Encoding Wide-Field Motion Characteristics in the Central Complex of the Cockroach, Blaberus Discoidalis

ENCODING WIDE-FIELD MOTION CHARACTERISTICS IN THE CENTRAL COMPLEX OF THE COCKROACH, BLABERUS DISCOIDALIS

NICHOLAS D. KATHMAN

Submitted in partial fulfillment of requirements

For the degree of Doctor of Philosophy

Advisor: Dr. Roy E. Ritzmann

Department of Biology

CASE WESTERN RESERVE UNIVERSITY

January 2015

CASE WESTERN RESERVE UNIVERSITY

SCHOOL OF GRADUATE STUDIES

We hereby approve the thesis/dissertation of

Nicholas D. Kathman

candidate for the Doctor of Philosophy degree*.

Committee Chair:

Hillel J. Chiel

Committee Member:

Roy E. Ritzmann

Committee Member:

Mark A. Willis

Committee Member:

Jessica L. Fox

Committee Member:

Daniel W. Wesson

Date of Defense: December 5th, 2014

*We also certify that written approval has been obtained for any proprietary material contained therein.

iii

Dedication

For my dad

Table of Contents

Thesis Summary ...... 1 Chapter 1: Introduction ...... 3 Summary ...... 4 Anatomy of the central complex of the insect brain ...... 5 Structural anatomy ...... 5 Neurochemical anatomy ...... 7 Sensory processing in the central complex of the insect brain ...... 9 Polarized light processing ...... 9 Retinotopic spatial information...... 11 Visual motion in identified cells ...... 12 Mechanreception ...... 14 Insect vision and visual interneurons of the brain ...... 15 Insect eyes ...... 15 Optic lobes ...... 17 Optic flow representing motion ...... 19 Ocelli ...... 21 Optomotor response ...... 22 The central complex and its role in behavior ...... 24 Locomotion control ...... 24 Courtship song and acoustic communication ...... 26 Visually guided behaviors...... 27 Basal ganglia homology ...... 28 My intended contribution to these fields ...... 29

Chapter 2: Encoding wide-field motion and direction in the central complex of the cockroach, Blaberus discoidalis ...... 34 Summary ...... 35 Introduction ...... 36 Methods ...... 38 Animals ...... 38 Animal preparation and electrophysiology ...... 38 Visual Stimuli ...... 40

Spike sorting and analysis ...... 41 Behavioral testing ...... 44 Procaine injection ...... 45 Histology and imaging ...... 46 Results ...... 47 Silencing neurons of the central complex reduces optomotor response ...... 47 Wide-field visual motion and the CX ...... 48 Response types to wide-field motion ...... 49 Unit response types and directional selectivity ...... 50 Sensitivity to stimulus temporal frequency ...... 51 Responses to wide-field motion with periodic firing ...... 54 Responses to vertical wide-field motion ...... 55 Discussion ...... 56 CX neurons convey diverse wide-field motion information ...... 56 Periodic responses to wide-field motion ...... 57 Response similarities with motion processing neurons upstream from the CX ...... 58 The role of CX circuits in optomotor responses ...... 59 Figures ...... 63

Chapter 3: Ocular dominance of directional responses to wide-field motion in neurons of the central complex of the cockroach, Blaberus discoidalis ...... 72 Summary ...... 73 Introduction ...... 74 Methods ...... 78 Animal preparation ...... 78 Visual stimuli ...... 79 Analysis ...... 80 Results ...... 82 General response classifications ...... 82 Temporal response type classifications in binocular conditions ...... 82 Ocular dominance testing ...... 83 Ocular dominance correlates with directional response for tonic and phasic responses ...... 84 Units with increased response amplitudes in monocular conditions ...... 86 Ocular dominance does not correlate with direction for entrained periodic responses ...... 88 vi

Few units completely lose response in monocular conditions ...... 89 Responses during control and ocellar inputs ...... 90 Units with changes in background activity ...... 91 Discussion ...... 92 Ocular dominance and directional motion in the CX ...... 92 Response similarities with motion processing neurons upstream from the CX ...... 94 Periodic neurons and neurons with background changes ...... 95 Increases in response amplitude compared to pre-binocular responses ...... 96 Responses during binocular occlusions and ocellar inputs ...... 98 Binocular vision, behavior, and the CX ...... 99 Figures ...... 101

Chapter 4: Conclusion ...... 110

Chapter 5: Appendix ...... 116

Bibliography ...... 119

vii

List of Tables

Chapter 5: Table 5.1: Activity levels of animals during treatment and control ...... 116

viii

List of Figures

Chapter 1: Figure 1.1: Schematic of the central complex (CX) ...... 8 Figure 1.2: Eyes of the cockroach ...... 16 Figure 1.3: Optic lobe ...... 18 Figure 1.4: Optic flow fields ...... 21

Chapter 2: Figure 2.1: Optomotor response is reduced after administration of local anesthetic in the CX ...... 63 Figure 2.2: Visual stimulus description and recording site locations ...... 64 Figure 2.3: Temporal properties of wide-field motion responses ...... 65 Figure 2.4: Distribution of temporal response types by horizontal direction and directional selectivity ...... 66 Figure 2.5: Changes in response duration with temporal frequency ...... 67 Figure 2.6: Temporal frequency response curves ...... 68 Figure 2.7: Sensitivity to direction could vary with temporal frequency ...... 69 Figure 2.8: Periodic responses to temporal frequency ...... 70 Figure 2.9: Responses to vertical motion ...... 71

Chapter 3: Figure 3.1: Visual stimulus description and recording site locations ...... 101 Figure 3.2: Response types to binocular visual motion in left and right directions ...... 102 Figure 3.3: Compound eye occlusions and effects on tonic and phasic responses ...... 104 Figure 3.4: Units with increased monocular and post-binocular responses ...... 105 Figure 3.5: Compound eye occlusions and effects on periodic responses ...... 106 Figure 3.6: Silenced responses during all occlusion conditions ...... 107 Figure 3.7: Responses during compound eye controls and ocellar occlusions ...... 108 Figure 3.8: Changes in background activity and responses with compound eye occlusions ...... 109

Chapter 5 Appendix: Figure 5.1: Baseline firing rates of all motion responding units by response type to each direction of motion ...... 117 Figure 5.2: Analog controls for digital stimulus ...... 118

Acknowledgments

I would especially like to thank my advisor, Dr. Roy Ritzmann. This has been a long, but wonderful journey, and Dr. Ritzmann has been a fantastic guide throughout. His unwavering support, instruction, faith, and patience have given me a critical eye and strength for my voice, without which this thesis would not be possible. I would also like to thank my committee, Dr.

Mark Willis, Dr. Jessica Fox, and Dr. Daniel Wesson. They have all touched this project in various ways throughout my degree. Mr. Alan Pollack has also been invaluable from the moment I started this endeavor. Every experimental idea first goes to Mr. Pollack for technical advice which invariably brings the idea from paper to reality in quick time. Dr. John Bender, Dr.

Josh Martin, Dr. Peiyuan Guo, Dr. Cynthia Harley, and Mrs. Ada Varga have offered extensive feedback, creative input, and support in all aspects of this work, from experimental design, analysis methods, and data interpretation. This work was supported by the National Science

Foundation [IOS-1120305 to R.E.R.].

Encoding Wide-Field Motion Characteristics in the Central Complex of the Cockroach, Blaberus discoidalis

Abstract

NICHOLAS D. KATHMAN

Wide-field motion is an important source of information used in visual control of

locomotion. It has been extensively studied in the optic lobes of flies and shown to be

sufficient to control flight maneuvers needed for course control. Additionally, the central

complex (CX), an associative in the arthropod brain, receives various forms of sensory

signals, including visual inputs, and is also involved in locomotor functions. Yet its

precise role in behavior is controversial. In this thesis, I used multi-channel extracellular

electrophysiology to record from large populations of neurons in the cockroach CX while

presenting wide-field visual motion that is varied across several parameters. The benefits

of long term recording of relatively large populations of neurons revealed the CX as a

structure that encodes rich and diverse visual motion information at a population level,

including characteristics such as speed, timing, and direction. Some of these sensitivities

were also been found to be conditional to others. For instance certain cells showed

directional selectivity, but only at narrow bands of stimulus speeds. This information may also contain localized regions of directional variability that correlate with specific rotational and translational movements. From experiments where individual compound eyes were occluded during stimulus presentations, certain neurons in the central complex

showed biases in response amplitude when only receiving input from one compound eye,

and these biases were correlated to the direction of motion inducing the response.

Finally, turning behavior was not only shown to correlate with wide-field motion, but could be inhibited by silencing the CX. These behavioral data were collected in conjunction with a colleague, Malavika Kesavan. These findings taken together, suggest the CX uses diverse wide-field visual motion, likely integrated with various other sensory information, in controlling locomotion. This system may work in parallel to a more direct visual motion control system, but is necessary for such control.

Chapter 1

Introduction

Summary

Navigating one’s environment is a rather complicated task, which involves assessing the surrounding environment and modulating motor commands to accommodate unexpected changes. All of these adaptations often take place in the context of some goal-oriented task, such as pursuing prey, finding one’s home, or evading a predator. While several sensory systems in insects have been found to directly modulate thoracic motor centers

(Strausfeld and Bassemir 1985; Iwano et al. 2010), the central brain also has associative structures that integrate different modalities of sensory inputs. These include the mushroom bodies and the central complex (CX) (Strausfeld 2012). The paired mushroom bodies receive massive input from the antennal lobes and play a key role in olfactory learning and memory (Mizunami et al. 1998; Heisenberg 2003). The CX, on

the other hand, is an unpaired midline structure which receives sensory input indirectly

from visual and mechanosensory systems (Strausfeld 1999; Homberg 2008) and has been

demonstrated to play a role in sensorimotor control of adaptive locomotor behavior

(Strausfeld 1999; Strauss 2002; Guo and Ritzmann 2013). Exactly what sensory

information is encoded in the neurons of the CX is still relatively unknown. Extensive

work has been done to demonstrate the CX as a map-like representation of polarized light

orientation in locusts (Heinze and Homberg 2007), crickets (Sakura et al. 2008), and

butterflies (Heinze and Reppert 2011). Visual motion signals were found in the CX in

response to both narrow features (Rosner and Homberg 2013; Seelig and Jayaraman

2013) and wide-field motion (Phillips-Portillo 2012; Weir et al. 2014). How extensively

this modality is represented is still unknown. Mechanical inputs from antennae also

project to the CX (Ritzmann et al. 2008).

In this introduction, I will provide background information that places my results in a wider context of sensorimotor integration and wide-field visual motion processing. I will

then describe my novel contributions to this field. I will also discuss the implications of having similar forms of wide-field visual motion information in both the optic lobes as well as the central complex, as possible parallel visual streams.

Anatomy of the central complex of the insect brain

Structural anatomy

The central complex (CX) is a group of interconnected midline neuropils found in the insect brain. Some variant of this structure is found in all insects, some crustaceans, and some annelids with bilateral appendages (Strausfeld 1999; Loesel et al. 2002;

Strausfeld 2012). Cells of the CX receive input from various sensory modalities, including visual and tactile antennal information (Strausfeld 1999; Ritzmann et al. 2008).

The CX consists of five structures: the fan-shaped body (FB) (also referred to as the

upper division of the central body), ellipsoid body (EB) (or lower division of the central

body), the protocerebral bridge (PB), and a pair of noduli (Pfeiffer and Homberg 2014).

In the cockroach, the PB is made of 16 columns (eight in each hemisphere), while the FB

and EB each have eight columns. Neurons connect these structures by means of a highly regular pattern of fiber crossings. Two additional neuropils called the lateral accessory

lobes (LALs) are attached to the CX. These structures lack a columnar arrangement and

carry ascending and descending information to and from the CX and thoracic ganglia.

Also in the LALs, outputs from the CX interact with other major pathways, including

heterolateral flip-flop neurons underlying searching actions of silk worm moths (Iwano et

al. 2010), that descend to make direct connections to the thoracic motor circuitry. This is

interesting, as this is a relatively simple sensorimotor circuit known to cause fast sensory driven tracking behaviors, yet olfactory processing in the mushroom bodies also integrate this information for other behavioral tasks. This is similar to the optomotor system

discussed in this thesis. Visual motion information is processed in the optic lobes and

projects directly to the motor circuitry, yet the associative central complex also receives

this information to be used for behavior regulation. Additionally, a pair of globular

neuropils (also without a columnar arrangement), called the noduli, are posterior-

ventrally attached to the EB and share connections between the EB, FB, and PB with

inter-hemispheric crossings (Pfeiffer and Homberg 2014).

The CX shares connections with many areas in the protocerebrum and some direct

connections to the optic lobes (Homberg et al. 1991). Otherwise, there are no known

direct connections between the CX and primary sensory areas (Pfeiffer and Homberg

2014). There are also connections found between the CX and the nearby mushroom

bodies (Strausfeld 1976; Phillips-Portillo 2012). Within the CX, there are four principle

cell types, classified through extensive work in both locusts and flies (Müller et al. 1997;

Heinze and Homberg 2008; Lin et al. 2013). First, tangential cells innervate the PB, EB,

FB, and noduli (Fig 1.1, red lines). Arborizations on these neurons outside the CX are

usually smooth but bear spines, suggesting input regions, while those within the CX are

6 varicose, suggesting output areas. Therefore, these cells are thought to be the main input neurons of the CX (Heinze and Homberg 2009). They arborize within various regions of the protocerebrum and are thought to distribute sensory inputs or ascending inputs from the thoracic ganglia to many or all of the columns of the CX substructures (Strausfeld

1999). The second cell type is the columnar neuron (Fig. 1.1, black lines). In contrast to the tangential cells, the columnar cells are thought to be the major output cells of the CX.

These cells either link the PB to other substructures of the PB or connect the FB to the

EB (reported only in flies), noduli, or LALs (Hanesch et al. 1989; Heinze and Homberg

2008). They occur in sets of 8 or 16, and may cross the midline, connecting contralateral columns of PB and FB. The final classes of cell types are intrinsic neurons called the pontine and amacrine neurons. Amacrine neurons are rare and innervate several columns of the FB. Pontine cells are more commonly found, and innervate only two columns lateral columns of the FB (Heinze and Homberg 2008).

Neurochemical anatomy

A large number of neurotransmitters have been found in neurons of the CX.

Glutamate and acetylcholine appear ubiquitously across substructures of the CX, while γ- aminobutyric acid (GABA), dopamine, octopamine, and histamine is associated with specific cell types (Kahsai and Winther 2011; Kahsai et al. 2012; Pfeiffer and Homberg

2014). GABA is the predominant transmitter of tangential cells of the EB and some tangential cells of the FB in several species (Homberg et al. 1987; Homberg et al. 1999).

Histamine, dopamine, and their receptors are found in some tangential cells of the FB,

7 while octopamine and its receptors are found in neurons ascending from the subesophageal ganglion to the FB. Serotonin and associated receptors are also found in various cells types, including columnar and tangential neurons (Homberg 1991; Kahsai et al. 2012).

Neuromodulators are also found in the CX in a variety of insects (Nässel and

Homberg 2006). The fly has at least eight classes of neuromodulatory neurons found in the FB (Kahsai and Winther 2011) and grasshoppers have six classes in the PB and LAL

(Herbert et al. 2010). These modulators likely report states of external stimuli and internal physiology to the CX circuitry. In flies some of these modulators influence locomotor behaviors (Kahsai et al. 2010). For example, flies deficient in the neuropeptide tachykinin showed a deficit in avoidance behaviors to open spaces and the short neuropeptide F played a role in regulating overall motor activity.

Figure 1.1. Schematic of the central complex (CX). Left: The CX consists of four substructures that are centered in the protocerebrum. Right: These four substructures are organized in a modular arrangement. The protocerebral bridge (PB) consists of 16 columns, while the fan-shaped body (FB) and ellipsoid body (EB) have 8 columns each. The inputs are tangential neurons (red) that project to each of the substructures. Columnar neurons interconnects the substructures and project to descending neurons (DN) of the lateral accessory lobes (LAL) as the outputs of the system.

Sensory processing in the CX of the insect brain

The modular nature of the CX with bilateral inputs and outputs suggests a possible role in relaying spatial or bilateral information. Two visual systems carry this type of information, the polarized light system and a feature detection system.

Polarized light processing

The polarized light mapping system is arguably the most comprehensively studied function of the CX. Polarized light is a feature of the atmosphere that indicates the location of the sun through angles of rotation (E-vector) patterned in the blue sky. Many insects, particularly ones that have long distance migrations, use this pattern for orientation and navigation (Wehner 2001). Through neurons studied specifically for their selectivity to E-vector angles, many of the known cell types of the CX have been characterized anatomically and functionally (Heinze and Homberg 2008; Pfeiffer and

Homberg 2014).

Most notably, the columns of the PB have been found to contain a map-like representation of polarized light E-vector orientation in locusts, crickets, and monarch butterflies (Heinze and Homberg 2007; Sakura et al. 2008; Heinze and Reppert 2011).

Through extensive studies using intracellular recordings, where cells were selected specifically for their sensitivity to linearly polarized light, much of the entire polarized light circuit leading to the CX has been mapped. In these studies cells are recorded and

filled with dyes, indicating not only their location, but what their inputs and output

regions are, depending on the anatomical characteristics of the dendrites.

Groupings of photoreceptors in a localized region of the compound eye (dorsal

rim) are arranged such that they are sensitive to particular E-vectors (Labhart and Meyer

1999). This signal passes through the lower anterior optic tubercle via TuLAL1a/b

neurons, which synapse on projections from the LALs and then enters the CX via

tangential neurons (TL2 and TL3) to both the EB and FB (Mappes and Homberg 2007).

In the CX, columnar neurons connect individual columns of the EB and FB to columns of the PB. The CL1 columnar neurons which have restricted arborizations in single columns

transmit polarized light information from the EB to the PB. At the output stage, TB1

tangential neurons of the PB integrate polarized light information from several CL1 neurons

across both hemispheres. Here, the topographic E-vector map is found (Heinze and

Homberg 2007). These output cells from the PB are specifically tuned to a range of E-

vector orientations through modulated firing rate. These cells represent a wide range of

rotation angles and are topographically arranged so that neighboring columns have the

most similar angles. Moreover, columnar neurons such as CPU1 and CP1 also have a topographic representation of e-vectors. This feature is possibly obtained through their connections with PB tangential neurons, project to the LAL, where they can influence descending fibers.

In addition to encoding specific angles of rotation, tangential cells in the CX and

LALs integrate rotation with spatial information (Bech et al. 2014). E-vector sensitivity

for these cells changes with stimulus location, systematically following the global

structure a specific circular E-vector pattern associated with sun position. Most of the

polarization sensitive neurons in this system receive bilateral inputs, with the exception

of one tangential cell of the EB that only receives ipsilateral input (Heinze et al. 2009).

It is important to note that, although this map has been hypothesized as part of a

sun compass used in the migration of some of these animals (Reppert et al. 2010), no

direct behavioral evidence has shown that the polarized light information encoded in the

CX participates in driving any specific behavior. This is primarily due to the limitations

of the recording techniques used for many of the polarized light neural studies. Also, even though responses to polarized light by corresponding central complex neuropils in two species is well established, this modality may be less well represented or even unrepresented in other nocturnal or cave dwelling insects that also have very well developed CX neuropils (Strausfeld 2012).

Retinotopic spatial information

The other topographic visual representation in the CX is a feature detection

system found in flies. Flies use vision as a primary feedback system for course control in

flight, which is discussed in further detail in the following section. Also, flies display

goal-driven orientation behaviors to vertical stripes moving left or right in their visual

field (Neuser et al. 2008). This task requires spatial information as well as spatial

working memory. Thus, it implies the insect’s capability to retain, recall, and integrate

positional information as the target moves in and out of sight. Input cells of the EB,

called ring neurons, are necessary for this task. Ring neurons are analogous to tangential

cells of the EB; a structure that is actually ellipse-shaped in the fly. Ring neurons also

encode retinotopic spatial information (Seelig and Jayaraman 2013). These cells arborize onto microglomeruli in the lateral triangle (LTR) of the LALs and have receptive fields with both excitatory and inhibitory subfields, resembling those of simple cells in the mammalian primary visual cortex (Bonin et al. 2011). Also like simple cells, ring neurons show orientation tuning to vertical bars and some with directional selectivity.

Ring neurons’ receptive fields cover large parts of the visual field, but with densities localized to regions of the ipsilateral eye and spatially correlated with the position of the microglomeruli. These responses also are reduced during flight but not with wing-beat patterns or turning, nor during walking (Seelig and Jayaraman 2013).

Responses to visual motion in identified cells

Although no studies have directly looked at wide-field visual motion responses in the cells of the CX, along with the optical imaging study mentioned above, a few studies have examined visual motion responses to moving narrow bars. One such study used intracellular recordings to assess response properties of EB, FB, and PB cells to single bars and stripe fields oriented in 45° rotational increments in the fly, Neobellieria bullata

(Phillips-Portillo 2012). PB inputs cells and pontine cells were identified that demonstrated motion related responses. Interestingly, no tangential cells showed responses to motion and only one of 14 showed a response to an increase in light intensity. The one PB input cell had dendrites extend to optic glomeruli of the optic lobe that receives input from the lobula plate (Strausfeld and Bassemir 1985). This cell showed directional selectivity to a moving bar, responding only to leftward motion.

Pontine cells also showed directional motion responses, but are intrinsic to the CX with

no direct connections to the optic lobes. None of these cells were sensitive to polarized

light.

One additional study tested small moving objects along with looming stimulus

which is the visual representation of an approaching object (Rosner and Homberg 2013).

This study was performed in the locust, with visual stimuli presented independently to

each eye, but not truly optically isolated. The more laterally positioned compound eyes of the locust, as compared to flies and cockroaches, allowed the investigators to point their stimuli at one eye. Nevertheless, it is possible that some receptors from each eye still detected contralateral stimuli. Columnar, tangential, and pontine neurons were sensitive to looming stimuli from both the ipsi- and contralateral eyes. A subset of tangential and columnar cells showed spatially antagonistic responses, where looming from one eye resulted in an excitatory response, but looming from the other eye resulted in inhibitory responses. When tested with small dark and light moving squares, all three cell types showed responses to motion, which always coincided with looming sensitivity.

Nevertheless, pontine and tangential cells were not directionally selective, while one columnar neuron of the FB was.

Together, these studies do not show much consistency, particularly between directionality and morphological cell type. Yet responses to visual motion were found in most cell types of the CX. Both of these studies employed intracellular techniques and therefore were limited in the number of neurons sampled as well as how much replication could be presented for stimulus parameters. This technique is excellent for assessing responses to a single identified neuron but cannot examine populations of neurons

simultaneously, nor can it allow the investigator to extend recordings over long periods of

time, while the sensory apparatus of the insect is manipulated. This is significant, as

subtle rate changes or precisely timed spiking information may go unnoticed. Large

populations of CX neurons can be recorded over long time periods in flies using optical

methods (Seelig and Jayaraman 2013), but currently, this technique requires that the

subject be restrained.

Mechanoreception

Although much work has focused on visual information, the CX is not exclusive

to visual processing. In fact, the CX is present in several blind species of arthropods

(Strausfeld 2012). Mechanosensory information processing has also been demonstrated

in the CX (Ritzmann et al. 2008). These studies were preformed in cockroaches and used

similar multi-channel techniques as those in this thesis, and therefore sampled large

groups of cells in the CX, simultaneously over several hours. Cockroaches use antennal cues to follow walls at constant distances (Camhi and Johnson 1999), show deficits in orientation after shaving off the hair plates from the antennae (Okada and Toh 2001) and rely heavily on antennal inputs when climbing over objects (Harley et al. 2009). Also, in tethered walking, deflecting an antenna consistently generates turning leg movements

(Mu and Ritzmann 2005). Considering the large emphasis on mechanoreception in cockroach behavior, it is expected that a sensorimotor region like the CX should encode this information. By deflecting each antenna in various combinations of direction and velocity, a large population of neurons was found encoding directional mechanosensory

motion. Cells in both the FB and EB receive inputs from both the left and right antennae

in the cockroach, with some cells being directionally selective while others were not.

Directional selectivity was observed at the level of a single antenna as well as both

antennae, as some cells responded to only one antenna or showed significantly stronger responses to one antenna. Most of these cells were also sensitive to changes in ambient light intensity, suggesting that many if not most CX neurons are multi-sensory in nature.

Intracellular studies have also shown cells in the CX of flesh flies (ring neurons and FB tangential cells) respond to flashes of light, visual motion and puffs of air on the head, believed to be sensed by the antennae (Phillips-Portillo 2012). Thus, it may be an over- simplification to think that the CX circuits are dedicated to one sensory modality such as polarized or non-polarized light.

Insect vision and visual interneurons of the brain

As described earlier, the CX integrates several sensory modalities that are important in insect locomotion and navigation. Vision, and specifically visual motion, has classically been a major field of study in sensory systems role in locomotive control and is the focus of this thesis. Thus, a review of the insect visual system would be useful in understanding the experiments that are described in this document.

Visual feedback is critical in behaviors such as flight stabilization, object tracking, self-motion measurements and distance travelled estimates (Egelhaaf and Kern,

2002; Srinivasan et al., 1999; Borst and Egelhaaf, 1989). Although less is known about if or how visual motion is processed in the CX, much work has been done to describe how

it is processed in the compound eyes and optic lobes, where motion characteristics are

first encoded.

Insect eyes

The visual sensory path begins at the eyes (Fig. 1.2). The both sets of eyes in the

cockroach are set in the front of the head. The compound eyes wrap around the posterior

lateral sides of the head, providing an almost complete panoramic visual field, with 40°

of dorsal overlap, along with 65° anterior overlap, and 56° posterior overlap (Butler

1973). Insect compound eyes consist of many facets called ommatidia, which are

arranged in a hexagonal array. Each cockroach compound eye contains 2000 ommatidia,

each with a corneal lens with an optical axis directed at a different point in space,

collecting light from 2.4° of visual space. Visual acuity is highest in the in the anterior

binocular overlap field.

Compound Eyes

Ocelli

Figure 1.2. Eyes of the cockroach. Both sets of eyes sit on the front side of the cockroach head. The compound eyes wrap around laterally behind the antennae and the ocelli are medial to the antennae.

Unlike many other nocturnal insects, the cockroach eye is an apposition eye type,

meaning each ommatidium is optically isolated. Although this system is less suited for

low light conditions, it is beneficial for motion detection. This is also the architecture

found in most dipteran species, which are classic motion detection models and the basis

of much of the vision background presented here. Also like the fly, the cockroach has

evolved a neural superposition arrangement, where 6-20 photoreceptor signals are pooled

at the neural level, into second-order cells which aids in amplifying the dim light signals

(Heimonen et al. 2006). Flicker fusion rates differ between cockroaches and flies. The

flicker fusion rate is the threshold frequency that intermittent light stimulus can be

perceived in the retina before fusing to one continuous signal. Unlike the fly, the flicker

fusion rate in the cockroach is relatively slow at 45-60 Hz (Guthrie and Tindall 1968).

The optic lobes

The retinotopic signals, while maintaining their spatial organization, are

processed by four successive optic ganglia—the lamina, medulla, lobula, and lobular

plate (Fig. 1.3). By maintaining their retinotopic arrangement, neighboring optical

components can be processed for spatial and directional information (Borst and Haag

2002; Borst 2014). Local motion detection first occurs in the medulla. One generally

accepted model for motion detection in biological systems the Hassenstein-Reichardt

model (Hassenstein and Reichardt 1956). This is a correlation-type model that consists of at least two inputs passing through asymmetrical channels and combined with a nonlinear element (Borst and Egelhaaf 1989). Two of these half detectors are combined

17 with mirror symmetry to form a directionally selective elementary motion detector

(EMD).

Figure 1.3. Optic lobe. Above: Three-dimensional schematic of the retina, optic lobe, and a tangential cell (blue). The columns in each layer are retinotopically arranged and represent ommatidia of the retina in a one-to-one ratio. Orange and yellow columns trace the path of two distinct tracts. Lower left: Two tangential cells (LPTCs), one pooling horizontally for horizontal motion and one pooling vertically for vertical motion. Lower right: intracellular recordings from a horizontal cell (HSE) during motion in its null and preferred direction. Figures and text were adapted from Borst and Haag, 2002 with permission.

Large motion sensitive cells, called lobular plate tangential cells (LPTCs), project

onto the lobular plate with large dendritic arborizations (Fig 1.3). Because of their size,

they can be recorded from relatively easily in the fly. There are approximately 60 of

these cells identified in the blowfly Calliphora (Hausen 1984). These cells have large

dendrites that spatially integrate various subpopulations of local EMDs. According to

their overall preferred direction, they are grouped into horizontal and vertical cells, which

are further subdivided into numerous specific directional motion fields. Most LPTCs

show a tonic spiking activity in response to motion along their preferred direction and are

hyperpolarized in response to motion in the null direction (Borst et al. 2010)(Fig. 1.3).

In the fly, three major routes are known for the visual motion information coming

out of the optic lobes. First, columnar neurons from the lobula project to various

glomeruli in the protocerebrum (Mu et al. 2012). Second, LPTCs synapse onto

descending neurons that connect, via the cervical connectives, to the motor centers in the

thoracic ganglion (Strausfeld and Bassemir 1985). Finally, LPTCSs could connect to neck motor neurons for controlling head movement (Strausfeld and Seyan 1985). It is unknown whether the cockroach possesses similar anatomical connectivity.

Optic flow representing motion

Visual motion cues often result from relative movement between the eyes of the

observer and the visual structures of the environment. The resulting motion pattern is

referred to as “optic flow” (Gibson 1950). A fly’s movements (or that of any other flying

object) can be described as a combination of six simple movements (e.g. Fig 1.4): three

translational directions along each of the body axes (thrust, slip and lift) and three

rotational movements around the same axes (roll, pitch and yaw). Each rotational or

translational movement generates a characteristic optic flow pattern, which can be used to

determine and maintain body orientation and flight directory. A characteristic of many of

these movements are complex patterns of directions of motion in different locations of

the receptive field. Some LPTCs have very complex dendritic trees that curl across

different localized regions with specific preferred directions. These cells act as matched

filters and result in sensitivities to more complex optic flow patterns (Krapp 2000).

These curled flow patterns are behaviorally relevant, as the neurons respond maximally

to certain flight maneuvers, such as a roll (Fig 1.4) (Karmeier et al. 2006).

Individually, all 60 of these classified neurons produce directionally selective

responses in different regions of the receptive field of the ipsilateral eye only, but

interactions among LPTCs shape their receptive fields. Of the cells responding to horizontal motion, a subset of 12 LPTCs extend their axons across the brain to synapse onto the contralateral lobula plate and respond to motion of the contralateral hemisphere

(Krapp et al. 2001; Farrow et al. 2006). Therefore, these cells are able to combine information from each eye even in the peripheral visual processing centers and thus possess spatially distinct fields. These cells have a larger response when binocular

horizontal motion is moving in the same direction and a smaller response when binocular

motion is moving in opposite directions (a different direction for each eye). The former

condition relates to the optic flow produced when a fly rotates about its vertical body

axes (yaw rotation) or when a fly translates sideways in the horizontal plain (slip

translation). The latter condition relates to forward and backward translation (thrust

translation). Therefore, LPTCs are capable of responding to complex flow patterns across

the entire visual field.

Figure 1.4. Optic flow fields during flight. Left: Flow field in a natural scene during forward translation. This flow field is more complex than a shifting stripe pattern. This scene has localized flow patterns, for instance the left and right sides of the scene have opposing directional vectors. Right top: Flow field during a roll maneuver. Right bottom: Receptive field of an LPTC in the optic lobe (Huston and Krapp 2008).

Ocelli

In addition to compound eyes, most insects have another set of much more simple photosensitive organs called ocelli (Fig. 1.2). Many flying insects, like flies and locusts, have three ocelli that are arranged in a triangle on the dorsal part of the head, while the cockroach has two, each medial to an antenna (Guthrie and Tindall 1968). The ocelli

have a very high light sensitivity, particularly to UV light (Wilson 1978). Yet, their

increased sensitivity is at the expense of spatial resolution (Mizunami 1993). Due to

underfocussing of the lenses, ocelli are likely used to measure small changes in light

intensity rather than motion detection or image formation. Although no known direct

projections exist from the ocelli to the CX, neurons do project to primary visual regions,

like the optic lobes, that are upstream from the CX, as well as other associative regions,

like the mushroom bodies (Mizunami 1995). Along with their high speed of signal

transmission, these qualities make them well suited for initiation and cessation of diurnal

activities, flight stabilization, navigation, and orientation behaviors (Mizunami 1995;

Taylor and Krapp 2007). Moreover, they play a critical role in the cockroach’s decisions

to either climb over or tunnel under a shelf (Harley et al. 2009). In bright light there is a

much greater tendency to tunnel. This bias goes away under darkened conditions or in

bright light when the ocelli are covered. Covering the compound eyes do not affect these

decisions.

Optomotor response

Motion stimuli parameters are relevant to the moving insect (typically in fly

flight), not just potential codes that are never decoded (Borst 2014). Most studies have

focused on experiments in which a fly is tethered in the middle of visual scenery, while measuring behavioral responses such as wingbeat asymmetry (Reiser and Dickinson

2008). This is an advantageous technique because it allows the visual stimulus to be

precisely presented to the animal and it also isolates the visual response component from

other proprioceptive information.

The optomotor response is a behavior typically studied in this manner. When the

visual scenery consist of some textured scene that globally rotates horizontally across the

entire visual field of the animal, such as a rotating striped drum, the animal rotates in the

same direction (Gotz 1968). The optomotor response is thought to stabilize the fly’s gaze

and, thus, its flight course against external disturbances. It is also interesting that this

response is highly conserved throughout many animal groups (Suthers 1966 - bats;

Mitchiner et al. 1976 - mice; Steinman and Collewijn 1980 - humans; Portugues and

Engert 2009 - zebrafish; Borst et al. 2010 - flies). In all of these groups, rotational motion in an animal’s visual field evokes either eye movements (Mitchiner et al. 1976; Steinman and Collewijn 1980; Miles 1997), head movements (Fox and Frye 2014), or body rotation

(Borst et al. 2010) in order to stabilize the gaze in the optic field. It is important to note that this is a distinct behavior from goal-oriented turning behaviors which involve higher-

order processing (Guo and Ritzmann 2013).

Due to their strong directional sensitivity, large receptive fields, and connectivity to

the thoracic ganglia, the LPTCs have long been proposed as the neural control elements

for optomotor response (Borst et al. 2010). Also, optogenetic stimulation of horizontal

cells induce yaw head movements and turning responses (Haikala et al. 2013).

The CX and its role in behavior

Locomotion control

As a central hub with connectivity to both sensory and motor regions of the

nervous system, the CX is a prime candidate for studying sensory motor integration as well as a command center for goal-oriented behaviors. Traditionally, these studies have

been limited to circuit manipulations such as lesioning, either through mechanical,

electrical, genetic or pharmacological methods, or electrical stimulation.

Early studies probed the CX for its role in various behaviors, including walking, flight, and even cricket song. Its role in locomotion was first demonstrated with evidence from surgical lesioning and stimulation (Huber 1960; Otto 1971). Although the neural control for straight, coordinated walking derives from central pattern generators and inter-joint reflexes in the thoracic ganglia (Ritzmann and Büschges 2007; Büschges et al.

2008), the CX was shown to increase walking activity with stimulation and decrease activity with lesions. More recently, neurogenetic tools used in Drosophila have been

useful for further exploring this role. More specific components of walking were

correlated with CX mutations, including the initiation and maintenance of walking, stride

symmetry, and visual orientation (Strauss 2002). Further manipulations either by genetic mutations or expression of tetanus toxin in the CX also revealed reductions in walking

duration (Martin et al. 1999) and walking speed (Poeck et al. 2008).

The CX has also been studied in the context of goal-directed locomotor behavior.

In the cockroach, the CX has been linked to turning behavior. After lesioning the CX

with foil lances, either frontally or sagitally, cockroaches exhibited significant deficits in

turning behavior (Ridgel et al. 2007). Animals could often turn in one direction and walk

straight, but would never turn in both directions and often turned into obstacles. Some

animals, on the other hand, continuously turned in circles. Electrolytic lesions in the FB

affected turning in a U-shaped arena (Harley and Ritzmann 2010). Typically in this

arena, when the animal encountered a wall with its antennae, it turned away from the antenna that made first contact. Lateral lesions in the FB produce significant increases in wrong turns (i.e. into the wall). On the other hand, insects with midline lesions made no turning mistakes, but did show climbing deficits.

More recently, through multi-channel extracellular recordings, work on the role of the central complex in these behaviors has made great strides. The primary means of recording from neurons in the central complex, both intracellular recording and optical recordings in genetic models are limited when dealing with moving animals. Through wire-tetrode extracellular recordings, as well as remote telemetry, animals are able to move, either with some motion limitations on a tether or even freely walking while not disturbing electrode connectivity (Guo et al. 2014). Multi-channel recordings using tetrodes take advantage of electrical transduction properties in neural tissue, which allow spatial comparisons of electrical fields created by neural activity to be made at four electrodes that make up a tetrode. The differences in recordings at these four sites allows one to separate neural signals based on the location, size, and morphology of the source neuron and distinguish large ensembles within a network (Buzsáki 2004). One can therefore record from relatively large populations of cells simultaneously for long periods of time even in behaving insects.

This technique has been used in some recent studies of cockroach locomotion. In

the cockroach, electrophysiological studies on tethered walking showed that activity in neurons of the CX were correlated with, and often predictive of, stepping frequency

(Bender et al. 2010). In the same study, stimulation of these neurons also increased stepping rate. In another example of multi-channel electrophysiology with flexible wire- bundles, recordings of CX neurons in the FB while walking on both an air-supported ball

(Guo and Ritzmann 2013) or while freely walking (Guo et al. 2014) showed that activity

in the CX also correlates and predicts turns and climbing behavior as well as walking

speed. These cells showed activity which correlated with discrete turning speeds and

orientations and were recorded from in the hemisphere ipsilateral the predicted turning

direction. Furthermore, stimulating these neurons induces turns, also in an ipsilateral

direction to the stimulation site.

Courtship song and acoustic communication

In addition to locomotion, the CX also plays a role in other behaviors, such as

initiation of sound production in the grasshopper (Heinrich et al. 2012). Stridulation, a

sound producing behavior used for courtship, mating, and rivalry, can be induced when

injecting cholinergic agonists into the CX. This behavior matched natural behaviors

quite well, including temporal structure, right/left coordination, sequencing, and other

body movements that coordinate with the stridulation. Flies also show a relationship

between the CX and song production, as mutants exhibit variable, unstable song patterns

with unusual temporal properties (Popov et al. 2003). Additionally, silencing FB neurons

reduces courtship activity and acoustic communication (Popov et al. 2005; Sakai and

Kitamoto 2006).

Visually guided behaviors

Visually guided gap crossing is affected by CX mutations (Triphan et al. 2010).

Gap crossing involves initiating a sequence of behaviors after visually estimating if the gap is insurmountable or not (Pick and Strauss 2005). With PB disruptions, flies properly access gap widths and begin to cross them, but fail at orienting themselves in the correct direction. The orientation component was partially restored when genomic rescue constructs that restored the PB were tested. Other climbing behaviors in flies are modulated by the CX, as silencing ring neurons of the EB lead to abnormal behavior in a gravitaxis maze, where flies make a series of up/down decisions to reach a light source

(Baker et al. 2007). Similar results have been shown in electrolytic FB lesion experiments in the cockroach, abnormalities of obstacle climbing were seen, such as decreased success rate, altered strategies that show less controlled rearing, and delayed response to the obstacle (Harley and Ritzmann 2010). These behaviors use both antennal and light information to make decisions about climbing or tunneling shelf-like obstacles (Harley et al. 2009).

The relationship between the CX and visually guided flight is unclear. Mutant flies with structural defects in the PB, EB, and FB all show deficits in strait walking as well as flight control (Ilius et al. 1994). Flight behavior alterations included reductions in the amplitudes and frequency of turning saccades, increases in optomotor reversal time

27 with changes in direction, and less persistent stripe fixation when compared to wild type flies. However, there is more evidence that demonstrates that visual information in the

CX is modulated by flight behavior. For instance, feature detection of some EB neurons were diminished in flight, but not during walking (Seelig and Jayaraman 2013), while a group of FB neurons were shown to be unresponsive while the fly was quiescent but respond to translational optic flow during flight (Weir et al. 2014).

Basal ganglia homology

Much like insects, mammals also must select actions based on sensory information (intrinsic and extrinsic) related to the proper context for that behavior. The basal ganglia is the suggested vertebrate structure that accomplishes this task (Grillner et al. 2008). A recent meta-analysis suggested that the insect CX is deeply homologous to the mammalian basal ganglia (Strausfeld and Hirth 2013). Functionally, the basal ganglia are involved in voluntary motor actions, including eye movements (Kandel et al. 2013), a suggested role for the CX. Lesions in and diseases of the basal ganglia lead to movement control disorders such as Parkinson’s disease and dystonia (Bhatia and Marsden 1994;

Redgrave et al. 2010). Anatomical similarities more strongly link this structure to the

CX. The basal ganglia consists of the striatum, globus pallidus, subthalmic nucleus, and substantia nigra. Like the CX, the striatum is modular, and receives information about both internal physiological states as well as sensory space from various modalities

(Flaherty and Graybiel 1994). The pallidum is also similarly organized to the EB, where

GABAergic neurons extend from the EB to the LALs with reciprocal glutamatergic

pathways (Kahsai et al. 2012). In vertebrates, GABAergic neurons connect the pallidum

to the subthalmic nucleus and the globus pallidus, also with reciprocal pathways.

Additionally, direct pathways project from the motor cortex to spinal tracts or, in the case

of the oculomotor system, the superior colliculus, but a parallel pathway through the

basal ganglia influences the strength of those commands (Alexander and Crutcher 1990).

Within the basal ganglia, direct and indirect pathways have opposite effects leading to

selection of appropriate behaviors at different times in the animal’s experience.

Another role of the basal ganglia is for integrating efference copies with other

afferents from the cortex and limbic system to contextualize motor commands relating to

ongoing goal-oriented actions (Redgrave and Gurney 2006; Fee 2012). This relates to a

common pattern in nervous systems where distinct neural pathways that convey

“outcome” and “context” are evaluated and reinforce behaviors in the proper context.

Experiments in songbirds suggest that vocal-related basal ganglia circuitry receives two

functionally distinct excitatory inputs (Fee and Scharff 2010). One input is from a cortical

region that carries context information about the current “time” in the motor sequence.

The other is an efference copy of motor commands from a separate cortical brain region

that generates vocal variability during learning. These are integrated where the efference

copy gates the learning rule for a given motor command in the given context.

My contribution to these fields

In this thesis, I present a far more thorough investigation of the wide-field motion information that is encoded in the CX than was previously available. My findings are

then put in the context of: 1) the behavioral relevance of this information and 2) a comparison of motion information that is contained in CX circuits with that which is also present upstream in the optic lobes. My data show that many of the wide-field motion characteristics that are found in the optic lobe are also found in the CX. As a population of cells found in various regions of the FB and EB, the neurons recorded encode for speed, timing, direction, and bilateral spatial information. All of these characteristics provide useful visual feedback for course control (Srinivasan et al. 1999; Borst 2014). As discussed earlier, the apparent motion of the visual environment provides information about ego-centric motion, such as rotational and translational velocity, and is known to invoke course correction responses when manipulated. The flow fields also can be spatially complex, with various directions of motion in specific regions of space. This observation is relevant because the cells found in this study were selective to direction and tuned to speed and all of these cells show directional bilateral biases in their responses to motion. Many cells receive weighted input for a direction of optic flow separately from each compound eye. Therefore, CX cells could potentially differentiate localized directional flow fields on each side of its visual field. This is useful information about complex global flow fields corresponding to body movements (see Fig. 1.4). This is an effect seen only in a subset of LPTCs (Farrow et al. 2006), but is a parallel example of integrating primary sensory information to encode more abstract information about the environment, or one’s own movements.

As I have said above, this visual motion information is also found in the optic lobe and has been shown to be sufficient to encode all relevant rotational and translational self-motion components and, therefore, control most visually guided course

30 control tasks in flies (Karmeier et al. 2006). Stimulation of LPTC cells also induces turning behaviors (Haikala et al. 2013). Yet additional data presented in this thesis, acquired in conjunction with a colleague, Malavika Kesavan, shows that a functional CX is also required for initiating visually guided behaviors. Therefore, I suggest that these two representations of wide-field visual motion, in both the optic lobes and the CX, are possibly part of a parallel network of motion inputs to the thoracic motor circuitry. The first branch offers both fast, direct course control inputs from the optic lobes to the thoracic motor centers, while the second parallel pathway, also collecting motion information from the optic lobes, is integrated in the CX with other relevant intrinsic and extrinsic sensory information and then descends to the thoracic circuitry to potentially alter behavior in a context dependent fashion. This model is similar to circuits seen in the basal ganglia of mammals (Alexander and Crutcher 1990; Hikosaka et al. 2000; Kandel et al. 2013), where direct pathways project from the motor cortex to spinal tracts or, in the case of the oculomotor system, the superior colliculus, but a parallel pathway through the basal ganglia influences the strength of those commands.

This study uses multichannel recordings in the context of visual motion processing, which is novel for this work in the CX. As discussed earlier, much of the work in the CX regarding motion vision, circuit mapping and polarized light studies in the central complex used intracellular recordings. This is an excellent technique, which allows not only electrically and temporally high resolution assessments of neural activity, but also can provide morphological and connectivity information. One limitation of the extracellular technique used in this thesis is the inability to identify morphological cell types that are the source of our recordings. For instance little can be said about the

anatomical classifications of the cells in this thesis, but many were found at the lower

margins of the EB, where large integrating regions of the EB tangential cells reside. This

is somewhat contradictory to the previous study of the fly CX where no tangential cells

were found to have motion responses (Phillips-Portillo 2012), which should be followed up with intracellular recordings in the cockroach.

Nevertheless, intracellular recordings have their shortcomings. They are far less stable than extracellular techniques. As a result, long term recording or recording during natural behavior is difficult or, in some cases, impossible. This is in addition to the sampling limitation, as only one neuron can be recorded from at a time and there is a sampling bias toward larger neurons. These limitations, specifically the duration and sampling limitations, are not shared with the extracellular techniques used in this study.

First, multi-channel recordings are highly stable, and can last for many hours, and some cases days. This allows one to test large, highly replicated yet diverse stimulus suites while probing much of the parameter space relevant to the sensory systems we examined.

Additionally, the longevity of the recordings allowed me to also look at changes in neural activity over long periods of time. In several units in this study, responses to a given stimulus changed from the beginning of the experiment to the end. Secondly, this technique has the ability to sample many more neurons in a recording. A fundamental principle of neurobiology is that complex signals in the brain are often represented across distributed populations of highly interconnected neurons (Sanger 2003). In order to understand how information is encoded in a network of neurons, it is useful to record simultaneously from many neurons of a structure. By using multi-channel recording, I record from up to 30 neurons simultaneously, distributed across four separate, precisely

32 spaced regions of the CX. In the experiments described in chapter 3, we were able to manipulate the eye coverings while maintaining our recordings. These experiments would have been very challenging for intracellular studies.

Although no behavioral recordings were made in this study, the recordings and data taken here can be followed up with identical recordings on a tether or in freely behaving insects moving in an arena as was done previously for antennal studies (Guo and Ritzmann 2013; Guo et al. 2014). Similar successes would allow one to correlate CX activity described in this thesis with various components of visually directed locomotion.

Chapter 2

Encoding Wide-field Motion and Direction in the Central Complex of the Cockroach, Blaberus discoidalis

Summary

In the arthropod brain, the central complex (CX) receives various forms of sensory signals and is associated with motor functions, but its precise role in behavior is controversial. The optomotor response is a highly conserved turning behavior directed by visual motion. In tethered cockroaches, 20% procaine injected into the CX reversibly blocked this behavior. We then used multichannel extracellular recording to sample unit activity in the CX in response to wide-field visual motion stimuli, moving either horizontally or vertically at various temporal frequencies. For the 401 units we sampled, we identified five stereotyped response patterns: tonically inhibited or excited responses during motion, phasically inhibited or excited responses at the initiation of motion, and phasically excited responses at the termination of motion. 67% of the units responded to horizontal motion, while only 19% responded to vertical motion. 38% of responding units were directionally selective to horizontal motion. Response type and directional selectivity were sometimes conditional with other stimulus parameters, such as temporal frequency. For instance, 16% of the units that responded tonically to low temporal frequencies responded phasically to high temporal frequencies. In addition, we found

26% of wide-field motion responding units showed a periodic response that was entrained to the temporal frequency of the stimulus. Our results show a diverse population of neurons within the CX that are variably tuned to wide-field motion parameters. Our behavioral data further suggest that such CX activity is required for effective optomotor responses.

Introduction

Motor control can be complex, requiring input from many sources within an animal’s central nervous system. Local reflexes and pattern generators in thoracic ganglia of arthropods or spinal cords of vertebrates provide local control to motor neurons.

Sensory receptors, including visual, tactile and chemosensors, monitor surroundings and lead to descending commands. Information from many of those sensors is processed through primary sensory systems and that processed information leads to descending commands that interact with the local control centers. As direct as this pathway seems, many behaviors also involve association areas of the brain. In insects, these include the mushroom bodies and central complex (Strausfeld 2012). The manner in which these regions are involved in motor control as well as the sensory information that is encoded in them is an open question in neuroethology.

Evidence from lesioning (Huber 1960) and electrical stimulation (Otto 1971) has long implicated the CX in regulating behavioral activities. More recently, neurogenetics have been used to disrupt the CC in Drosophila and show evidence for its role in many visually-directed behaviors, such as orientation, gap climbing, and visually-directed walking (Bausenwein et al. 1994; Strauss 2002; Poeck et al. 2008; Triphan et al. 2010;

Kahsai et al. 2010). Visual motion responses of CX neurons are modulated by behavior in flies. Feature detection of some EB neurons were diminished in flight, but not during walking (Seelig and Jayaraman 2013), while a group of FB neurons were shown to be unresponsive while quiescent but respond to translational optic flow while in flight (Weir

36 et al. 2014). Also, the CX has been shown to be necessary for visually mediated place learning in insects (Ofstad et al. 2011).

In the cockroach, further lesion studies have implicated the CX in forward walking, turning, climbing and tunneling (Ridgel et al. 2007; Harley and Ritzmann 2010).

Moreover, extracellular neural activity recorded in the CX has been correlated with walking speed, turning direction, and climbing (Bender et al. 2010; Guo and Ritzmann

2013; Guo et al. 2014). Stimulating the CX was also found to invoke behaviors, such as turning, where the direction of the turn was predicted by the stimulus location (Guo and

Ritzmann 2013).

The CX also plays a role in visual sensory processing. Extensive work has been done characterizing CX cells anatomically in relation to their role in polarized light vision in locusts, crickets, and monarch butterflies (Sakura et al. 2008; Heinze and

Reppert 2011; Pfeiffer and Homberg 2014). Intracellular recordings of CX neurons reveal representations of various types of visual motion in CX neurons of various insects, indicating directional preference (Phillips-Portillo 2012)(flesh fly), and looming sensitivity (Rosner and Homberg 2013)(locust). Two-photon calcium imaging in fruit flies has shown that ring neurons in the EB display directionally selective orientation tuning to narrow features that are arranged retinotopically with respect to their receptive fields (Seelig and Jayaraman 2013).

Questions remain about the range of the visual response properties for CX neurons, especially relative to particular behaviors such as optomotor response and spatial orientation. The optomotor response is an orientation behavior in response to rotating wide-field motion (Borst et al. 2010) and is found in the cockroach (Szczecinski

et al. 2014). Although the necessary directional motion is characterized in visual lobes

(Borst and Haag 2002; Borst et al. 2010), it is possible that the complete behavior could

involve associative regions such as the CX. In this study, we first established that a

functional CX is critical for the optomotor response by silencing CX activity with

procaine. We then used multi-channel recording to monitor groups of CX neurons while

presenting the cockroach with a wide range of repeated visual motion parameters, and

thereby established ranges of response properties that are relevant to optomotor and other

visually directed behaviors.

Methods

Animals

Adult male cockroaches, Blaberus discoidalis, were used in all experiments.

Animals were housed in 5 gallon buckets, given food and water ad libitum, and kept on a

12h light/dark cycle at 27°C. Recordings and behavioral testing typically began 2-3 h into the dark cycle.

Animal Preparation and Electrophysiology

Animals were anesthetized with ice and wings were removed. For extracellular recordings, a thread tourniquet was tied loosely around the neck, tight enough to reduce hemolymph flow but not so tight as to damage neck connectives. Reduction of hemolymph flow was necessary, because continued flow to the head results in a large clot forming over the brain, making it difficult to target specific regions. The insect was

placed into a vertical plastic tube, held in place with cork and wax to fill the spaces in the

tube. A plastic plate was fixed with wax on top of the tube and the head was fixed to the plate with wax across the mouthparts and behind the head, with negligible visual occlusion. A small portion of the cuticle between the ocelli was removed, with some

connective and fatty tissue, to expose the ventral surface of the brain. Care was taken to

avoid damage to the ocellar nerve or optic tracts. Saline (Tryba and Ritzmann, 2000)

added to the head cavity just covered the brain tissue. A copper reference electrode was inserted into the mesothoracic spiracle. After the experiments the animals typically walked normally.

Prior to insertion in the brain, the two shanks of the 16-channel silicon probes

(NeuroNexus® A-series 2x2 tetrodes, Ann Arbor, MI, USA) were dipped 5-10 times in

DiI paste (Invitrogen NeuroTrace® CM-DiI Paste, Eugene, OR, USA) for fluorescent labeling of the electrode tracks. Each shank is 15 µm thick and 150 µm apart from center to center and contains two diamond shaped iridium recording site tetrodes, also spaced

150µm apart vertically from center to center. The impedance of each channel was 2-3.5

MΩ. The probe was mounted on the headstage of the Neuralynx Cheetah32 digital data

acquisition system (Bozeman, MT, USA) and driven into the brain with a

micromanipulator. Unit activity was sampled at 30.3 KHz and band-pass filtered between

600 and 6,000 Hz. Only recording waveforms exceeding a predetermined voltage

threshold for any channel of the tetrode were saved.

Visual Stimuli

All visual stimuli were generated with a custom graphics program written in

Python utilizing the OpenGL graphics library (provided by John Bender). Visual stimuli

generated at 75 fps were displayed by a PLUS UA-1080 DLP projector, lit with a

mercury arc lamp with a UV filter, projecting onto a screen 9 cm from the animal that

subtended 135° horizontally and 100° vertically, centered in the frontal visual field,

resulting in a resolution of 820x645 pixels. The wide-field motion stimuli consisted of a shifting, sharp edged, black and white stripe field, covering the entire screen. Each trial consisted of the stripe field shifting for a duration of 4 s with some combination of parameters, varied randomly between trials. The varied parameters were temporal frequency, orientation, and direction. The temporal frequencies tested were 0.25, 0.5,

0.75, 1.0, 1.25, 1.5, 1.75, 2.0, 2.5, 3.0, 4.0, 5.0 Hz. Only two orientations were tested: vertically oriented stripes that shifted horizontally (horizontal motion) and horizontally oriented stripes that shifted vertically (vertical motion). Each orientation shifted in both directions, left/right and up/down, respectively. After a 6-10 s latent period, the stripe field would shift again for the next trial. Repeated parameter sets occurred randomly throughout the experiment so that, for example, all 20 trials of 0.25 Hz motion to the right direction did not repeat sequentially over a brief time period, but were presented randomly across the entire recording period. The stripes had a Michelson contrast of 0.96 calculated with luminances of 53.7 cd/m2 (white stripe) and 1.1 cd/m2 (dark stripe) measured with a Pentax Spotmeter V (TI Asahi, Saitama, Japan) and were 33° wide

(therefore the stripe field had a spatial wavelength of 66°). Contrast and width were held constant across all trials of all conditions.

To verify that the refresh rate of the stimuli was sufficient, four animals were

tested with an analog display system. Neural responses were compared for various

stimulus parameters and were found to have no significant differences from the digital

system (Appx. Fig. 5.1). Stimulus timing was synchronized to neural data with direct

TTL output from the stripe generator for digital trials and from a photodiode during

analog trials.

For behavior trials, the animal was facing an LCD monitor (NEC LCD17-BK,

Tokyo, Japan), driven at 60 Hz with a spatial resolution of 1280 x 1024 pixels. The

monitor was positioned 12 cm from the animals head and subtended 110° horizontally

and 110° vertically, centered in the frontal visual field. The stripes had a Michelson

contrast of 0.97, calculated with luminances of 71.6 cd/m2 (white stripe) and 1.1 cd/m2

(dark stripe). The stripe field parameters were held constant at 2 Hz temporal frequency and 66° spatial wavelength, moving horizontally for 10 s. Only direction was altered, which was randomly alternated between trials.

Spike Sorting and Analysis

Spike data from each tetrode were sorted off-line into unit clusters. Automated spike sorting was first performed using KlustaKwik (v1.5; K. Harris, Rutgers University, as part of the MClust toolbox, v3.5; A.D. Redish, University of Minnesota), which utilizes an expectation-maximization algorithm to separate spikes by the waveform parameters peak and energy (L2 norm). Cluster selection was very conservative, discarding units that had greater than 2% of its spikes with less than 100 ms ISI or units that did not maintain separation throughout the experiment. Great lengths were taken to

get accurate cluster separation over these long recording periods. All waveforms were

examined over time and manually corrected using Offline Sorter™ (Plexon Inc, Dallas,

TX, USA), which is excellent at visualizing waveform parameters over time. In addition, approximately five trials (1 Hz temporal frequency, both in the left and right direction), were compared from the beginning and end of each experiment. If any significant change in response was observed, in magnitude or timing, the unit was discarded.

After unit sorting, spike times were imported into Matlab (The Mathworks,

Natick, MA, USA), where all further data analysis was performed. Unit responses to

particular parameter sets of stimuli were shown with peristimulus time histograms

(PSTH), calculated by binning spikes into 50ms bins and convolving that histogram with a Gaussian kernel (σ = 150 ms) to estimate instantaneous firing rate.

For temporal response patterns and response curve classification, a fuzzy K-

means clustering algorithm was used, with a fuzzy factor of 1.5 (based on code from J.M.

Fellous, University of Arizona). Fuzzy K-means clustering is a weighted clustering algorithm derived from a standard K-means clustering, which identifies cluster centers and determines the Euclidean distance for each data point to each of the centers. Standard

K-means clustering assigns the data points to one and only one of K number of clusters.

The fuzzy K-means clustering algorithm, on the other hand, assigns a probability to each data point of belonging to each one of the clusters and therefore may belong to two or more clusters (details can be found in Fellous et al. 2004). This is an optimal clustering algorithm when few data points are available, as standard K-means clustering may lead to a large number of local minima. For temporal response patterns, the parameters used for clustering were the normalized peristimulus firing rates, at 50 ms intervals from 5 s

42 before the beginning of the stimulus through 5 s after the end of the stimulus period.

Mean firing rates were normalized between 0 and 1 by subtracting the minimum firing rate and dividing by the maximum rate.

Once response temporal patterns were established via clustering, individual response times were defined for each unit response. Responses for phasic units were determined by searching for peaks or nadirs within a time range defined for that response type. This search period was defined by the stimulus and the response envelope for that cluster (two standard deviations after the mean response time). Response period was determined by half of the peak/nadir height from mean baseline firing rate. For instance, for a unit response in the cluster of “phasic initiation - excitatory”, a peak was found between 0 s and 0.2 s (the response envelope for that cluster) from stimulus onset. A period of 0 s to 0.5 s was used to search for phasic responses to ambient light changes.

The response period was then determined by the width of the response peak at half its height from baseline. Tonic response periods were simply the stimulus period. This response period was used for significance testing against the baseline (paired two-tailed t- test, comparing the spike counts of the time periods divided by duration of the time periods of each trial) to verify a response of that type. In addition, a standard response period for a unit was found if the unit had only one response type across a given parameter, such as direction. This period was used for response comparisons (two-sample two-tailed t-test, also comparing the spike counts of the time periods divided by duration of the time periods of each trial) across different stimulus parameters, such as temporal frequency. The standard response period was determined by the minimum and maximum response periods of all responses of that given type for that unit.

An ANOVA was also used to identify units with variable responses across

different temporal frequencies to be included in tuning curve classification. For response

curves to temporal frequency, the parameters used for clustering were normalized

differences of mean firing rate for that unit’s standard response period from baseline.

These differences were normalized between -1 and 1 by subtracting the minimum

absolute value of the differences and dividing them by the maximum difference. For both

instances, K was selected based on pair-wise cluster separation statistics and visual assessment. Clusters with two or less units were merged with the closest cluster.

Behavioral Testing

Healthy male animals were tethered onto a 15.24 cm air-supported Styrofoam ball that was monitored by optical mouse sensors and recorded by a customized Matlab program (Guo and Ritzmann 2013). Two trials were performed before injection to assess the animal’s ability to perform an optomotor response. Animals that did not walk or noticeably display the response were discarded. After injection of either procaine or saline (n = 15, 15 animals, respectively), the animal was placed back on the tether.

Behavior was monitored every 15 min (recording ball movements for 90 s, with one visual motion trial 30 s into the trial) for 60 total min (5 total trials). After 60 min, the animal was removed from the ball and we then observed the animal during free walking and noted any abnormal behaviors.

The occurrence of an optomotor response was indicated when: (1) the mean angular turning velocity during the stimulus was 2*standard error ± mean of the turning velocity for 10 s before the stimulus and (2) the mean turning velocity was in the same

44 direction of the stimulus. Significant differences between treatment and control groups were found using a Pearson’s chi squared test.

Procaine Injection

Pretested animals were removed from the ball and injected using single barrel capillary tubes (World Precision Instruments, OD/ID, 1.0/.58 mm), pulled to a large tip size (700c, DKI Vertical Puller, Tujunga, CA, USA), which was then scored and broken so the new tip was between 20-40 μM OD. These pipettes were backfilled with either a procaine solution or cockroach saline. Procaine hydrochloride 99% (Acros Organics

AC20731) was dissolved to 20% and 10% concentrations, both in saline and 2% dextran fluorescein (Invitrogen D1822). Procaine, a sodium channel blocker, reversibly inhibits neural activity in insect brains (Muller et al. 2003; Devaud et al. 2007). Animal preparation was identical to restrained electrophysiology protocols above, except the brain was de-sheathed. The solution was injected into the central brain, frontally, using a

PM 2000 (B) 4-Channel pressure injection system (Microdata Instrument Inc- S.

Plainfield, NJ). After the experiment, injections volumes were estimated by injecting the solution into white petrolatum. Estimated injection volumes averaged 2.07 nL ± 1.8 nL.

Two controls were performed, a saline injection and a sham consisting of an identical surgery but with no injection. After the injection or sham, the neck tourniquet was removed and the head capsule was sealed.

For electrophysiological testing of procaine injected brains, the animal was prepared identically to the previous restrained recording preparations. The only exception was that the brain was de-sheathed before the recording electrodes were inserted. After

45 recording was initiated and some units were roughly identified, a baseline was recorded for 5-10 min. The injection pipette was then carefully lowered into the central brain at approximately 45° from vertical (to fit both injection and recording apparatus in the available space). If previously identified units still remained, the procaine or saline solution was then injected into the brain (at the same specifications as behavior trials).

Recordings then persisted for approximately an hour after injection. All further recording and spike sorting procedures were identical to previously discussed methods.

Histology and Imaging

At the end of both neural and behavioral experiments, brains were dissected from the animals, fixed in 4% paraformaldahyde, rinsed in 0.1M phosphate buffer solution, dehydrated in an increasing ethanol series, and cleared with methyl salicylate. The entire brain was then mounted in DPX mounting medium and optical sections were taken coronally using a confocal microscope (LSM700, Zeiss, Oberkochen, Germany) and 10x objective. No physical sectioning was needed, as the 400-500 µm thick tissue samples were imaged from both sides. Autofluorescence of the neural structures was used to identify the location of either the DiI lined electrode tracks or the injected procaine-dye mixture.

Results

Silencing Neurons of Central Complex Reduces Optomotor Response

In order to establish behavioral relevancy to the visual responses we recorded in the CX, we first observed the cockroach optomotor response. 75% of untreated cockroaches walking on an air suspended ball responded in a directional manner to right and left moving stripe patterns in their visual field (Fig. 2.1A). A 20% solution of the local anesthetic, procaine, injected into the CX reversibly blocked this response (Fig.

2.1B). The location of the injection site was determined by histological identification of dextran fluorescein that was mixed with the procaine. To control for general effects of fluid injection, we compared the procaine trials to insects that had been injected with saline solution. A significant reduction in the number of subjects that responded to optomotor stimuli was found in initial trials and trials taken 15 min after injection (chi squared test, p < 0.05). At 30 min and beyond, subjects performed normal optomotor responses. A 10% procaine solution only reduced the response at the initial testing time

(Fig. 2.1B).

To control for additional effects of the surgery, we performed sham trials, where the insect’s brain was exposed but with no injection. These trials showed no effect on optomotor response. Animals were also still able to walk after procaine injection. To verify that procaine injected animals were not simply paralyzed we measured general activity levels of each animal during drug trials and compared them with saline injected animals. We found no significant difference between the groups at any time after injection (two-sample t-test, p < 0.05) (Appx. Table 5.1).

To verify that the procaine effectively blocked neural activity near the injection

site, we performed 7 multi-channel recordings in and outside of the CX during a procaine

injection. Most, if not all, units recorded in or next to the CX were completely silenced

(Fig. 2.1C). As with the behavioral trials, all units silenced by the drug were affected almost immediately (mean = 0.2 min ± s. d. 0.4 min from injection) then recovered 10 min ±5 min after injection (mean of 4 animals, 42 units), with some silencing lasting over

20 min (Fig. 2.1C). Recordings taken from regions distinctly outside of the CX showed little to no effects from procaine injected within the CX (Fig. 2.1D). These patterns were consistent in all 7 recordings that were performed.

Wide-Field Visual Motion and the CX

Extracellular multichannel activity was recorded within or on the margins of the

central complex of 14 insects. These recordings yielded a total 401 units separated from

the multiunit activity. Units were sampled from both the FB and EB, sampling across

most of the dorsal-ventral plane of the CX, with a large proportion of the recording sites

located near the margins of the CX (Fig. 2.2D).

A series of 4 s visual stimulus trials (Fig. 2.2A and B) was presented to each

animal in random order. Unit activity was monitored while the stripe field shifted for that

trial. The stimulus for any given trial was either horizontal or vertical motion, shifting

among 10 temporal frequencies between 0.25 Hz and 5 Hz. Each parameter combination

was replicated randomly 15-20 times over a period of 2-3 h.

Response Types to Wide-Field Motion

Our analysis of the resulting data followed several stages that led to ever more precise indications of the temporal properties of each response. First, we identified general responses as mean changes in firing rate, over all 15-20 trials, that were at least two standard deviations from their baseline firing rates between the beginning of motion and 500 ms after motion cessation. Responses were found for all units to each parameter combination. This analysis identified 268 of 401 (67%) units that consistently responded in some way to visual motion.

To identify more representative temporal patterns occurring in these units, we used a fuzzy k-means algorithm that clustered the responses into common subgroups. We first normalized the firing rate of each response between 0 and 1. The clustering algorithm then separated the units based on temporal properties (Fig. 2.3). Responses in clusters with broad positive or negative changes that occurred throughout the stimulus were classified as tonic. Those with narrow peaks or nadirs were classified as phasic and grouped according to their relative location within the stimulus period. Three consistent phasic patterns were identified in this way: peaks or nadirs at the beginning or peaks at the end of the stimulus.

Once the unit responses were placed into consistent groupings, we could examine the actual peaks and nadirs of each response within a group to identify exactly where a significant response occurred relative to baseline (described in Methods). We then tested these more precise temporal response periods for significance from baseline (paired t-test

(p<0.05)). Responses in the tonic groups were tested over the entire stimulus period while

49 phasic groups were more restricted (see below). This analysis yielded 210 of 401 (52%) units with significant responses in a representative temporal pattern of the cluster they were assigned. While the remaining 58 motion sensitive units that did not meet these additional criteria may represent important responses patterns, they occurred rarely enough that we did not consider them further in our analysis.

With this temporal information, we could then describe the phasic response characteristics quantitatively. Units responding with phasic excitation and inhibition to the initiation of motion had median response durations of 0.45 s and 0.50 s, respectively

(Fig. 2.3F). But units with phasic excitatory responses to termination had a median response duration of 1.73 s. The first quartile for these responses was at 0.60 s, similar to the initiation phasic responses. This reflects the high variability in the distribution of these termination responses. The median response delay of phasic excitation to initiation, phasic inhibition to initiation, and phasic excitation to termination were 0.15 s, 0.20 s, and 0.48 s, respectively.

Unit Response Types and Directional Selectivity

The analysis described above allowed us to classify units according to common response types that varied across visual stimulus parameters. Individual units displayed consistent response types for a given direction. For each unit we determined a standard response time interval (described in Methods) then compared responses to left and right motion. 79 of 210 (38%) responding units were directionally selective to motion along the horizontal axis (two-sample t-test p < 0.05, for respective response period). These units consistently showed significantly higher mean firing rates in one direction. The

response in the non-preferred direction was reduced significantly (Fig. 2.4A), eliminated

(Fig. 2.4B) or depressed below baseline levels (Fig. 2.4C). 60% of directional units were biased to the left, and 40% to the right. Units were only classified as directional if they had similar response periods. That is, directionality was not established for units with tonic activity in one direction paired with a phasic response in the opposite direction (Fig.

2.4D).

Fig. 2.4E shows the distribution of units with regard to response types and direction. Most units (93%) maintained their tonic or phasic properties irrespective of direction. For instance, 62 units were tonic-excitatory in both directions. 16 of those units had a right preferred direction, 7 units had a left preferred direction, and 39 were not directionally selective (Fig. 2.4E, row 2, column 2). Only five units responded phasically in one direction and tonically in the other (Fig. 2.4D). Among those with phasic responses, no units responded to initiation in one direction and termination in the other, or vice versa. Baseline firing rates for each unit class were variable (Appx. Fig. 2.1).

Histological analysis revealed no patterns relative to directional selectivity. Indeed, various units from the same recording site would often select for different directions or different phasic and tonic response properties.

Sensitivity to Stimulus Temporal Frequency

We also examined the sensitivity of individual units to various other visual parameters. 168 units from 5 animals were tested with motion of at least 10 different temporal frequencies between 0.25 Hz and 5 Hz, in both the left and right directions.

There were several changes in response properties as temporal frequency varied. 26 units

changed from tonic responses at low frequencies to phasic responses at higher

frequencies, typically over a narrow range between 1.5-2.5 Hz (Fig. 2.5). The reverse

change never occurred. Interestingly, frequency effects often occurred (22 units) only in responses to one direction of movement (14 units had changes in responses to left movement and 8 for right movement). They included both excitatory and inhibitory responses.

Temporal frequency could also affect the directional bias of a unit. To examine

this effect, we selected 93 units with significant responses that retained consistent phasic

vs. tonic properties across all frequencies. We then identified units that showed some

significant differences in response properties as frequency varied (ANOVA, p <0.05) (50

units). These units were pooled and sorted by fuzzy k-means clustering and binned by mean response to each frequency (Fig. 2.6A and B). Of these units, 15 had responses that varied across frequencies only for motion in the left direction, 17 had responses that only varied for motion in the right direction, and 18 had responses that varied for both directions of motion. Our analysis revealed four clusters found from responses to left motion and five from right motion, representing variably tuned response curves to temporal frequency. Of the four average curves for responses to left motion, two (clusters

1 and 2) were of inhibitory responses and two (clusters 3 and 4) were of excitatory responses. These groupings were tuned to different ranges of temporal frequencies, distinguished by the range that evoked the strongest responses. For example, units in cluster 3 had the strongest responses at low frequencies, while the units in cluster 4 had stronger responses at higher frequencies. Similar relationships could be found in clusters

52 with inhibitory responses. Right motion data had three clusters that were similar to left motion (1, 4 and 5) and two (2 and 3) that were more restricted to lower frequencies.

An examination of individual unit response to various temporal frequencies shows that directional selectivity could vary with temporal frequency. Units with relatively broad tuning curves for each direction maintained their directional selectivity (two- sample t-test, †p<0.05, *p<0.005) at all frequencies (Fig. 2.7A and B), along with units with relatively narrow curves that were similar in each direction (Fig. 2.7C and D). But variation in frequency tuning for different directions of motion could lead to directional selectivity only over discrete frequency ranges (Fig. 2.7E and F). Narrow regions of directional bias across frequencies could also occur where units showed steadily increasing responses with frequency for one direction while maintaining consistently higher responses for the other direction (Fig. 2.7G). Some units had less stereotypical curves, sometimes with narrow bands of reduced or increased responses that led to narrow bands of directional selectivity across frequencies (Fig. 2.7H). The unit depicted in figure 2.7H failed to differentiate direction at a narrow frequency band around 1.5 Hz but consistently did so at frequencies above and below this band. Moreover, since the various stimuli were presented randomly 12-15 times at each parameter combination over the entire test procedure, the gap in directionality did not come about as a result of a transient change in responsiveness, but rather occurred whenever the ineffective frequencies were presented.

When taken in concert, a population of cells with this type of temporal frequency tuning can display different patterns of directionality depending on the temporal frequency used (Fig. 2.7I). Clearly, these properties emphasize the necessity to consider a

53 range of stimulus properties before labeling a neuron as either directional or non- directional. For example, in the data depicted in figure 2.7I (all units from one experiment), a test relying solely on 1 Hz stimulation (yellow box) would generate very different results than one taken with 1.75 Hz (red box) or 3 Hz stimulation (green box).

Responses to Wide-field Motion with Periodic Firing

In 92 of 401 units tested with wide-field motion (23%), we observed periodic firing rate changes that entrained to the temporal frequency of the stimulus (Fig. 2.8A).

Responses were classified as periodic if their spike phase distribution, relative to the temporal period of the stimulus, was non-uniform (Hodges-Ajne non-parametric test, p<0.01). These units often displayed distinct responses to changes in light intensity (Fig.

2.8B). 72% of these units had a phasic response to either ambient light turning on, off, or both often followed by a period of depressed activity. Bursts of spikes in these units were synchronized to a particular phase of the stripe cycle (determined by the angular mean of the spike phase distribution), yet this phase changed with frequency (Fig. 2.8C).

Individual units were often tuned to different phases at a given frequency (off-set) or with different phase-frequency relationships (slope). For one unit the phase tuning progressed linearly with frequency for stripes moving to the right (linear regression, slope = 79.84), but stayed entrained to the same phase, regardless of frequency, with left moving stimulation (slope = 4.85) (Fig. 2.8D).

No trends were seen between preferred phase or phase-frequency relationships and recording site location. These units were excluded from the cluster analysis and all analysis that relied on response type determined from the clustering. Moreover, entrained

responses were rarely directionally selective and no units were entrained in one direction

and not in the other.

Responses to Vertical Wide-field Motion

Responses to vertical motion, in either direction, were far less prevalent or

consistent. Only 14 of 74 (18%) units tested with vertical stimuli showed responses to

motion. Of those 14 units, 13 also responded to horizontal motion. Of the 13, 6 had

consistent tonic responses to horizontal motion, yet responses to vertical motion were

inconsistent (Fig. 2.9). Increases in firing rate occurred randomly during the stimulus

presentation, giving inconsistent response times, both across motion trials of the same

stimulus parameters as well as other motion parameters. Although some of these units

showed a statistical change from baseline, the response onset and duration changed for each stimulus condition and often was not in the defined response envelope from the cluster analysis. An additional 6 of the 13 units showed periodic activity to both horizontal and vertical motion (similar to Fig. 2.8). The remaining horizontal and vertical motion responding unit was phasically excited to right motion onset only. During vertical motion, it was phasically excited to with upward motion only. The unit that only responded to vertical motion also only responded phasically to the onset of upward motion.

Discussion

CX neurons convey diverse wide-field motion information

Diverse visual motion information is encoded by the neurons of the CX.

Responses to the stimuli provide timing information about the initiation, termination, and

duration of the motion, as well as direction and temporal frequency. These units can work in concert with one another as a population of numerous cells with numerous tuning properties. Information represented in these units, such as timing of onset and/or offset of motion, as well as direction and frequency, would be useful for evoking many visually

directed behaviors, such as the optomotor response. Given that these units are simply a

representative sample of a large population of neurons in the CX, variations in the

magnitude of these responses associated with different parameters of the stimulus may

provide subtle changes of input to downstream motor control circuits.

One of the striking features of our data is that individual units are tuned to a wide

range of stimulus parameters but that their response characteristics can change with certain other parameters. For example, responsiveness of the population can vary with the temporal frequency of the moving stripe field. Some broadly tuned units responded equally well over most of the frequency range that was tested (e.g. Fig. 2.6B cluster 4), while others showed significant responses to a narrower band of frequencies (e.g. Fig.

2.6B clusters 2, 3, and 5). Moreover the frequency band that generated significant responses varied among units. Finally, these curves could shift, become broader, or narrower with other parameters like direction. Therefore, response characteristics of

these units, such as directional selectivity and temporal frequency tuning, can be

dependent upon one another.

We also observed varying responses to different stripe orientations. Far fewer

units responded to the stripe field moving vertically, and those that did respond rarely

showed a consistent pattern. This could represent a behaviorally relevant bias to

horizontal motion. It is worth noting that previous work has shown a shift in response

properties of CX units when cockroaches were induced to climb objects, which is visually represented as vertical motion (Guo 2014). An alternate explanation is that the temporal frequency range for vertical motion responses is different than for horizontal motion responses, and our stimuli fell outside of the preferred range.

Periodic responses to wide-field motion

Many units displayed periodic firing entrained to the temporal frequency of the stimulus. It is likely that these units were responding to stripe features or light intensity changes rather than wide-field motion. This is supported by the strong phasic light intensity responses of many of these units. Retinotopically arranged feature detector neurons have been previously described in the fly CX (Seelig and Jayaraman 2013). Like the rest of the units described here, no spatial correlations of recording site were found with periodic response properties. Further testing would need to be done to verify if these units have small receptive fields, as hypothesized, and whether those units were topographically organized. In addition, the periodic units tested for vertical motion responded similarly to both orientations, unlike all other response types tested with

vertical motion. This also would suggest that at least some of these units are small field light intensity detectors.

The linear changes in mean response phase most likely represent a fixed latency.

Therefore, as the frequency of the stripe field progressed, the temporal entrainment would

progress as well. Differences in slope, between units, as well as within individual units

but to different directions, likely represent different latencies.

Response similarities with motion processing neurons upstream from the CX

The diversity of responses seen here and in other studies suggests a convergence

of sensory inputs, supported by the anatomy of the region (Strausfeld 1999). It is likely

that many of the motion responses we report are processed in the optic lobes and

integrated in the CX with other visual responses, as well as those of other sensory

modalities. Many units displayed tonic, directionally selective, and variable temporal

frequency responses to the stimuli. This is similar to the responses of motion sensitive

neurons in the periphery, such as the lobula plate tangential cells (LPTC) of the fly (Borst

and Haag 2002). LPTCs connect to thoracic motor centers, via descending neurons, as

well as other brain areas, and are involved in visual course control during flight

(Heisenberg et al. 1978; Geiger and Nässel 1981; Haikala et al. 2013). LPTCs can be

grouped by a preference to two orientations: horizontal (HS) and vertically (VS) sensitive

cells. A functional bias to HS cells could also relate to the orientation bias in our CX

cells.

The visual interneurons of the fly also have an initial transient peak of firing rate

before the activity settles to a steady state at high temporal frequencies (Egelhaaf and

Borst 1989; Reisenman et al. 2003). A similar velocity effect is seen in some CX units in this study, where units with tonic responses to slower stimuli changed to phasic responses for faster stimuli. In flies, this transient spiking is only seen in relatively high temporal frequencies and is always coupled with a velocity sensitive steady state response at lower frequencies, which is not the case for most phasic units in this study.

Additionally, transients occurring only at the cessation of motion (Fig. 2.1A), are not, to our knowledge, characteristic of LPTC neurons. This distinction could also help explain the difference in response duration of these responses (Fig. 2.3F). We suspect some of these responses may be after-excitation from a weak or obscured tonic inhibition.

Similar durations of increased firing rate were observed anecdotally in some tonic inhibitory units after the motion ended. Units with these rebounds were rare and had varying durations, and, therefore, were not separated via clustering.

The role of CX circuits in optomotor responses

To establish a link between the representations of wide-field motion found in the

CX and sensorimotor behaviors, we described the effects of disrupting the CX on a visually directed behavior. The reversible reduction of optomotor response with procaine injection implies some role for this structure in conveying this information to motor regions. Although other systems have been described as critical, and possibly sufficient, to regulate behaviors such as optomotor response (Haikala et al. 2013), in this study we found that the CX is necessary to this behavior. It is likely that the information from the

CX is not limited to involvement in optomotor behavior, yet it does play a role in this more basic, highly conserved response.

The implications of these findings do not exclude other pathways used in optomotor response, independent of the CX. At least two models for the role of the CX in sensorimotor behaviors exist. One consists of a serial pathway that includes the CX. In this model, peripheral sensory structures collect and process information which converges at the CX for further processing and integration, and is then relayed to motor regions in the thoracic ganglia. The serial nature of this model is consistent with the observation that silencing CX neurons reversibly blocks optomotor responses. Whether the CX circuits serve to create descending commands based upon visual information or modify them as they pass through, they represent a bottleneck and must be active to allow visual signals to descend from the brain. An alternative model involves parallel pathways that convey primary sensory information to both motor circuitry as well as associative regions, such as the CX. Under this model, the direct sensory-motor pathway yields fast commands based on specific sensory conditions, while the CX monitors conditions within and surrounding the animal, then modifies descending commands appropriately.

This parallel model is similar to circuits seen in the basal ganglia of mammals

(Alexander et al. 1990; Hikosaka et al. 2000; Kandel et al. 2013). In the mammalian motor system, direct pathways project from the motor cortex to spinal tracts or, in the case of the oculomotor system, the superior colliculus. But a parallel pathway through the basal ganglia influences the strength of those commands. Within the basal ganglia, direct and indirect pathways have opposite effects leading to the selection of appropriate behaviors at different times in the animal’s experience. Those pathways rely heavily upon inhibitory connections that are altered by neuromodulators such as dopamine.

Interestingly, a recent meta-analysis strongly suggests that the CX is deeply homologous

60 to the mammalian basal ganglia (Strausfeld and Hirth 2013). Moreover, the CX is heavily invested in GABAergic (inhibitory) receptors as well as receptors for numerous neuromodulators, including dopamine (Homberg et al. 1999; Kunst et al. 2011; Kahsai and Winther 2011; Kahsai et al. 2012).

Under the parallel model, it is not as clear why silencing the CX would block optomotor responses or other behaviors (Ridgel and Ritzmann 2005; Harley and

Ritzmann 2010). A possible explanation is that the CX’s role is so critical that some baseline activity is necessary for the direct pathways to express any behavior (Hikosaka et al. 1993). In either model the CX circuits can have profound effects on the optomotor response or any other behavior. Taking advantage of the large amounts of sensory information available to them as well as the rich supply of neuromodulators within the

CX, they could play an important role in adjusting the insect’s behaviors to match the internal and external context on a moment-to-moment basis. This ability to rapidly adapt to changing conditions is, indeed, a hallmark of animal behavior that easily distinguishes it from more static reflexive behavior.

Additional Contributions: Malavika Kesavan designed, performed and analyzed all data from the behavioral procaine experiments. All experiments involving electrophysiology, including those with procaine injections, were designed, performed, and analyzed by Nicholas Kathman.

Acknowledgements: I would like to thank Dr. John Bender for providing Python and MATLAB software along with advice in experimental design and analysis, Dr. Josh Martin for providing MATLAB software along with advice on analysis, Dr. Jean-Marc

Fellous for providing MATLAB software, and Alan Pollack for assistance with equipment, histology, and microscopy, and for maintaining the cockroach colonies.

Funding: This work was supported by the National Science Foundation [IOS-1120305 to R.E.R.].

Figures

Figure 2.1. Optomotor response is reduced after administration of local anesthetic in the CX. A. Animal’s turning response to shifting stripes was measured while tethered to an air-supported Styrofoam ball. B. Proportions of animals with a successful optomotor response at 15 min time intervals after the injection of 20% procaine (blue), 10% procaine (green), or saline (orange) into the CX (n = 15, 16, and 15 animals, respectively). Both treatments were significantly different (chi squared, p < 0.05) from saline controls at 0 min and only the 20% procaine was reduced at 15 min (asterisk). This effect was reversed from 30 min on for both treatments. C, D. Unit recordings during procaine injection verify that procaine did inactivate neural activity near the injection site, but not in regions of the brain outside the CX. Activity of units recorded from the margin (C) and outside (D) of the CX, before and after procaine injection (t = 0), is depicted by heatmaps of normalized firing rate of all units (rows) from the tetrode locations indicated in the image to the left (red arrow). Unit silence (when true instantaneous firing rate = 0) is indicated as white in the heatmaps. Green dye seen in the confocal images (yellow arrow) was injected along with the procaine. A clear gap exists between the procaine and the recording site in D, but not in C. Scale bar is 200 µm.

Figure 2.2. Visual stimulus description and recording site locations. A. Schematic of experimental setup for electrophysiological recordings of animals. B. Depiction of the stimulus at the two orientations used. Below the two stripe fields is an example of a series of trials with randomized temporal frequency, orientation, and direction. Deflections from baseline only indicate the times of initiation and termination of motion. C. Optical section of the CX with DiI indicating recording probe tracks (orange) in the left center and left margin of the fan shape body (FB). Scale bar is 200 µm. D. Positions of all recording sites for visual motion experiments, indicating position at the approximate depth of the recording site and the number of units recorded from that position. Many ellipses represent multiple recordings which sampled the same location.

Figure 2.3. Temporal properties of wide-field motion responses. Responses to wide- field visual motion had variable temporal properties that related to specific features of the stimulus. To determine what temporal response patterns were present, all unit responses (when firing rate was 2σ above or below baseline) were sorted based on their normalized mean firing rate over time (by fuzzy k-means clustering, binned at 50 ms) into 5 temporal response types. These types included responses lasting the entirety of motion (tonic) as well as brief responses (phasic) that corresponded to either the initiation or termination of motion. In all temporal classifications, with the exception of termination, responses were found as both increases and decreases of activity (excitatory and inhibitory, respectively). The heatmap in the center of the figure shows a response (normalized mean firing rate aligned to the initiation of motion, t=0) from each unit (rows) in the dataset, grouped by response type. A-E. Selected unit responses as examples for each response type (indicated by arrows to the appropriate row in the heatmap). These displays include a raster plot of neural activity from each trial as well as a peristimulus time histogram (PSTH, binned at 50 ms). In each PSTH the red line represents mean firing rate (smoothed with gaussian kernel, σ=150 ms) of all trials. The gray box on each PSTH indicates the time when stripes were moving. F. Box and whisker plots of the means of both the response delay from stimulus onset (left) and response duration (right) for responses for each unit with phasic responses (phasic excitation to initiation, phasic inhibition to initiation, and phasic excitation to termination; n = 58, 12, 5). Red lines represent medians, blue boxes represent first and third quartiles, and black whiskers are the extreme data points, excluding outliers. Outliers (red +) are points greater than 1.5 times interquartile range.

Figure 2.4. Distribution of temporal response types by horizontal direction and directional selectivity. Directional selectivity was seen in different forms, either: (A) reduced response in the null direction, (B) a loss of response in the null direction, or (C) as directional opponency (excitation in one direction and inhibition in the other). 5 units were phasic in one direction and tonic in the other (e.g. D). E. Table gives the number of units for each type of directional response. 60% of directional units were biased to the left (blue numbers), and 40% to the right (red numbers). Black numbers represent units that were not directional. Preferred direction was determined as that with the higher mean firing rate for respective response period. Units with different response periods were not classified as directional. Gray shaded boxes indicate pairs of response types with no representative units. Response type abbreviations: Phas - phasic, Excit - excitation, Inhib - inhibition, Init – to initiation of motion, Term – to termination of motion.

Figure 2.5. Changes in response duration with temporal frequency. An example of one unit with variable response duration with temporal frequency of stimulus for one direction only. For each histogram, the gray box represents the time of stimulus motion. Blue (left motion) and red (right motion) boxes represent the response duration, defined by when the smoothed firing rate (red line) exceeded 2σ of the baseline (green dashed lines). Plot on the right displays response width for each temporal frequency tested for both directions (blue for left motion, red for right motion). No data points are shown if no significant response was found (e.g. there was no significant response to left motion at 0.5, 0.75, 1.75, and 4 Hz). A short response duration (indicating phasic response) is seen in all conditions except right motion of frequencies below 1.75 Hz, where the responses are tonic.

Figure 2.6. Temporal frequency response curves. A,B. Groups of similar unit response curves to temporal frequency were identified for each direction of motion (A, left; B, right). To classify response curves, unit responses that vary with frequency (ANOVA, p<0.05) were normalized between 1 and -1 for each frequency, pooled and sorted by fuzzy k-means clustering for each direction of motion (heatmaps). Each cluster’s mean normalized response curve was plotted in the color of the cluster label (right of heatmaps).

Figure 2.7. Sensitivity to direction could vary with temporal frequency. A-H. Examples of response curves of individual units, for both left (blue) and right (red) motion. Each line represents mean difference in firing rate of response period from baseline for 12-15 trials of that frequency +/- standard deviation (shading). Responses to left and right motion were significantly different (two sample t-test, †p<0.05, *p<0.005) at various frequencies. I. Directional selectivity of all units (rows) over a range of temporal frequencies (columns) from one recording (2 tetrodes) with non-periodic motion responses. White grid cells indicate the unit is directionally selective at that frequency (p<0.05), black cells indicate it is not selective. Three highlighted frequencies show different distributions of selective units. The yellow box shows that five units were directionally selective at 1 Hz, the orange box shows that no units were selective at 1.75 Hz, and the green box shows that three units were selective at 3 Hz (two are different units than those selective at 1Hz).

Figure 2.8. Periodic responses to temporal frequency. A. One unit response to four different temporal frequencies of motion. The gray box in the PSTH represents the stimulus time interval. Circular histograms to the right depict the spike distribution (binned at 18° of the stimulus temporal period) of the same trials as their adjacent PSTH, with 0-360° representing the temporal period of the stimulus, starting from the onset of motion. All four frequency distributions are non-uniform (Hodges-Ajne test, p<0.01), but responding to different phases of the stimulus (red line is the angular mean of the spike phases). B. The unit had transient phasic excitation for ambient light turning on, followed by inhibition; and a strong but longer phasic excitation to ambient light turning off; also followed by inhibition. C. The relationship of the mean phase of the unit from A’s responses with temporal frequency of the stimulus in different directions (red = right, blue = left). The relationships were similar for each direction (linear regression, left motion: slope = 52.47; right motion: slope = 45.82). D. Phase-frequency relationship of another unit with different slopes for each direction (linear regression, left direction: slope = 79.84; right direction: slope = 4.85).

Figure 2.9. Responses to vertical motion. One unit’s response to three temporal frequencies moving vertically (left two columns) and horizontally (right two columns). Although three conditions (*) statistically showed a response to the stimulus (paired t- test, p<0.05), response periods greatly varied across conditions, unlike the responses to horizontal motion. This property was seen in all tonic responding units that also responded to vertical motion.

Chapter 3

Ocular dominance of Directional Responses to Wide-field Motion in Neurons of the Central Complex of the Cockroach, Blaberus discoidalis

Summary

Wide-field motion is an important source of information used in visual control of locomotion. This information may contain localized regions of directional variability that correlate with specific rotational and translational movements. Neurons in the central complex of the cockroach are sensitive to several characteristics of wide-field visual motion, including speed, orientation, and direction. Here, I demonstrate that certain neurons in the central complex show biases in response strength when only receiving input from one compound eye, and these biases are correlated to the direction of motion inducing the response. Therefore, these neurons receive directionally-distinct information from each compound eye, which suggests the ability to discriminate global optic flow with spatial differences in direction. This correlation with direction was only seen in some cell types, while other cells with periodic bursting entrained to the stimulus pattern showed the same bias to both directions of motion for one particular eye. The majority of responding cells (56 of 69) always showed responses, regardless of which eye was covered. Along with the previous characteristics, this suggests that, as a population, the cells of the central complex encode complex visual motion information that can be integrated with the various other sensory inputs and help direct locomotion.

Introduction

The central complex (CX) encodes various characteristics of wide-field visual

motion (Kathman et al. 2014). Wide-field visual motion is an important source of

sensory information to animals as they navigate through the world. As an animal moves,

visual motion is created as relative movements between the eyes of the observer and the

visual components of the environment. The resulting field of motion, often referred to as

the “optic flow” field (Gibson 1950), contains structures that indicate the animal’s direction of motion, including rotational and translational movements (Lappe et al. 1999).

Information from both visual hemispheres can be used to disambiguate the structure of

the flow field, as specific combinations of local visual motion characteristics can be

attributed to the animal’s self-motion.

Wide-field motion is initially processed in the nervous system’s peripheral

structures. In insects, each compound eye collects light that is processed by

retinotopically arranged direction selective elements found in the optic lobes (Borst

2014). Lobula plate tangential cells (LPTC) receive ipsilateral visual input from these

elementary movement detectors (EMD) (Egelhaaf and Borst 1993). They pool

information from many EMDs resulting in very large receptive fields of directional

selectivity.

Neurons that analyze complex optic flow fields must be able to respond to

different directions at different locations within a receptive field. Well studied examples

include MT and MST neurons in monkeys (Lagae et al. 1994) and looming sensitive

neurons of the locust (Gabbiani et al. 1999). Some individual LPTCs are able to select

74 for more complex fields. Integrating binocular information, at the level of these neurons, tunes the LPTC neurons to respond preferentially to specific flow fields (Krapp et al.

2001). Commissural pathways have been found that directly link these cells to the contralateral hemisphere (Farrow et al. 2006).

Certain behaviors, such as course control, are dependent upon this visual motion information (Lappe and Grigo 1999; Egelhaaf et al. 2002; Borst 2014). The CX plays a role in many of these same behaviors. Flies with structural defects in the CX show deficits in straight walking (Strauss and Heisenberg 1993; Strauss 2002) and visual flight control (Ilius et al. 1994). Gap crossing is another visually dependent behavior that the

CX plays a role in. In flies, gap crossing involved initiating a sequence of new behaviors while using a visual distance estimation to avoid insurmountable gaps (Pick and Strauss

2005). Mutant flies with defects in the CX often fail at gap crossing, where they are unable to properly orient their bodies in the appropriate direction, which requires coordinating spatial information (Triphan et al. 2010).

Goal-oriented behaviors are suggested to be a main function of the CX (Strauss and Berg 2010; Ritzmann et al. 2012; Strausfeld and Hirth 2013; Pfeiffer and Homberg

2014). Like gap crossing, many goal-oriented behaviors require distance assessments to be made, and having inputs from both eyes can be useful for making such measurements.

Praying mantises are excellent at assessing distance (Rossel 1980; Kral 2012). They are also quite closely related to the cockroach (Misof et al. 2014), the animal used in this study, and have a highly developed CX (Strausfeld 2012). They use this information for two important distance-dependent, goal-driven behaviors: prey strikes (Rossel 1980) and gap crossing (e.g. moving from perch to perch (Poteser and Kral 1995)). These behaviors

use two different distance calculations, binocular disparity for closer proximity prey

strikes and velocity parallax via a head shifting behavior called peering (Kral 2012).

Binocular disparity calculations require binocular inputs which compare spatial

information of static images from the perspective of each eye.

Visual motion information is also processed in the CX. Intracellular recordings of

CX neurons reveal representations of various types of visual motion in CX neurons of various insects, indicating directional preference (Phillips-Portillo 2012)(flesh fly), and

looming sensitivity (Rosner and Homberg 2013)(locust). Two-photon calcium imaging

in fruit flies has shown that ring neurons in the EB display directionally selective

orientation tuning to narrow features that are arranged retinotopically with respect to their

receptive fields (Seelig and Jayaraman 2013). Multi-channel recordings revealed detailed

patterns of motion sensitive responses throughout the cockroach CX that appeared to be

related to optomotor responses (Kathman et al. 2014).

Various pieces of sensory information must be integrated and transformed into

useful motor commands for such behaviors to occur. At the level of the CX, neurons

process information from different sensory modalities as well as activity associated with

locomotion behaviors (Guo and Ritzmann 2013; Pfeiffer and Homberg 2014). Visual

motion information in the CX is known to modulate behavior in insects. Feature

detection of some EB neurons were diminished in flight, but not during walking (Seelig

and Jayaraman 2013), while a group of FB neurons were shown to be unresponsive while

quiescent but respond to translational optic flow while in flight (Weir et al. 2014). Also,

in the cockroach, the optomotor response, a turning response to shifts in wide-field

motion, was transiently eliminated when the CX was reversibly lesioned (Kathman et al.

2014).

Neurons of the CX that are sensitive to the rotational angle of polarized light (E-

vectors) receive both bilateral and unilateral inputs in the locust. Anatomical and

physiological data shows that most cells were binocular, with the exception of some

tangential neurons of the ellipsoid body (EB) that received ipsilateral inputs only (Heinze

et al. 2009). Similar, neurons found in monarch butterfly have been associated with

migratory behaviors. These cells were also shown to respond strongly to non-polarized

light (Heinze and Reppert 2011). Some cells that were sensitive to looming and small- field motion also showed bilateral inputs in the locust, as well as excitation or inhibition depending on which eye was stimulated (Rosner and Homberg 2013).

Here, we report on cells recorded in the CX of the cockroach, Blaberus discoidalis that seem to also receive binocular inputs with functional distinctions between each eye. We altered visual inputs from binocular, to monocular (for each eye), and back to binocular conditions in order to determine which compound eyes provide their inputs and if there are biases between them (i.e. ocular dominance). A previous study identified several categories of wide-field motion sensitive neurons in the CX (Kathman et al.

2014), which are used as groupings in this study as well. After, classifying binocular response types, monocular effects were analyzed. Most responding units were found to receive inputs from both compound eyes, and some possibly with weaker inputs from the ocelli. We found that responses to left and right directional motion showed corresponding biases to the ipsilateral compound eye. That is, many neurons showed greater response amplitude to right motion when only receiving input from the right

compound eye. This pattern was seen more often in types of neurons that were

directionally sensitive. In contrast, units with entrained periodic bursting activity showed

monocular biases unrelated to the direction of motion or, often, no bias at all.

Methods

Animal preparation, recording, analysis, and histology procedures were consistent with

those described in Ch. 2 with the following additions and changes.

Animal Preparation

Following the binocular response testing (lasting 1-2 h), eye occlusion conditions were created by alternately painting over and then peeling off the coverings for each eye to generate the following sequence of conditions in all 6 insects used in this study: (1) binocular, (2) left compound eye monocular (right eye occluded), (3) both compound eyes occluded, (4) right compound eye monocular (left eye occluded). In four of these subjects (38 of the 69 responding units) this sequence was followed by a fifth set of observations in which both eyes were uncovered to return the insect to binocular vision.

Testing the units again under binocular conditions was a measure taken to control for dark-adaptations generated by eye covering and other changes in unit activity over time, such as state changes that were not light dependent, along with changes in recording strength or neuron health. The latter scenario was a concern due to the amount of handling of the animal when applying and removing paint to the eyes.

For compound eye occlusions, acrylic paint was applied to the appropriate

compound eye, taking care to have thorough coverage of the entire eye while not

reaching the ocellar lens or antennae. Red paint was first applied (PR108 opaque,

Utrecht Mfg. Corp., Cranbury, NJ, USA), followed by black paint to verify the opacity

(PBk 11 opaque, Utrecht Mfg. Corp., Cranbury, NJ, USA). After recordings were taken

for occlusion condition with both eyes covered, the dried paint was easily removed by

pealing it off with a forceps. The acrylic paint was often removed in one piece. The

contrast between the red paint and the dark eye surface made it possible to verify that no

occlusions remained.

Visual Stimuli

During the initial binocular test condition individual units were placed into

response classifications, following stimulus protocols in Ch. 2. During occlusion and

occlusion control trials, a series of 4 s visual stimulus trials (Fig. 3.1A) was presented to

each animal and unit activity was monitored while the stripe field shifted for that trial.

Stimulus trials were presented randomly, varying direction (left and right) and temporal

frequency (only 1 Hz and 2 Hz). Orientation (vertical bars), spatial wavelength (33°), and

contrast (0.96 Michelson contrast) were all held constant. Each stimulus parameter

combination was replicated randomly 15-20 times for each eye occlusion condition. The entire recording period lasted from 1-2 h.

Analysis

Response periods were determined for each unit to leftward and rightward motion

according to standards set in Ch. 2. Identified response periods were used for significance

testing against the baseline (paired two-tailed t-test, comparing the spike counts of the

time periods divided by duration of the time periods of each trial). Standard response

periods for each unit were used for response comparisons (two-sample two-tailed t-test, also comparing the spike counts of the time periods divided by duration of the time periods of each trial) across different response conditions, such as various eye occlusions.

All significance testing for tonic and phasic responses were performed on mean firing rates of these time periods, although some figure plots show mean differences from baseline to indicate excitation or inhibition, or monocular responses normalized to the binocular response for showing monocular bias.

An ANOVA was used to identify units with variable background activity levels across occlusion conditions and separated from analysis procedures involving baseline normalization.

Units with periodic activity during the stimulus period were identified as having significant non-uniformity (Hodges-Ajne nonparametric test for circular uniformity, p <

0.01) across temporal periods of the stimulus, starting at time 0. To measure the response amplitude of these units, various measures of stimulus modulation were tested. The best indicator of activity levels during the preferred phase of the response as well as conformity to the stimulus was the spectral density at the temporal frequency of the stimulus.

To measure how closely the periodic responses followed the temporal pattern of the stimulus, an estimate of the power spectral density (PSD) was found for each response and the power at the temporal frequency at the stimulus was measured. To estimate the PSD for the response, rate histograms for each trial were found for each stimulus period (0.01 s bin width). The data was multiplied by a Hann window

2 1 cos ( + 1) ( ) = 휋2푖 − � � 푛 � 푤 푖

where n = the number of bins. A periodogram was used to then estimate the PSD, which is found by determining the discrete-time Fourier transforms for the samples and scaling the magnitude squared of the result. The periodogram estimate of the PSD of length-L signal xL(n) is

1 ( ) = 퐿−1 ( ) 2 −푗2휋푓푛 �퐹푠 푥푥 퐿 푃 푓 푠 �� 푥 푛 푒 � 퐿퐹 푛=0

where Fs is the sampling frequency. A fast Fourier transform (FFT) was used with frequencies 0-50 Hz at 0.25 Hz intervals. The mean log power (log10 of the raw power multiplied by 10) at the stimulus temporal frequency was used for all periodic response comparisons. Spectragram heat maps were found by normalizing the mean PSD (power spectrum divided the largest value) for response with the lowest p-value found by the

Hodges-Ajne non-uniformity test so the sum of all the spectrum values equals one.

Results

General response classifications

In this study, we analyzed the ocular dominance of several groups of wide-field motion responding units to elucidate how much information each neuron receives from either eye. Extracellular multichannel activity was recorded within or on the margins of the CX of six cockroaches. These recordings yielded a total 117 units separated from the multiunit activity. Units were sampled from both the FB and EB, sampling across most of the dorsal-ventral plane of the CX, with a large proportion of the recording sites located near the margins of the EB (Fig. 3.1B).

Temporal response type classifications in binocular conditions

Before examining the monocular effects of these units, we first classified them by their response to wide-field motion under binocular conditions. This provided a classification system for sorting monocular effects. Of 117 units recorded, 69 of those units (59%) responded to some set of these stimuli. Unit responses to directional motion

(Fig. 3.2) were generally consistent with those reported in the previous chapter.

Some notable unit types were found. For example, there were 12 tonic opponency

units, which were tonic excitatory to a preferred direction and tonic inhibitory to the

opposing direction and 12 units that were phasic excitatory to the initiation of motion in

both directions. The tonic opponency units were found in three of the six recordings

performed. Twenty-three units showed periodic activity that was entrained to the temporal frequency of the stimulus. In some instances, units could have periodic

82 responses to one direction of motion and a tonic or phasic response in the other direction, but these units were still classified as periodic. Although entrained units were found in all six experiments, they were somewhat more concentrated to two subjects. 14 of the 23 periodic units came from three recording positions of these experiments, all located in the ellipsoid body, at the lower and upper margins (Fig 2F). Both tonic and phasic units were also found at these sites, but no tonic opponency units were found in them.

There were a few differences from the larger data set described in the previous chapter. These consisted mostly of the absence of units in categories that were found rarely in the larger study. Far fewer units with tonic excitatory responses in both directions were found and no units were found that responded to the termination of motion. Also, only two units were found that changed response change from tonic to phasic responses with increased stimulus temporal frequency.

Ocular dominance testing

Ocular dominance of CX units was tested by comparing unit activity while occluding either compound eye to unit activity with neither or both compound eyes occluded. Few differences were seen among monocular effects of any units between responses to the slower or faster stimulus speeds, therefore, responses to 2 Hz stimulus was used for all units, except one that only responded to 1 Hz, which was the testing frequency for that unit.

We also verified that background activity under these conditions was consistent, as some units showed changes that possibly correlated with response amplitude. Units with significantly variable background firing rates (9 tonic/phasic units and 8 periodic

units, 17 units total, ANOVA p < 0.05) between monocular and binocular conditions

were omitted from the initial ocular bias classification and are discussed later.

Ocular dominance correlates with directional response for tonic and phasic responses

Responses recorded when one eye was occluded revealed a relationship between

ocular dominance and the direction of stimulus movement. Of the 44 units with tonic or

phasic responses to motion (i.e. not showing any periodic responses), all but five showed

some change in response amplitude in monocular versus binocular conditions. The

remaining five showed no change in monocular responses from binocular ones. Of the

39 units with monocular effects, a strong correlation was seen between monocular bias and the direction of stimulus motion. This is best exemplified in the units with tonic opponency responses (i.e. units with tonic excitation for one direction of motion and tonic inhibition for the opposite direction). For all 12 units, the excitatory response to the preferred direction of motion for that unit was significantly larger (two-sample t test,

p<0.05) when only receiving input from the compound eye ipsilateral to that direction

when compared to the other monocular condition. That is, MonoL (right eye covered)

showed larger responses to leftward movement than MonoR for units preferring leftward

movement and MonoR had larger responses to rightward movement than MonoL for

units preferring rightward movement (Fig. 3.3A and B).

This rule often held true for the responses to the non-preferred direction as well.

Of the seven units with left preferred direction, three of them also showed a larger

response in the ipsilateral monocular condition to the non-preferred direction (Fig. 3.3B,

84 left). These results require some explanation. Since the response to the non-preferred direction is inhibitory in tonic opponency units, a larger response means the activity was more inhibited, resulting in lower activity when only the ipsilateral eye was uncovered than when only the contralateral eye was uncovered. It is also important to note that in figure 3.3B, the absolute value of response amplitude is plotted, therefore larger values in the y-axis represent greater inhibition (i.e. lower activity) for the non-preferred direction of each unit. The other four left-preferred units had no significant difference between the responses taken under the two monocular conditions. Of the five units with a right preferred direction, all of them also showed a larger inhibitory response (i.e. less activity) in the ipsilateral (MonoR) monocular condition for that direction (Fig. 3.3B, right). Thus

8 of 12 opponency units showed an ocular bias to movement in either direction.

This trend of monocular biases correlating with directional responses also holds true for many phasic units. Of the 12 units with phasic responses for both directions of motion, five had significant differences between responses during the two monocular conditions to at least one direction of motion. All of them showed a larger response in the monocular conditions ipsilateral to the direction of motion, but only one showed this bias for both directions of motion (Fig. 3.3C). Unlike the tonic opponency units, only two of these phasic units were directionally selective, both to right motion. The seven remaining units had no bias.

Units with tonic inhibitory responses in both directions were affected by eye occlusion in a manner similar to the inhibitory responses found in opponency units. Of six units, three showed a larger response to left motion in the ipsilateral monocular condition and three showed larger responses to right motion in the ipsilateral condition,

but only one unit showed both of these trends. Unlike tonic opponency and phasic units,

two units did actually show the reverse trend where responses to right motion were

actually larger during the contralateral monocular condition.

Remarkably, very few units failed to show stronger responses to a given direction

of motion during the ipsilateral monocular condition. Of the 33 units with tonic or phasic

responses with some response to leftward motion, 16 units showed significantly larger

responses in the left monocular condition (ipsilateral) and only one unit was larger in the

right monocular condition (contralateral). The remaining 15 units showed no monocular

bias for this direction. Of the 34 units with some response to rightward motion, 15 were

larger in the right monocular condition (ipsilateral) in this direction, and only three showed the opposite trend. The remaining 16 units had no monocular bias for this direction. Beyond the tonic opponency units, no correlation in preferred direction was seen with monocular bias.

Additionally, two units changed temporal response type during a monocular condition. One changed from tonic to phasic responses and the other from phasic to tonic, both for responses to right motion and both in the monocular right condition.

Units with increased response amplitudes in monocular conditions

Some tonic and phasic units showed increases in monocular conditions when compared to the binocular condition taken before the painting treatment. Interestingly, all units with a significant increase in response during a monocular condition that was also tested with a post-paint treatment binocular condition also showed a significant

increase in response in that condition (Fig. 3.4A). None of these units showed increased monocular responses compared to this post-binocular response.

Although few units showed this response amplitude increase, a disproportionate amount of the ones that did were tonic opponency units. Only ten units out of the 69 total showed greater monocular responses compared to the pre-binocular test, but seven of them were tonic opponency units. As seen in Fig. 3.3B and C, which were normalized to the pre-treatment binocular response for each unit, approximately half (7 of 12) of the tonic opponency units had increased monocular responses over the initial binocular recordings (Fig. 3.3B, Left PD units - all responses marked with circles and squares,

Right PD units – only red circles and squares). Four of seven units with left preferred direction showed monocular increases (two to both direction of motion) and three of five units with right preferred direction show the increase (one to both directions). One unit with phasic responses to both directions of motion had a monocular increase for right motion only (Fig. 3.3C, only the open square), and two periodic units had monocular increases for both directions (discussed later, Fig. 3.5D, open and closed squares).

When we re-examined the binocular condition after the eye covering trials were completed, we found a correlation between increased monocular responses and increased post-binocular responses. Of the 7 tonic opponency units with a monocular increase, only two were tested with a post binocular condition, but both showed significantly increased responses when compared to the pre-treatment binocular condition for the relevant direction (Fig. 3.4B). The monocular response of both of these units had responses contralateral to the preferred direction that were significantly smaller than that found in the post-binocular condition. The ipsilateral condition to the preferred direction

was also significantly smaller in one of the units (Fig. 3.4B, right) but not significantly

different in the other (Fig. 3.4B, left) for the preferred direction. In the non-preferred

direction, only one of these units showed greater inhibition in a monocular condition

compared to the pre-treatment binocular condition (Fig. 3.4B, right), which was not significantly different from the post-treatment binocular condition, while the contralateral condition was significantly smaller.

Therefore, the directional response correlation between monocular conditions discussed above still holds for these units. Also, when comparing these monocular conditions to the post-binocular, rather than the pre-binocular condition, the relationships are the same with binocular conditions. In this case, the response to a particular direction is reduced from the binocular condition when the ipsilateral eye is covered.

Ocular dominance does not correlate with direction for entrained periodic responses

The relationship between ocular dominance and stimulus direction in CX neurons with entrained periodic responses was very different from that found for the tonic opponency neurons. Relatively few periodic units showed any response bias between monocular conditions, and those that did, showed the same bias for responses to stripes moving in either direction. To measure response amplitude, we calculated the spectral density at the temporal frequency of the stimulus to measure the level of entrainment to the stimulus. The spectral density at this frequency was a good indicator of how entrained the activity was to the stimulus (Fig. 3.5A). Many units had varying degrees of inhibition and excitation associated with each temporal period so that the mean firing rate across the

88 entire stimulus period was not necessarily changed from baseline, but spectral density captured these changes well (Fig. 3.5B, below).

Three general groupings emerged from periodic units. The majority (14 of 23) were not biased to either compound eye, nine of which showed no change from the binocular condition at all. But four units were biased to the left compound eye (larger in this condition) for motion in both directions (Fig. 3.5C) and five units were biased to the right compound eye, but only two that also were biased for motion in the right direction

(Fig. 3.5D). This is a different effect from the pattern seen in tonic and phasic units.

These units show greater response amplitudes for the same monocular condition, regardless of the direction of motion inducing the response.

Increases in monocular responses, when compared to the pre-binocular condition, were rare, but again, where they occurred, this change correlated with post-binocular increases. Of all 23 units with periodic responses, only two had a significantly larger response for any monocular conditions compared to the binocular condition (Fig. 3.5D, open and closed squares). These units showed the increased monocular response to both directions of motion and also showed an increased response in the post-binocular control, but less so than the monocular condition. This was true for the spectral power measurement of these responses, as well as overall firing rate during the stimulus period.

Few units completely lose response in monocular conditions

Although many of the units showed reduced response amplitude under monocular conditions, most retained some response, suggesting that few, if any, units were exclusively monocular. Only 13 out of 69 units failed to respond during at least one

89 monocular condition. Even these could have responded below threshold for detection in our extracellular recordings. This group was made up of 10 phasic excitatory units and three periodic units. Four units (all phasic) lost their responses in the monocular left condition and seven units (four phasic and three periodic) lost their response in the monocular right condition. Four additional units (all phasic) lost their responses during both the left and right monocular conditions suggesting that these units required input from both eyes to reach threshold for a response (Fig. 3.6). No units with only tonic excitatory, tonic inhibitory, or tonic opponency responses ever completely lost their response for both monocular conditions.

Responses during control and ocellar inputs

A small proportion of units retained responses when both compound eyes were occluded. In each case, these reduced responses were eliminated by further covering the ocelli. Of the units tested, 14 did not completely lose their response when both compound eyes were painted, although they were greatly diminished. Of these units, 11 showed periodic responses to motion (Fig. 3.7A) and three had tonic responses (all showing tonic opponency) (Fig. 3.7B). Periodic responses that persisted during the compound eye occlusion control were always greatly reduced from the binocular and monocular responses. These units were found in every animal tested, with no notable distribution biases. Not all periodic units with persistent responses during control conditions remained periodic in this condition. Some units only showed a brief transient at the beginning of the stimulus, similar to phasic responses found in other units. Of these

11 units, four were also tested with an additional condition where the ocelli were also

occluded (Fig 7C). The periodic responses of all four of these units were silenced in this

condition.

Three units that retained weak responses with both eyes occluded reversed their preferred direction under that condition (Fig. 3.7B). All three of these were tonic opponency units. For example, the unit shown in figure 3.7B preferred leftward motion during binocular and monocular conditions, with tonic excitatory responses to left motion and tonic inhibitory responses to rightward motion. When both compound eyes were occluded, this was reversed, so that rightward motion became slightly preferred with small tonic excitatory responses. Meanwhile a small inhibitory effect was now visible to leftward motion.

Units with changes in background activity

Since eye occlusion can clearly affect the impact of ambient light on the system along with the response to discrete stimuli, we also examined background activity under all compound eye occlusion conditions. Of the 69 units that responded to the visual stimulus, 17 units had significantly different background firing rates (ANOVA, p < 0.05), measured as the mean firing rate for 4 s before the stimulus for all trials of that condition.

All 17 units came from two experiments, the same two experiments that had a disproportionate amount of periodic units and no tonic opponency or inhibitory units.

There were four types of occlusion effects seen in the background of these units.

Many of these units (nine) showed reduced background activity in all conditions with at least one eye covered (Fig. 3.8A). All of these units were found in the same recording, at the lower margin EB sites and were of different types (tonic, phasic, or periodic). The

majority of these units also showed reduced background activity during the post-

treatment binocular control, where all occlusions were removed (Fig. 3.8A, left). Two

groups of units, all periodic, showed background changes only for the control condition

where both eyes were occluded (Fig. 3.8B). One additional unit showed increased

activity for all occlusion conditions.

All units where background activity was reduced for all occlusion conditions also

had a reduced or eliminated response in monocular conditions. Also, occlusion effect on

background activity did not always recover in the post-binocular condition, but some responses did. Seven of 11 units showed reduced background activity in the post- binocular condition (Fig. 3.8A and C), but in five units, response activity recovered during the post-binocular control, independently from the baseline activity (Fig. 3.8C,

left). Additionally, four of these units showed monocular bias in both response activity

and background activity (Fig. 3.8C, right).

Discussion

Ocular dominance and directional motion in the CX

Ocular dominance was analyzed to determine how much information sample

neurons recorded in the CX received from either compound eye. Although few absolute

biases were seen with any given group of units, several strong trends were seen. Few

cells completely lost their visual response when either eye was covered. Those that did

have their responses eliminated when one eye was covered consisted primarily of phasic

excitatory units and a few periodic units. Even these neurons could have experienced

subthreshold activation that went undetected with our extracellular recordings. This

suggests that most, if not all, neurons in the CX, especially those that respond tonically to

moving stripe stimulation, receive inputs from both compound eyes.

Many neurons, however, did show a preference for one compound eye over the

other and these biases differed with response type. For tonic and phasic units, a strong

correlation was seen between the direction of motion a cell was responding to and which

compound eye provided the dominant input. The response was always greater during the

ipsilateral monocular condition. This suggests that the directional responses to these units receive more input from the eye ipsilateral to that direction (e.g. right eye for rightward movement). This pattern was best distinguished when looking at the tonic opponency neurons. These units showed a very clear trend where all 12 were dominant to the eye ipsilateral to their preferred direction of motion. This trend suggests these neurons may provide a localized directional code that could potentially discriminate various types of egocentric-motion.

Although the preferred direction response (i.e. the excitatory response) always showed this relationship, the responses to the non-preferred direction (i.e. the inhibitory responses) were less consistent, as several showed no bias for either eye. This was also true for neurons with only tonic inhibitory responses to motion. The activity of these neurons was often completely silenced during the response period, so this may simply reflect the notion that inhibition from either eye was sufficient to significantly decrease firing. There also were not as many neurons with phasic responses in both directions following this trend, yet only one showed the opposite trend.

Some phasic neurons completely lost their response in one monocular condition,

suggesting they received input exclusively from one eye for some directional responses,

although it is likely that these cells all receive binocular inputs lateral inhibition could

account for a complete loss of response. This is further suggested because the monocular

responses rarely seem to add linearly to equal the binocular response. This was not

thoroughly examined, as several units also showed increases in monocular response

compared to the binocular condition (discussed later), so these linear summations were

unlikely.

The two tonic opponency units that changed response type possibly suggest that

these tonic and phasic responses exist together in some of these neurons. Yet both were

changes in right directional responses in the right monocular condition, therefore these responses are not dropping out to reveal another type. Also, no neurons were found that had inhibitory responses in one monocular condition and excitatory effects in the other, as seen in polarization sensitive neurons in the locust (Rosner and Homberg 2013).

Response similarities with motion processing neurons upstream from the CX

It is possible that many of the motion responses examined here are processed in

the optic lobes and integrated in the CX with other visual responses, as well as those of

other sensory modalities. Many neurons displayed tonic, directionally selective responses

to wide-field motion. This is similar to the responses of motion sensitive neurons in the

periphery, such as the lobula plate tangential cells (LPTC) of the fly (Borst and Haag,

2002). LPTCs connect to thoracic motor centers, via descending neurons, as well as other

brain areas, and are involved in visual course control during flight (Geiger and Nässel,

1981; Haikala et al., 2013; Heisenberg et al., 1978). Some also receive binocular inputs

(Farrow et al. 2006), therefore the binocular inputs contributing to responses in these cells may be added upstream. The similarity with these optic lobe cells may suggest that the CX serves as a center for efference copy receiving information similar to descending commands that can be used to monitor the behavioral state of the animal.

The visual interneurons of several species of flies also have an initial transient

peak of firing rate before the activity settles to a steady state at high temporal frequencies

(Egelhaaf and Borst, 1989; Reisenman et al., 2003). A similar velocity effect is seen in

some of these neurons, where neurons with tonic responses to slower stimuli changed to

phasic responses for faster stimuli. It is also possible that some phasic neurons described

here are, in fact, these transients. This may explain why tonic and phasic neurons show

similar trends in ocular dominance with directional responses.

Periodic neurons and neurons with background changes

Periodic neurons seem to have distinct inputs from those with exclusively tonic or

phasic responses, and likely receive signals for both directions of motion from the same

eye or equally from both eyes. Almost half of the periodic neurons showed a monocular

bias, which was the same bias for both directional responses. In fact, these units rarely

show any differences between directional responses at all (Kathman et al. 2014). It is

possible that these neurons are not encoding wide-field directional motion, but rather small-field motion or simply light intensity in discrete regions of the visual field. This would explain the periodic pattern, if the neurons was being excited or inhibited as individual stripes pass over the neurons receptive field. Of the remaining neurons, nine

of 23 showed no change at all in monocular conditions when compared to the binocular

conditions, and three showed no monocular bias. These cells are likely receiving inputs

from both compound eyes but showed no bias.

Neurons with periodic responses also often showed changes in their background

firing rate when eyes were occluded. Specifically, these were neurons with background

activity that was only reduced in the control condition. This also indicates that these

neurons may be more sensitive to light intensity than motion. Many neurons showed

responses that correlated with the background changes, further suggesting that the light

levels may be influencing response amplitude more than motion properties.

All of the neurons with background changes addressed above were found in the

same three recording sites. These were the same three recording sites where the majority of periodic neurons were found and no tonic opponency neurons were found. In our previous study (Kathman et al. 2014), no correlations with recording location were found for any response characteristics. It is possible that the relative locations, in relation to the structures of the CX, are not important, but certain functional cell types may be grouped together and dispersed throughout the structures.

Increases in response amplitude compared to pre-binocular responses

Few neurons showed an increased response to any monocular conditions, and those that did so also had increased response amplitudes during the post-binocular control condition. For these cells, after one eye was occluded, the same ocular dominance patterns seen in other cells for that type seem to hold, but only when compared to the increased post-binocular condition. These effects seem to reflect a gain increase in

response amplitude, often for both directional motion responses, after either eye was

occluded. This effect is possibly a dark-adaptation from one or both of the occlusions

that could involve circuit modifications, at the level of the CX or upstream from it,

relating to the reduced light input. Although light- and dark- adaptive effects on motion vision are thought to be related to the nature and extent of spatio-temporal filtering in the retina and lobula (Pick and Buchner 1979), it is not surprising to see other effects, such as changes in response gain in an associative region like the CX.

This effect was seen in tonic opponency, phasic, and periodic neurons. This suggests either a state change that affects all visual streams (motion, intensity, etc) in the

CX or an effect taking place in the periphery, possibly at the photoreceptor level, before visual streams are separated. Yet the effect was seen disproportionately in tonic opponency cells, so upstream wide-field motion processing systems could be more affected. Additionally, if the cells showing this gain with periodic activity are not truly responding to motion but intensity, as suggested previously, these cells may be reflecting more primary dark-adaptations affecting sensitivity (Butler 1973), as the increase only

occurred in the conditions following the control where both eyes were occluded.

These potential dark-adaptations are not unexpected and may be present in other

neurons that did not show increases from binocular response. Recent studies in another

cockroach, Periplaneta americana, have described a unique variability in structure and

function of their photoreceptors, where population coding has been show as a means for

pooling these signals to obtain reliable intensity information (Heimonen et al. 2006).

This is thought to be an adaptation to better make sense of the inherently unreliable light

signals present in dim environments when using apposition eyes. The variability in these

CX cells may also reflect this retina variability, and may use a similar coding method for its visual information.

Responses during binocular occlusions and ocellar inputs

The majority of neurons lost all forms of a response in the control condition where both compound eyes were occluded and all neurons had at least a reduction of response amplitude. This suggests that all responding neurons are dominated by inputs from the compound eyes, at least with regards to wide-field motion responses. Yet some neurons appear to have weak ocellar inputs as well. A notable correlation we found was between neurons that had periodic responses and those that had a response when both compound eyes were covered. Although this could be a result of incomplete eye occlusion, the specificity to unit type suggests that these neurons may have alternate inputs. This if further supported by the fact that many other unit responses recorded in the same animals under the same level of occlusion were silenced during controls.

Ocellar light information is used for various behavioral tasks in insects, including locomotion and flight control (Taylor and Krapp 2007). In the cockroach, the ocelli inputs are known to affect obstacle negotiation (Harley et al. 2009), which is a behavior modulated by the CX (Harley and Ritzmann 2010; Guo et al. 2014). It is also known that neurons from the ocelli plexus project to various regions of the brain, including primary sensory structures, such as the optic lobes, as well as associative structures, such as the mushroom body, but not the CX (Mizunami 1995). Therefore, ocellar inputs are likely upstream of the CX, possibly directly modulating motion sensitive neurons in the optic lobe that then project to the CX. Visual systems using the ocelli are often very fast

(Taylor and Krapp 2007), and likely are not well suited for direct inputs to an associative structure like the CX.

Binocular vision, behavior, and the CX

Many behaviors involve assessing visual objects in an insect’s path and directing some movement based on that assessment. These behaviors involve collecting information from various sensory structures and integrating them in the brain. The central complex has long been thought of as a candidate for this task.

Such behaviors, like navigation and course control, are dependent upon determining direction and speed of ego-motion, often using vision (Lappe et al. 1999;

Egelhaaf and Kern 2002). By having binocular inputs that are directionally distinct from each eye, these neurons may be integrating spatially distinct directional information.

Therefore, the CX could be capable of differentiating between optic flow structures associated with various movements, such as translation or rotation (Lappe and Grigo

1999; Krapp et al. 2001). Binocular dominance has previously been shown in the CX for polarized light sensitive cells (Heinze et al. 2009) and looming sensitive cells (Rosner and Homberg 2013), which further supports this idea.

Since the CX plays a role in walking control (Strauss and Heisenberg 1993;

Strauss 2002; Bender et al. 2010), turn initiation (Ritzmann et al. 2012; Guo and

Ritzmann 2013) and visual flight control (Ilius et al. 1994), it is likely this information is encoded in the CX, and possibly the neurons recorded from in this study. Considering the relationship with directional responses and ocular biases of the tonic opponency neurons, along with the similarity of these responses to LPTCs, this function seems more

likely than for distance measurements which rely on small-field motion information.

More characterizations of the neurons with periodic responses need to be performed to assess if they do, in fact, display small-field motion characteristics or encode other static light characteristics. In the future, more complexly structured flow fields should be tested on tonic neurons to determine if they can discern differences, such those associated with rotational and translational motion.

Acknowledgements: The authors thank Dr. John Bender, Dr. Josh Martin, and Mr. Alan

Pollack for help throughout this project.

Funding: This material is based upon work supported by the National Science

Foundation (NSF) under Grant No. IOS-1120305 to RER.

100

Figures

Figure 3.1. Visual stimulus description and recording site locations. A. Schematic of experimental setup for electrophysiological recordings of animals. B. Depiction of the stimulus used. Below the stripe field is an example of a series of trials with randomized temporal frequency and direction. Deflections from baseline only indicate the times of initiation and termination of motion. C. Optical section of the CX with DiI indicating recording probe tracks (orange) in the left center and left margin of the fan shape body (FB). Scale bar is 200 µm. D. Positions of all recording sites for visual motion experiments, indicating position at the approximate depth of the recording site and the number of units recorded from that position. Many ellipses represent multiple recordings which sampled the same location. PB, protocerebral bridge; FB, fanshape body; EB, ellipsoid body.

101

Figure 3.2. Response types to binocular visual motion in left and right directions. A- E. Selected examples of various response types to left and right wide-field motion for individual units. Each row displays spike time rasters (top) and peristimulus time histograms (PSTH, binned at 50 ms, bottom) of the unit’s activity for each direction of motion (left motion for left columns and right motion for right columns). In each PSTH the red line represents mean firing rate of all trials. The gray box on each PSTH indicates the time when stripes were moving. A. This unit showed tonic directional opponency (excited tonically in one direction and inhibited tonically in the opposing direction). B.

102

This unit showed phasic excitation at the initiation of motion in both directions. C. This unit showed phasic inhibition at the initiation of motion in both directions. D. This unit showed periodic activity entrained to the temporal frequency of the motion for both directions. E. This unit also showed periodic activity for both directions but only showed an excitatory response (p < 0.05) in one phase of the stimulus. It is notable that this unit had very little spontaneous activity during the baseline. F. Recording sites where all periodic units were found. Three recording sites (yellow) had a disproportionate amount of units with periodic activity and no tonic opponency units. G. Table displaying the distribution of all units that responded to either binocular or monocular wide-field motion. Rows represent the number of units with the given temporal response type to left motion and columns for right motion. 37% of the responding units were directional and 46% of directional units were biased to the left (blue numbers), and 54% to the right (red numbers). Black numbers represent units that were not directional. Gray shaded boxes indicate pairs of response types with no representative units. Response type abbreviations: Phas - phasic, Excit - excitation, Inhib - inhibition, Init – to initiation of motion, Term – to termination of motion, Period – periodic entrained to the stimulus. PB, protocerebral bridge; FB, fanshape body; EB, ellipsoid body.

103

Figure 3.3. Compound eye occlusions and effects on tonic and phasic responses. A. Example unit response to occlusion conditions. PSTHs of responses to left and right motion (columns) for each occlusion condition (rows). The conditions are presented in the sequence they were performed during the experiment: 1) no occlusion, Binoc; 2) right compound eye occluded, MonoL; 3) both compound eyes occluded, BothOccl; 4) left compound eye occluded, MonoR. The responses for each occlusion condition are shown (bargraph, bottom). Responses to each direction of motion (left – blue, right – red) are shown as the mean difference in firing rate (spikes/s) between the response period and

104

baseline (4 s before the stimulus). The response to rightward motion in the monocular left condition is significantly less than (symbols, p<0.05, matched symbols represent response from same unit) those in the binocular condition (dashed lines) as are both responses when both eyes were occluded. The responses to right motion in the monocular left condition are also significantly less than those in the monocular right condition. B. Monocular bias of all units with tonic opponency responses. They all show a significant bias to a compound eye corresponding to the preferred direction (symbols, p<0.05, matched symbols represent responses from same unit). Line graphs show the mean responses during monocular conditions normalized to the binocular conditions (absolute value of the difference between activity during the response period and baseline activity during monocular condition divided by the difference during binocular condition) for monocular left conditions and monocular right conditions for all units with a left preferred direction (left) and all units with right preferred direction (right). Dashed line is the response for the binocular condition. Blue lines are response biases for leftward motion and red lines are response biases for rightward motion. C. Most units with phasic excitatory responses to both directions of motion show no monocular bias. Four units showed significant bias while ten did not.

Figure 3.4. Units with increased monocular and post-binocular responses. Units with increases in response amplitudes of monocular conditions compared to the pre-treatment binocular condition (dashed lines) also showed a significant increase in the post- binocular condition (PostBinoc). A. Monocular biases (absolute value of response amplitude normalized to pre-treatment binocular response, indicated by dashed line) of all tonic and phasic units that were tested with post-treatment binocular control. The only units with increased responses compared to pre-binocular treatment (all points above dashed line) also had increased responses in post-binocular conditions (†, p<0.05). B. The only two tonic opponency units that showed increased response amplitude for any direction of motion in a monocular condition that were tested with a post-binocular condition (all significant differences: *, p<0.05). Both also show increased response to post-treatment compared to pre-treatment binocular test. Unit 1 shows increases only for right motion (red), while unit 2 increases in responses to right and left (blue) directions of motion (inhibitory responses considered increased when response is more negative).

105

Figure 3.5. Compound eye occlusions and effects on periodic responses. A Periodograms of spectral densities for all responding units, where each row is the spectrogram of one responding unit (normalized to the maximum power for that unit). The upper group of units were classified as periodic by a non-uniformity test (Hodges-

106

Ajne, p < 0.05), and show clear peaks at the temporal frequency of the stimulus (2 Hz). B. Responses of two periodic units during occlusion conditions and bargraphs of the mean log base 10 of spectral powers at 2 Hz for each condition. Both units have significantly reduced responses for all monocular and control conditions when compared to the binocular condition (dashed lines). Only the right unit shows a significantly reduced response (*, p < 0.05) to the monocular right condition for both directions of motion. Most periodic units showed no monocular bias. C. Four periodic units had a significantly reduced responses (symbols, p<0.05, matched symbols represent responses from same unit) in the monocular right condition compared to the monocular left, which was seen for responses to both directions (blue – left motion, red – right motion) in all four units. All monocular responses were significantly less than the binocular condition (dashed line). Graphs show the mean log base 10 of the spectral power for each monocular condition. D. Five units were reduced in the monocular left condition for responses to right motion, but only two of them were also reduced for left motion. Two of these units had significantly larger responses than the binocular (above the dashed line) also had significantly larger responses in the post-treatment binocular condition compared to the pre-treatment binocular condition (open and closed squares, p<0.05). These were the only two periodic units with increased monocular responses compared to the binocular condition.

Figure 3.6. Silenced responses during all occlusion conditions. Example of a unit with phasic responses to both directions of motion in binocular conditions, but is silenced in all treatments where any compound eye is occluded.

107

Figure 3.7. Responses during compound eye controls and ocellar occlusions. A. PSTHs and bargraphs of response amplitude of a unit with periodic activity that persists during the control when both compound eyes are occluded. Bargraph shows response amplitude (mean spectral power at the stimulus temporal frequency) for these occlusion conditions for leftward (blue) and rightward (red) motion. Significance (*) was determined by a circular non-uniformity test (Hodges-Ajne, p < 0.05). Error bars are standard error. B. PSTHs and bargraphs of response amplitude from one unit showing tonic opponency to motion and with a response during the control conditions, but with reversed preferred direction. The bargraph shows response amplitude (difference of mean firing rate during response period and baseline) of these conditions. Significance (*) was determined by comparing response window activity to baseline (paired t-test, p < 0.05). C. Units lose the response that persists with compound eyes occluded when ocelli are also occluded.

108

Figure 3.8. Changes in background activity and responses with compound eye occlusions. A. Two units with reduced background activity in all occlusion conditions. One (left), the background activity was also reduced in the post-treatment binocular control and one (right) the activity recovered in the post-treatment control. Bargraphs depict the mean activity (spikes/s) of 4 s before the start of stimulus (blue bar) of each trial for each occlusion condition (not ordered by experimental sequence). Asterisk indicates significant differences from the pre-treatment binocular condition (dashed line). Error bars are standard error. B. Two types of background changes seen in periodic units, one with reduced activity when both eyes are occluded (left) and one with increased activity for same condition. C. Background activity and activity during response period for left and right motion of two periodic units, all of which are reduced during occlusion conditions. Background activity and responses were normalized to their respective binocular conditions. The left unit also has reduced background activity in the post- occlusion binocular control, but the responses to both directions of motion recover. The right unit response does not recover for either direction, similar to the background activity. This unit also shows a monocular bias toward the left compound eye in both background activity and response to both directions of motion, which is typical for all units with a bias.

109

Chapter 4:

Conclusion

110

In this thesis, I have demonstrated that the central complex (CX) is encoding rich

visual motion information in various groupings of neurons. As a population, the neurons

in the CX convey various motion characteristics, including speed, timing, direction, and

bilateral spatial information. These are all useful data for an animal that is moving

through its environment, particularly for maintaining a heading or speed. But most of

this information is found upstream from the CX as well, in neurons tangential to the

lobular plate of the optic lobes (Borst et al. 2010). Therefore, the question must be asked: why is this structure needed? The lobular plate tangential cell (LPTC) system is a rapid

system with direct connections to the motor circuitry in flies (Strausfeld and Bassemir

1985). As an associative region, there are several more synapses in the path to and from

the CX that information must pass before being transduced to a motor output. This

requires more time, which is a scarcity when dealing with such tasks as course correction

(Dickinson 2005). But these synapses are also involved in integrating much more

information.

In the complex environments animals navigate through, sensory information

exists in many modalities, including visual, tactile, olfactory, and auditory modalities.

Therefore, the ability to respond appropriately to an external stimulus is optimized by extracting as much of this information as possible. Polarized light rotation (Heinze and

Homberg 2007), narrow-field translational motion (Phillips-Portillo 2012; Rosner and

Homberg 2013), stripe position and orientation (Seelig and Jayaraman 2013), and spatial memory (Neuser et al. 2008; Ofstad et al. 2011) are all types of visual information encoded in the CX. In addition to visual inputs, neurons in the CX of the cockroach and fly are involved in encoding mechanosensory information from the antennae (Ritzmann

111 et al. 2008; Phillips-Portillo 2012). These cells also were often sensitive to changes in ambient light intensity or visual motion. Anecdotally, many cells sampled in both chapters of this manuscript were also tested with antennal deflections and a large proportion of them that responded to visual motion also responded to these mechanical stimuli. The multisensory properties of the CX suggest that this structure is not simply used for encoding individual sensory modalities, but rather integrates many aspects of the entire sensory surround to represent more abstract components of the environment. This is seen in the polarized light system, as E-vector orientation and location is integrated to match neural sensitivities to complex patterns of polarized light associated with sun position (Bech et al. 2014). This concept is likely extended to integrating various other modalities that are relevant, in concert with one another, to important features of the environment (e.g. shelters, food sources, or prey). This is reinforced by the presence of numerous neuromodulators in the CX, which possess the capacity to report the states of external and internal physiology (Nässel and Homberg 2006; Kahsai et al. 2010).

I believe it is also useful to pose the question of whether the CX, or other brain structures, across insects shares a common functional ground pattern or whether in different species they serve different and specific behavioral functions. Although experimental models are often chosen exactly because of what behaviors they excel at, what so far is known about the sensory representation in cockroach, locust (or monarch butterfly), or fly CX suggests that they might support different behavioral roles. Relevant differences exist between the fly and cockroach systems, which could also suggest reasons why additional visual control systems may exist. For instance, the mode of locomotion these systems operate in should not be ignored. While flight is a high speed,

112

dynamic behavior, the cockroach primarily walks on the ground at relatively slow speeds

(Bender et al. 2011). The rate of change of sensory information between these behaviors

is vastly different, and the amount of time the animal has to sample its surroundings is

dependent on this. This is particularly relevant in odor tracking, where adequately

sampling odor plumes to make useful comparisons between concentrations takes more

time and therefore walking insects and flying insects may have divergent mechanisms for

accomplishing these tasks (Willis et al. 2011).

Another important consideration is that cockroaches are nocturnal walking insects, but have eyes similar to that of flies. They have apposition eyes, which are not typically found in nocturnal insects since the ommatidia are optically isolated and therefore cannot take advantage of pooling light from many photoreceptors in dim light conditions. But apposition eyes are an excellent system for collecting visual motion, as seen in flies. This is also seen in the relatively close relative to the cockroach, the praying mantis (Misof et al. 2014). The mantis, with appositions eyes like the cockroach and fly, is a visual hunter and uses peering behaviors for moving between branches in its habitat (Rossel 1980; Poteser and Kral 1995; Kral 2012).

Ultimately, the LPTC and CX systems both encode directional wide-field visual motion that is known to be used in the optomotor response. Despite the studies that show the LPTC system to be sufficient to induce turning, the optomotor response still requires the CX. As discussed in chapter one, we proposed a parallel pathway model, where the

CX can function to provide additional, context dependent information to another, faster sensorimotor system, like the LPTC system. Under this model, the direct sensory-motor pathway provides a means for fast commands based on specific sensory conditions, while

113

the parallel structure of the CX monitors conditions within and surrounding the animal,

then modifies descending commands appropriately. This parallel model is similar to

circuits seen in the basal ganglia of mammals (Alexander et al. 1990; Hikosaka et al.

2000; Kandel et al. 2013), where direct pathways project from the motor cortex to spinal

tracts or, in the case of the oculomotor system, the superior colliculus, but a parallel

pathway through the basal ganglia influences the strength of those commands. This

structural analogy has also been proposed in a recent meta-analysis showing these

systems as having both functionally, anatomically, and neurochemically deep homology

to the mammalian basal ganglia (Strausfeld and Hirth 2013).

Under the parallel model, it is not as clear why silencing the CX would block optomotor responses or other behaviors (Ridgel et al. 2007; Harley and Ritzmann 2010).

It is possible that the CX inputs, even as spontaneous activity, is gating the optic lobe

descending pathway, so that only disinhibition from the CX allows behaviors to be

performed. This is another characteristic of the basal ganglia (Hikosaka et al. 1993).

Absent any such activation in the parallel pathway, the animal fails to react to external stimuli at all. A similar situation is seen in the cockroach escape system. Wind or tactile stimulation to the abdomen generates rapid leg movements that turn the insect away from a potential threat (Camhi and Tom 1978; Schaefer 1994). There is a direct pathway from wind sensors on the rear of the insect or tactile receptors on the abdomen to interneurons in the thoracic ganglia that ultimately activate fast motor neurons that rapidly extend or flex the leg joints (Ritzmann and Eaton 1997). This direct pathway excludes any involvement of circuits within the brain or suboesophageal ganglia, and yet, bilateral lesion of the neck connectives that disconnect these structures from the thoracic ganglia

114

have profound effects on the escape movements (Schaefer and Ritzmann 2001). In particular, lesion of the neck connectives or reversible sucrose block of these structures prevents the fast motor neurons of the prothoracic legs from reaching threshold in response to stimuli that would otherwise readily activate them in effective escape responses. In this way, external stimuli, such as walls, can influence the direct circuits and alter the behavior in parallel to the direct pathway (Ritzmann et al. 1991).

In this thesis, I described many sensory processing characteristics of the CX, most of which are novel. Although not at the level of detail the visual motion system has been described in the optic lobes, many key components related to locomotion control were found. Taken with the extensive intracellular recordings and morphology that has been documented above (Heinze and Reppert 2011; Homberg et al. 2011; Phillips-Portillo

2012) as well as neurogenetic studies including recent optical imaging (Strauss and Berg

2010; Seelig and Jayaraman 2013), the CX and its role in behavior is beginning to be understood. Whether speed, directionality, or binocular disparities are actually used in locomotion control still needs to be examined. The methods used in this thesis are well suited for behavioral studies (Guo et al. 2014). Therefore, these visual parameters can, and should, be used to probe the CX’s role in sensorimotor integration in the context of cockroach locomotion.

115

Appendix

This appendix includes additional data as a supplement to chapter 2.

Time after Injection Mean Activity Level – 20% Procaine Mean Activity Level – Saline (min) (%) (%) 0 58.4 58.3 15 54.4 45.8 30 50.5 50.3 45 53.8 52.6 60 55.4 62.6

Appendix Table 5.1. Activity levels of animals during treatment and control. To verify that the animal was still moving after injection and not paralyzed, we measured activity levels of the animals for treatments and controls. Activity levels were classified as the proportion of non-zero angular velocity samples, in both axes of rotation, to all samples. A motionless animal standing on the ball has a steady velocity of zero, in both axes. Proportions were found for each animal at each time after injection, and means (shown in table) were compared between procaine and saline injected groups (two- sample t-test, p < 0.05, n = 15 for each condition). Activity levels were not significantly different between the groups at any time interval. This data shows that the animals were, in fact, moving after injections. More specific analysis on parameters such as walking speed or trajectory was not performed.

116

Appendix Figure 5.1. Baseline firing rates of all motion responding units by response type to each direction of motion. Each cell in the grid contains the mean baseline firing rate (spikes/s) ± s.d. (spikes/s ) for all units with that combination of response types (sample sizes, n, given in cell). Orange cells represent units that only have tonic responses and yellow cells contain units that only have phasic responses. White cells contain units with both tonic and phasic responses. The green cell contains units with periodic responses entrained to the temporal frequency of the stimulus. Gray cells represent combinations of response types with no representative units found. Notable trends include units with phasic excitatory responses (to either direction or to one direction with no response to the other direction) had relatively lower baseline firing rates and were fairly consistent (2.5 spikes/s ± 2.3 spikes/s, mean ± s.d.). Units with only tonic excitatory responses, had higher baseline firing rates of 7.4 spikes/s ± 8.4 spikes/s. Phasic inhibitory units also had higher baseline firing rates (17.6 spikes/s ± 12.0 spikes/s), while tonic inhibitory units had lower baseline firing rates (4.3 spikes/s ± 4.2 spikes/s). Tonic opponency units (with excitatory responses in one direction and inhibitory responses in the other direction) were also moderately high, similar to tonic excitatory responses (8.3 spikes/s ± 7.7 spikes/s). Response type abbreviations: Phas - phasic, Excit - excitation, Inhib - inhibition, Init - to initiation of motion, Term - to termination of motion.

117

Appendix Figure 5.2. Analog controls for digital stimulus. To verify that the refresh rate of the stimuli was sufficient, four animals were run with an analog display system, in addition to the digital system. This consisted of a handheld projector (3M MPro110, LED backlit) used as a light source, centered in a motorized rotating cylinder of opaque and transparent stripes used to project stripe fields onto the screen. The striped cylinder was powered by a servomotor (HiTec HSR-5995TG) and controlled with an Isopod™ control board (New Micros, Inc.) using custom software written by William Lewinger. The cylinder was also run with the light off, to control for any additional artifacts from the motor or cylinder. Here, we show two units representative of all units from the four experiments. Both units shown are from the same recording and responded similarly to both conditions and had no response to the control. The first condition (first row) was the same digital projector and stripe generator used for all other experiments. The second condition (second row) was the analog stripe generator and the third (third row) was a control of the analog condition with the light off. The first column is a unit with similar phasic responses to both stripe conditions moving to the right (the unit did not respond to stripes moving left). The second and third columns are the responses to left and right motion, respectively, from a tonic unit with directional opponency. No statistical difference was found between responses to digital and analog motion stimulus. All temporal frequencies were 4 Hz.

118

Bibliography

Alexander GE, Crutcher MD (1990) Functional architecture of basal ganglia circuits: neural substrates of parallel processing. Trends Neurosci 13:266–71.

Alexander GE, Crutcher MD, DeLong MR (1990) Basal ganglia-thalamocortical circuits: parallel substrates for motor, oculomotor, “prefrontal” and “limbic” functions. Prog Brain Res 85:119–46.

Baker DA, Beckingham KM, Armstrong JD (2007) Functional Dissection of the Neural Substrates for Gravitaxic Maze Behavior in Drosophila melanogaster. J Comp Neurol 764:756–764. doi: 10.1002/cne

Bausenwein B, Müller NR, Heisenberg M (1994) Behavior-dependent activity labeling in the central complex of Drosophila during controlled visual stimulation. J Comp Neurol 340:255–268. doi: 10.1002/cne.903400210

Bech M, Homberg U, Pfeiffer K (2014) Receptive Fields of Locust Brain Neurons Are Matched to Polarization Patterns of the Sky. Curr Biol 24:1–6. doi: 10.1016/j.cub.2014.07.045

Bender J a, Simpson EM, Tietz BR, et al. (2011) Kinematic and behavioral evidence for a distinction between trotting and ambling gaits in the cockroach Blaberus discoidalis. J Exp Biol 214:2057–2064. doi: 10.1242/jeb.056481

Bender JA, Pollack AJ, Ritzmann RE (2010) Neural activity in the central complex of the insect brain is linked to locomotor changes. Curr Biol 20:921–926. doi: 10.1016/j.cub.2010.03.054

Bhatia KP, Marsden CD (1994) The behavioural and motor consequences of focal lesions of the basal ganglia in man. Brain 117:859–76.

Bonin V, Histed MH, Yurgenson S, Reid RC (2011) Local diversity and fine-scale organization of receptive fields in mouse visual cortex. J Neurosci 31:18506–21. doi: 10.1523/JNEUROSCI.2974-11.2011

Borst A (2014) Fly visual course control: behaviour, algorithms and circuits. Nat Rev Neurosci. doi: 10.1038/nrn3799

Borst A, Egelhaaf M (1989) Principles Motion Detection. Trends Neurosci 12:297–306.

Borst A, Haag J (2002) Neural networks in the cockpit of the fly. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 188:419–37. doi: 10.1007/s00359-002-0316- 8

119

Borst A, Haag J, Reiff DF (2010) Fly motion vision. Annu Rev Neurosci 33:49–70. doi: 10.1146/annurev-neuro-060909-153155

Büschges A, Akay T, Gabriel JP, Schmidt J (2008) Organizing network action for locomotion: insights from studying insect walking. Brain Res Rev 57:162–171. doi: 10.1016/j.brainresrev.2007.06.028

Butler R (1973) The Anatomy of the Compound Eye of Periplaneta americana L. J Comp Physiol 83:223–238.

Buzsáki G (2004) Large-scale recording of neuronal ensembles. Nat Neurosci 7:446–451. doi: 10.1038/nn1233

Camhi JM, Johnson EN (1999) High-frequency steering maneuvers mediated by tactile cues: antennal wall-following in the cockroach. J Exp Biol 202:631–643.

Camhi JM, Tom W (1978) The Escape Behavior of the Cockroach Periplaneta americana I . Turning Response to Wind Puffs *. J Comp Physiol A 128:193–201.

Devaud J-M, Blunk A, Podufall J, et al. (2007) Using local anaesthetics to block neuronal activity and map specific learning tasks to the mushroom bodies of an insect brain. Eur J Neurosci 26:3193–3206. doi: 10.1111/j.1460-9568.2007.05904.x

Dickinson MH (2005) The Initiation and Control of Rapid Flight Maneuvers in Fruit Flies. Integr Comp Biol 281:274–281.

Egelhaaf M, Borst a (1989) Transient and steady-state response properties of movement detectors. J Opt Soc Am A 6:116–27.

Egelhaaf M, Borst A (1993) Movement detection in arthropods. In: Miles FA, Wallman J (eds) Vis. Motion its Role Stabalization Gaze. Elsevier, Amsterdam, pp 53–77

Egelhaaf M, Kern R (2002) Vision in flying insects. Curr Opin Neurobiol 12:699–706. doi: 10.1016/S0959-4388(02)00390-2

Egelhaaf M, Kern R, Krapp HG, et al. (2002) Neural encoding of behaviourally relevant visual-motion information in the fly. Trends Neurosci 25:96–102.

Farrow K, Haag J, Borst A (2006) Nonlinear, binocular interactions underlying flow field selectivity of a motion-sensitive neuron. Nat Neurosci 9:1312–20. doi: 10.1038/nn1769

Fee MS (2012) Oculomotor learning revisited: a model of reinforcement learning in the basal ganglia incorporating an efference copy of motor actions. Front Neural Circuits 6:38. doi: 10.3389/fncir.2012.00038

120

Fee MS, Scharff C (2010) The Songbird as a Model for the Generation and Learning of Complex Sequential Behaviors. ILAR J 5:362–377.

Fellous J-M, Tiesinga PHE, Thomas PJ, Sejnowski TJ (2004) Discovering spike patterns in neuronal responses. J Neurosci 24:2989–3001. doi: 10.1523/JNEUROSCI.4649- 03.2004

Flaherty AW, Graybiel AM (1994) Input-Output Organization of the Sensorimotor Striatum in the Squirrel Monkey. J Neurosci 14:599–610.

Fox JL, Frye MA (2014) Figure-ground discrimination behavior in Drosophila. II. Visual influences on head movement behavior. J Exp Biol 217:570–9. doi: 10.1242/jeb.080192

Gabbiani F, Krapp HG, Laurent G (1999) Computation of object approach by a wide- field, motion-sensitive neuron. J Neurosci 19:1122–41.

Geiger G, Nässel DR (1981) Visual orientation behaviour of flies after selective laser beam ablation of interneurones. Nature 293:398–399.

Gibson (1950) The perception of the visual world. Houghton Mifflin, Oxford, England

Gotz KG (1968) Flight Control in Drosophila by Visual Perception of Motion. Kybernetic 9:

Grillner S, Wallén P, Saitoh K, et al. (2008) Neural bases of goal-directed locomotion in vertebrates--an overview. Brain Res Rev 57:2–12. doi: 10.1016/j.brainresrev.2007.06.027

Guo P (2014) The role of the central complex in adaptive locomotor behavior in cockroaches. Case Western Reserve University

Guo P, Ritzmann RE (2013) Neural activity in the central complex of the cockroach brain is linked to turning behaviors. J Exp Biol 216:992–1002. doi: 10.1242/jeb.080473

Guo P, Varga AG, Martin JP, Ritzmann RE (2014) Extracellular wire tetrode recording in brain of freely walking insects. J Vis Exp 86:e51337.

Guthrie DM, Tindall AR (1968) The Biology of the Cockroach. Z Allg Mikrobiol 9:145– 151. doi: 10.1002/jobm.19690090615

Haikala V, Joesch M, Borst A, Mauss AS (2013) Optogenetic control of fly optomotor responses. J Neurosci 33:13927–34. doi: 10.1523/JNEUROSCI.0340-13.2013

Hanesch U, Fischbach K, Heisenberg M (1989) Neuronal architecture of the central complex in Drosophila melanogaster. Cell Tissue Res 257:343–366.

121

Harley CM, English B a, Ritzmann RE (2009) Characterization of obstacle negotiation behaviors in the cockroach, Blaberus discoidalis. J Exp Biol 212:1463–1476. doi: 10.1242/jeb.028381

Harley CM, Ritzmann RE (2010) Electrolytic lesions within central complex neuropils of the cockroach brain affect negotiation of barriers. J Exp Biol 213:2851–2864. doi: 10.1242/jeb.042499

Hassenstein B, Reichardt (1956) Systemtheoretische Analyse der Zeit-Reihenfolgen- und Vorzeichenauswer- tung bei der Bewegungsperzeption des Rüsselkäfers Chlorophanus. Z Naturforsch 11b:513–524.

Hausen K (1984) The lobula-complex of the fly: structure, function, and significance in visual behavior. Photoreception Vis. Invertebr. pp 523–559

Heimonen K, Salmela I, Kontiokari P, Weckström M (2006) Large functional variability in cockroach photoreceptors: optimization to low light levels. J Neurosci 26:13454– 62. doi: 10.1523/JNEUROSCI.3767-06.2006

Heinrich R, Kunst M, Wirmer A (2012) Reproduction-related sound production of grasshoppers regulated by internal state and actual sensory environment. Front Neurosci 6:89. doi: 10.3389/fnins.2012.00089

Heinze S, Gotthardt S, Homberg U (2009) Transformation of polarized light information in the central complex of the locust. J Neurosci 29:11783–11793. doi: 10.1523/JNEUROSCI.1870-09.2009

Heinze S, Homberg U (2007) Maplike representation of celestial E-vector orientations in the brain of an insect. Science 315:995–7. doi: 10.1126/science.1135531

Heinze S, Homberg U (2008) Neuroarchitecture of the central complex of the desert locust: Intrinsic and columnar neurons. J Comp Neurol 511:454–478. doi: 10.1002/cne.21842

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. J Neurosci 29:4911