecture for angular integration in Drosophila. Nature, 546, 101–106.Find this resource:
Groh, C., Lu, Z., Meinertzhagen, I. A., & Rössler, W. (2012). Age-related plasticity in the synaptic ultrastructure of neurons in the mushroom body calyx of the adult honeybee Apis mellifera. Journal of Comparative Neurology, 520, 3509–3527.Find this resource:
Guerra, P. A., Gegear, R. J., & Reppert, S. M. (2014). A magnetic compass aids monarch butterfly migration. Nature Communications, 5, 4164.Find this resource:
Guerra, P. A., Merlin, C., Gegear, R. J., & Reppert, S. M. (2012). Discordant timing between antennae disrupts sun compass orientation in migratory monarch butterflies. Nature Communications, 3, 958.Find this resource:
Guerra, P. A., & Reppert, S. M. (2013). Coldness triggers northward flight in remigrant monarch butterflies. Current Biology, 23, 419–423.Find this resource:
Guerra, P. A., & Reppert, S. M. (2015). Sensory basis of lepidopteran migration: Focus on the monarch butterfly. Current Opinion in Neurobiology, 34C, 20–28.Find this resource:
Guo, P., & Ritzmann, R. E. (2013). Neural activity in the central complex of the cockroach brain is linked to turning behaviors. Journal of Experimental Biology, 216, 992–1002.Find this resource:
Haferlach, T., Wessnitzer, J., Mangan, M., & Webb, B. (2007). Evolving a neural model of insect path integration. Adaptive Behavior, 15, 273–287.Find this resource:
Heinze, S. (2014). Polarized-light processing in insect brains: Recent insights from the desert locust, the monarch butterfly, the cricket, and the fruit fly. In G. Horváth (Ed.), Polarized light and polarization vision in animal sciences (pp. 61–111). Berlin: Springer.Find this resource:
Heinze, S. (2015a). Polarization vision. In R. Jung & D. Jaeger (Eds.), Encyclopedia of computational neuroscience (pp. 1–30). Berlin: Springer.Find this resource:
Heinze, S. (2015b). Neuroethology: Unweaving the senses of direction. Current Biology, 25, R1034–R1037.Find this resource:
Heinze, S. (2017a). Neural coding: Bumps on the move. Current Biology, 27, R409–R412.Find this resource:
Heinze, S. (2017b). Unraveling the neural basis of insect navigation. Current Opinion in Insect Science, 24, 1–10.Find this resource:
Heinze, S., Florman, J., Asokaraj, S., el Jundi, B., & Reppert, S. M. (2013). Anatomical basis of sun compass navigation II: The neuronal composition of the central complex of the monarch butterfly. Journal of Comparative Neurology, 521, 267–298.Find this resource:
Heinze, S., Gotthardt, S., & Homberg, U. (2009). Transformation of polarized light information in the central complex of the locust. Journal of Neuroscience, 29, 11783–11793.Find this resource:
Heinze, S., & Homberg, U. (2007). Maplike representation of celestial E-vector orientations in the brain of an insect. Science, 315, 995–997.Find this resource:
Heinze, S., & Homberg, U. (2008). Neuroarchitecture of the central complex of the desert locust: Intrinsic and columnar neurons. Journal of Comparative Neurology, 511, 454–478.Find this resource:
Heinze, S., & Homberg, U. (2009). Linking the input to the output: New sets of neurons complement the polarization vision network in the locust central complex. Journal of Neuroscience, 29, 4911–4921.Find this resource:
Heinze, S., & Reppert, S. M. (2011). Sun compass integration of skylight cues in migratory monarch butterflies. Neuron, 69, 345–358.Find this resource:
Heinze, S., & Warrant, E. J. (2016). Bogong moths. Current Biology, 26, R263–R265.Find this resource:
Held, M., Berz, A., Hensgen, R., Muenz, T. S., Scholl, C., Rössler, W., . . . Pfeiffer, K. (2016). Microglomerular synaptic complexes in the sky-compass network of the honeybee connect parallel pathways from the anterior optic tubercle to the central complex. Frontiers of Behavioral Neuroscience, 10, 186.Find this resource:
Homberg, U. (1985). Interneurones of the central complex in the bee brain (Apis mellifera, L.). Journal of Insect Physiology, 31, 251–264.Find this resource:
Homberg, U. (1994). Flight-correlated activity changes in neurons of the lateral accessory lobes in the brain of the locust Schistocerca gregaria. Journal of Comparative Physiology A, 175, 597–610.Find this resource:
Homberg, U. (2008). Evolution of the central complex in the arthropod brain with respect to the visual system. Arthopod Structure & Development, 37, 347–362.Find this resource:
Homberg, U. (2015). Sky compass orientation in desert locusts—Evidence from field and laboratory studies. Frontiers of Behavioral Neuroscience, 9, 346.Find this resource:
Homberg, U., Heinze, S., Pfeiffer, K., Kinoshita, M., & el Jundi, B. (2011). Central neural coding of sky polarization in insects. Philosophical Transactions of the Royal Society of London B, 366(1565), 680–687.Find this resource:
Homberg, U., & Paech, A. (2002). Ultrastructure and orientation of ommatidia in the dorsal rim area of the locust compound eye. Arthopod Structure & Development, 30, 271–280.Find this resource:
Hore, P. J., & Mouritsen, H. (2016). The radical-pair mechanism of magnetoreception. Annual Review of Biophysics, 45, 299–344.Find this resource:
Immonen, E.-V., Dacke, M., Heinze, S., & el Jundi, B. (2017). Anatomical organization of the brain of a diurnal and a nocturnal dung beetle. Journal of Comparative Neurology, 525, 1879–908.Find this resource:
Ito, K., Shinomiya, K., Ito, M., Armstrong, J. D., Boyan, G., Hartenstein, V., . . . Insect Brain Name Working Group. (2014). A systematic nomenclature for the insect brain. Neuron, 81, 755–765.Find this resource:
Iwano, M., Hill, E. S., Mori, A., Mishima, T., Mishima, T., Ito, K., & Kanzaki, R. (2010). Neurons associated with the flip-flop activity in the lateral accessory lobe and ventral protocerebrum of the silkworm moth brain. Journal of Comparative Neurology, 518, 366–388.Find this resource:
el Jundi, B., Foster, J. J., Byrne, M. J., Baird, E., & Dacke, M. (2015a). Spectral information as an orientation cue in dung beetles. Biology Letters, 11, 20150656.Find this resource:
el Jundi, B., Foster, J. J., Khaldy, L., Byrne, M. J., Dacke, M., & Baird, E. (2016). A snapshot-based mechanism for celestial orientation. Current Biology, 26, 1456–1462.Find this resource:
el Jundi, B., Heinze, S., Lenschow, C., Kurylas, A., Rohlfing, T., & Homberg, U. (2009). The Locust Standard Brain: A 3D standard of the central complex as a platform for neural network analysis. Frontiers in Systems Neuroscience, 3, 21.Find this resource:
el Jundi, B., & Homberg, U. (2012). Receptive field properties and intensity-response functions of polarization-sensitive neurons of the optic tubercle in gregarious and solitarious locusts. Journal of Neurophysiology, 108, 1695–1710.Find this resource:
el Jundi, B., Pfeiffer, K., Heinze, S., & Homberg, U. (2014a). Integration of polarization and chromatic cues in the insect sky compass. Journal of Comparative Physiology A, 200, 575–589.Find this resource:
el Jundi, B., Pfeiffer, K., & Homberg, U. (2011). A distinct layer of the medulla integrates sky compass signals in the brain of an insect. PLoS ONE, 6, e27855.Find this resource:
el Jundi, B., Smolka, J., Baird, E., Byrne, M. J., & Dacke, M. (2014b). Diurnal dung beetles use the intensity gradient and the polarization pattern of the sky for orientation. Journal of Experimental Biology, 217, 2422–2429.Find this resource:
el Jundi, B., Warrant, E. J., Byrne, M. J., Khaldy, L., Baird, E., Smolka, J., & Dacke, M. (2015b). Neural coding underlying the cue preference for celestial orientation. Proceedings of the National Academy of Sciences of the United States of America, 112(36), 11395–11400.Find this resource:
Kakaria, K. S., & de Bivort, B. L. (2017). Ring attractor dynamics emerge from a spiking model of the entire protocerebral bridge. Frontiers of Behavioral Neuroscience, 11, 8.Find this resource:
Kamikouchi, A., Shimada, T., & Ito, K. (2006). Comprehensive classification of the auditory sensory projections in the brain of the fruit fly Drosophila melanogaster. Journal of Comparative Neurology, 499, 317–356.Find this resource:
Kanzaki, R., Soo, K., Seki, Y., & Wada, S. (2003). Projections to higher olfactory centers from subdivisions of the antennal lobe macroglomerular complex of the male silkmoth. Chemical Senses, 28, 113–130.Find this resource:
Kim, I. S., & Dickinson, M. H. (2017). Idiothetic path integration in the fruit fly Drosophila melanogaster. Current Biology, 27, 2227–2238, e3.Find this resource:
Kim, S. S., Rouault, H., Druckmann, S., & Jayaraman, V. (2017). Ring attractor dynamics in the Drosophila central brain. Science, 356, 849–853.Find this resource:
Kinoshita, M., Pfeiffer, K., & Homberg, U. (2007). Spectral properties of identified polarized-light sensitive interneurons in the brain of the desert locust Schistocerca gregaria. Journal of Experimental Biology, 210, 1350–1361.Find this resource:
Knaden, M., & Graham, P. (2016). The sensory ecology of ant navigation: From natural environments to neural mechanisms. Annual Review of Entomology, 61, 63–76.Find this resource:
Knaden, M., & Wehner, R. (2006). Ant navigation: Resetting the path integrator. Journal of Experimental Biology, 209, 26–31.Find this resource:
Krapp, H. G., & Hengstenberg, R. (1996). Estimation of self-motion by optic flow processing in single visual interneurons. Nature, 384, 463–466.Find this resource:
Kühn-Bühlmann, S., & Wehner, R. (2006). Age-dependent and task-related volume changes in the mushroom bodies of visually guided desert ants, Cataglyphis bicolor. Journal of Neurobiology, 66, 511–521.Find this resource:
Kurylas, A. E., Rohlfing, T., Krofczik, S., Jenett, A., & Homberg, U. (2008). Standardized atlas of the brain of the desert locust, Schistocerca gregaria. Cell Tissue Research, 333, 125–145.Find this resource:
Labhart, T. (1988). Polarization-opponent interneurons in the insect visual system. Nature, 331, 435–437.Find this resource:
Labhart, T. (2000). Polarization-sensitive interneurons in the optic lobe of the desert ant Cataglyphis bicolor. Naturwissenschaften, 87, 133–136.Find this resource:
Labhart, T. (2016). Can invertebrates see the e-vector of polarization as a separate modality of light? Journal of Experimental Biology, 219, 3844–3856.Find this resource:
Labhart, T., & Meyer, E. P. (1999). Detectors for polarized skylight in insects: A survey of ommatidial specializations in the dorsal rim area of the compound eye. Microscopy Research and Technique, 47, 368–379.Find this resource:
Labhart, T., & Meyer, E. P. (2002). Neural mechanisms in insect navigation: Polarization compass and odometer. Current Opinion in Neurobiology, 12, 707–714.Find this resource:
Labhart, T., Petzold, J., & Helbling, H. (2001). Spatial integration in polarization-sensitive interneurones of crickets: A survey of evidence, mechanisms and benefits. Journal of Experimental Biology, 204, 2423–2430.Find this resource:
Liedvogel, M., & Mouritsen, H. (2010). Cryptochromes—a potential magnetoreceptor: What do we know and what do we want to know? Journal of the Royal Society, 7(Suppl. 2), S147–S162.Find this resource:
Lin, C.-Y., Chuang, C.-C., Hua, T.-E., Chen, C.-C., Dickson, B. J., Greenspan, R. J., & Chiang, A.-S. (2013). A comprehensive wiring diagram of the protocerebral bridge for visual information processing in the Drosophila brain. Cell Reports, 3, 1739–1753.Find this resource:
Liu, G., Seiler, H., Wen, A., Zars, T., Ito, K., Wolf, R., . . . Liu, L. (2006). Distinct memory traces for two visual features in the Drosophila brain. Nature, 439, 551–556.Find this resource:
Loesel, R., & Homberg, U. (2001). Anatomy and physiology of neurons with processes in the accessory medulla of the cockroach Leucophaea maderae. Journal of Comparative Neurology, 439, 193–207.Find this resource:
Lohmann, K. J., Lohmann, C. M. F., & Putman, N. F. (2007). Magnetic maps in animals: nature’s GPS. Journal of Experimental Biology, 210, 3697–3705.Find this resource:
Mappes, M., & Homberg, U. (2004). Behavioral analysis of polarization vision in tethered flying locusts. Journal of Comparative Physiology A, 190, 61–68.Find this resource:
Martin, J., Guo, P., Mu, L., Harley, C. M., & Ritzmann, R. E. (2015). Central-complex control of movement in the freely walking cockroach. Current Biology, 25, 2795–2803.Find this resource:
Martin, J.-R., Raabe, T., & Heisenberg, M. (1999). Central complex substructures are required for the maintenance of locomotor activity in Drosophila melanogaster. Journal of Comparative Physiology A, 185, 277–288.Find this resource:
Menzel, R. (2014). The insect mushroom body, an experience-dependent recoding device. Journal of Physiology—Paris, 108, 84–95.Find this resource:
Menzel, R., de Marco, R. J., & Greggers, U. (2006). Spatial memory, navigation, and dance behaviour in Apis mellifera. Journal of Comparative Physiology A, 192, 889–903.Find this resource:
Menzel, R., Greggers, U., Smith, A., Berger, S., Brandt, R., Brunke, S., . . . Watzl, S. (2005). Honey bees navigate according to a map-like spatial memory. Proceedings of the National Academy of Sciences of the United States of America, 102(8), 3040–3045.Find this resource:
Menzel, R., Kirbach, A., Haass, W.-D., Fischer, B., Fuchs, J., Koblofsky, M., . . . Greggers, U. (2011). A common frame of reference for learned and communicated vectors in honeybee navigation. Current Biology, 21, 645–650.Find this resource:
Merlin, C., Gegear, R. J., & Reppert, S. M. (2009). Antennal circadian clocks coordinate sun compass orientation in migratory monarch butterflies. Science, 325, 1700–1704.Find this resource:
Merlin, C., Heinze, S., & Reppert, S. M. (2012). Unraveling navigational strategies in migratory insects. Current Opinion in Neurobiology, 22, 353–361.Find this resource:
Mishima, T., & Kanzaki, R. (1998). Coordination of flipflopping neural signals and head turning during pheromone-mediated walking in a male silkworm moth Bombyx mori. Journal of Comparative Physiology A, 183, 273–282.Find this resource:
Mizunami, M., Weibrecht, J. M., & Strausfeld, N. J. (1998). Mushroom bodies of the cockroach: Their participation in place memory. Journal of Comparative Neurology, 402, 520–537.Find this resource:
Mota, T., Gronenberg, W., Giurfa, M., & Sandoz, J.-C. (2013). Chromatic processing in the anterior optic tubercle of the honey bee brain. Journal of Neuroscience, 33, 4–16.Find this resource:
Mouritsen, H., & Frost, B. J. (2002). Virtual migration in tethered flying monarch butterflies reveals their orientation mechanisms. Proceedings of the National Academy of Sciences of the United States of America, 99(15), 10162–10166.Find this resource:
Müller, M., Homberg, U., & Kühn, A. (1997). Neuroarchitecture of the lower division of the central body in the brain of the locust (Schistocerca gregaria). Cell Tissue Research, 288, 159–176.Find this resource:
Müller, M., & Wehner, R. (2007). Wind and sky as compass cues in desert ant navigation. Naturwissenschaften, 94, 589–594.Find this resource:
Namiki, S., Iwabuchi, S., Pansopha Kono, P., & Kanzaki, R. (2014). Information flow through neural circuits for pheromone orientation. Nature Communications, 5, 5919.Find this resource:
Namiki, S., & Kanzaki, R. (2016a). Comparative neuroanatomy of the lateral accessory lobe in the insect brain. Frontiers in Physiology, 7, 244.Find this resource:
Namiki, S., & Kanzaki, R. (2016b). The neurobiological basis of orientation in insects: Insights from the silkmoth mating dance. Current Opinion in Insect Science, 15, 16–26.Find this resource:
Neuser, K., Triphan, T., Mronz, M., Poeck, B., & Strauss, R. (2008). Analysis of a spatial orientation memory in Drosophila. Nature, 453, 1244–1247.Find this resource:
Ofstad, T. A., Zuker, C. S., & Reiser, M. B. (2011). Visual place learning in Drosophila melanogaster. Nature, 474, 204–207.Find this resource:
Okamura, J.-Y., & Strausfeld, N. J. (2007). Visual system of calliphorid flies: Motion- and orientation-sensitive visual interneurons supplying dorsal optic glomeruli. Journal of Comparative Neurology, 500, 189–208.Find this resource:
Olberg, R. M. (1983). Pheromone-triggered flip-flopping interneurons in the ventral nerve cord of the silkworm moth, Bombyx mori. Journal of Comparative Physiology A, 152, 297–307.Find this resource:
Omoto, J. J., Keleş, M. F., Nguyen, B.-C. M., Bolanos, C., Lovick, J. K., Frye, M. A., & Hartenstein, V. (2017). Visual input to the Drosophila central complex by developmentally and functionally distinct neuronal populations. Current Biology, 27, 1098–1110.Find this resource:
Otsuna, H., & Ito, K. (2006). Systematic analysis of the visual projection neurons of Drosophila melanogaster. I. Lobula-specific pathways. Journal of Comparative Neurology, 497, 928–958.Find this resource:
Paulk, A. C., & Gronenberg, W. (2008). Higher order visual input to the mushroom bodies in the bee, Bombus impatiens. Arthopod Structure & Development, 37, 443–458.Find this resource:
Paulk, A. C., Phillips-Portillo, J., Dacks, A. M., Fellous, J.-M., & Gronenberg, W. (2008). The processing of color, motion, and stimulus timing are anatomically segregated in the bumblebee brain. Journal of Neuroscience, 28, 6319–6332.Find this resource:
Pfeiffer, K., & Homberg, U. (2007). Coding of azimuthal directions via time-compensated combination of celestial compass cues. Current Biology, 17, 960–965.Find this resource:
Pfeiffer, K., & Homberg, U. (2014). Organization and functional roles of the central complex in the insect brain. Annual Review of Entomology, 59, 165–184.Find this resource:
Pfeiffer, K., & Kinoshita, M. (2012). Segregation of visual inputs from different regions of the compound eye in two parallel pathways through the anterior optic tubercle of the bumblebee (Bombus ignitus). Journal of Comparative Neurology, 520, 212–229.Find this resource:
Pfeiffer, K., Kinoshita, M., & Homberg, U. (2005). Polarization-sensitive and light-sensitive neurons in two parallel pathways passing through the anterior optic tubercle in the locust brain. Journal of Neurophysiology, 94, 3903–3915.Find this resource:
Phillips-Portillo, J., & Strausfeld, N. J. (2012). Representation of the brain’s superior protocerebrum of the flesh fly, Neobellieria bullata, in the central body. Journal of Comparative Neurology, 520, 3070–3087.Find this resource:
Reiser, M. B., & Dickinson, M. H. (2010). Drosophila fly straight by fixating objects in the face of expanding optic flow. Journal of Experimental Biology, 213, 1771–1781.Find this resource:
Reppert, S. M., Guerra, P. A., & Merlin, C. (2016). Neurobiology of monarch butterfly migration. Annual Review of Entomology, 61, 25–42.Find this resource:
Reppert, S. M., Zhu, H., & White, R. H. (2004). Polarized light helps monarch butterflies navigate. Current Biology, 14, 155–158.Find this resource:
Ritz, T., Adem, S., & Schulten, K. (2000). A model for photoreceptor-based magnetoreception in birds. Biophysical Journal, 78, 707–718.Find this resource:
Ritz, T., Ahmad, M., Mouritsen, H., Wiltschko, R., & Wiltschko, W. (2010). Photoreceptor-based magnetoreception: optimal design of receptor molecules, cells, and neuronal processing. Journal of the Royal Society, 7(Suppl. 2), S135–S146.Find this resource:
Ritzmann, R. E., Ridgel, A. L., & Pollack, A. J. (2008). Multi-unit recording of antennal mechano-sensitive units in the central complex of the cockroach, Blaberus discoidalis. Journal of Comparative Physiology A, 194, 341–360.Find this resource:
Sakura, M., Lambrinos, D., & Labhart, T. (2008). Polarized skylight navigation in insects: Model and electrophysiology of e-vector coding by neurons in the central complex. Journal of Neurophysiology, 99, 667–682.Find this resource:
Schwarz, S., Narendra, A., & Zeil, J. (2011). The properties of the visual system in the Australian desert ant Melophorus bagoti. Arthopod Structure & Development, 40, 128–134.Find this resource:
Seelig, J. D., & Jayaraman, V. (2015). Neural dynamics for landmark orientation and angular path integration. Nature, 521, 186–191.Find this resource:
Shlizerman, E., Phillips-Portillo, J., Forger, D. B., & Reppert, S. M. (2016). Neural integration underlying a time-compensated sun compass in the migratory monarch butterfly. Cell Reports, 15, 683–691.Find this resource:
Srinivasan, M. V. (2014). Going with the flow: A brief history of the study of the honeybee’s navigational “odometer.” Journal of Comparative Physiology A, 200, 563–573.Find this resource:
Srinivasan, M. V. (2015). Where paths meet and cross: Navigation by path integration in the desert ant and the honeybee. Journal of Comparative Physiology A, 201, 533–546.Find this resource:
Stalleicken, J., Labhart, T., & Mouritsen, H. (2006). Physiological characterization of the compound eye in monarch butterflies with focus on the dorsal rim area. Journal of Comparative Physiology A, 192, 321–331.Find this resource:
Stalleicken, J., Mukhida, M., Labhart, T., Wehner, R., Frost, B. J., & Mouritsen, H. (2005). Do monarch butterflies use polarized skylight for migratory orientation? Journal of Experimental Biology, 208, 2399–2408.Find this resource:
Stieb, S. M., Muenz, T. S., Wehner, R., & Rössler, W. (2010). Visual experience and age affect synaptic organization in the mushroom bodies of the desert ant Cataglyphis fortis. Developmental Neurobiology, 70, 408–423.Find this resource:
Stone, T., Webb, B., Adden, A., Weddig, N. B., Honkanen, A., Templin, R. M., . . . Heinze, S. (2017). An anatomically constrained model for path integration in the bee brain. Current Biology, 27, 3069–3085.e11.Find this resource:
Strausfeld, N. J. (1999). A brain region in insects that supervises walking. Progress in Brain Research, 123, 273–284.Find this resource:
Strausfeld, N. J. (2009). Brain organization and the origin of insects: An assessment. Proceedings. Biological Sciences, 276, 1929–1937.Find this resource:
Strausfeld, N. J. (2012). Arthropod brains. Cambridge, MA: Harvard University Press.Find this resource:
Strausfeld, N. J., & Okamura, J.-Y. (2007). Visual system of calliphorid flies: Organization of optic glomeruli and their lobula complex efferents. Journal of Comparative Neurology, 500, 166–188.Find this resource:
Strauss, R., Hanesch, U., Kinkelin, M., Wolf, R., & Heisenberg, M. (1992). No-bridge of Drosophila melanogaster: Portrait of a structural brain mutant of the central complex. Journal of Neurogenetics, 8, 125–155.Find this resource:
Strauss, R., & Heisenberg, M. (1993). A higher control center of locomotor behavior in the Drosophila brain. Journal of Neuroscience, 13, 1852–1861.Find this resource:
Strauss, R., & Pichler, J. (1998). Persistence of orientation toward a temporarily invisible landmark in Drosophila melanogaster. Journal of Comparative Physiology A, 182, 411–423.Find this resource:
Stürzl, W., Zeil, J., Boeddeker, N., & Hemmi, J. M. (2016). How wasps acquire and use views for homing. Current Biology, 26, 470–482.Find this resource:
Träger, U., Wagner, R., Bausenwein, B., & Homberg, U. (2008). A novel type of microglomerular synaptic complex in the polarization vision pathway of the locust brain. Journal of Comparative Neurology, 506, 288–300.Find this resource:
Triphan, T., Poeck, B., Neuser, K., & Strauss, R. (2010). Visual targeting of motor actions in climbing Drosophila. Current Biology, 20, 663–668.Find this resource:
Turner-Evans, D., Wegener, S., Rouault, H., Franconville, R., Wolff, T., Seelig, J. D., . . . Jayaraman, V. (2017). Angular velocity integration in a fly heading circuit. eLife, 6, e04577.Find this resource:
Varga, A. G., Kathman, N. D., Martin, J. P., Guo, P., & Ritzmann, R. E. (2017). Spatial navigation and the central complex: Sensory acquisition, orientation, and motor control. Frontiers of Behavioral Neuroscience, 11, 4081.Find this resource:
Varga, A. G., & Ritzmann, R. E. (2016). Cellular basis of head direction and contextual cues in the insect brain. Current Biology, 26, 1816–1828.Find this resource:
Vitzthum, H., Müller, M., & Homberg, U. (2002). Neurons of the central complex of the locust Schistocerca gregaria are sensitive to polarized light. Journal of Neuroscience, 22, 1114–1125.Find this resource:
Warrant, E. J., & Dacke, M. (2016). Visual navigation in nocturnal insects. Physiology (Bethesda), 31, 182–192.Find this resource:
Warrant, E. J., Frost, B., Green, K., Mouritsen, H., Dreyer, D., Adden, A., . . . Heinze, S. (2016). The Australian Bogong moth Agrotis infusa: A long-distance nocturnal navigator. Frontiers of Behavioral Neuroscience, 10, 77.Find this resource:
Webb, B., & Wystrach, A. (2016). Neural mechanisms of insect navigation. Current Opinion in Insect Science, 15, 27–39.Find this resource:
Wehner, R. (1989). Neurobiology of polarization vision. Trends in Neuroscience, 12, 353–359.Find this resource:
Wehner, R. (2003). Desert ant navigation: How miniature brains solve complex tasks. Journal of Comparative Physiology A, 189, 579–588.Find this resource:
Weir, P. T., Henze, M. J., Bleul, C., Baumann-Klausener, F., Labhart, T., & Dickinson, M. H. (2016). Anatomical reconstruction and functional imaging reveal an ordered array of skylight polarization detectors in Drosophila. Journal of Neuroscience, 36, 5397–5404.Find this resource:
Wittlinger, M., Wehner, R., & Wolf, H. (2006). The ant odometer: Stepping on stilts and stumps. Science, 312, 1965–1967.Find this resource:
Wolff, T., Iyer, N. A., & Rubin, G. M. (2015). Neuroarchitecture and neuroanatomy of the Drosophila central complex: A GAL4-based dissection of protocerebral bridge neurons and circuits. Journal of Comparative Neurology, 523, 997–1037.Find this resource:
Wu, M., Nern, A., Williamson, W. R., Morimoto, M. M., Reiser, M. B., Card, G. M., & Rubin, G. M. (2016). Visual projection neurons in the Drosophila lobula link feature detection to distinct behavioral programs. eLife, 5, e21022.Find this resource:
Yarger, A. M., & Fox, J. L. (2016). Dipteran halteres: Perspectives on function and integration for a unique sensory organ. Integrative and Comparative Biology, 56, 865–876.Find this resource:
Zeil, J. (2012). Visual homing: An insect perspective. Current Opinion in Neurobiology, 22, 285–293.Find this resource:
Zeller, M., Held, M., Bender, J., Berz, A., Heinloth, T., Hellfritz, T., & Pfeiffer, K. (2015). Transmedulla neurons in the sky compass network of the honeybee (Apis mellifera) are a possible site of circadian input. PLoS ONE, 10, e0143244–25.Find this resource:
Ziegler, P. E., & Wehner, R. (1997). Time-courses of memory decay in vector-based and landmark-based systems of navigation in desert ants, Cataglyphis fortis. Journal of Comparative Physiology A, 181, 13–20.Find this resource: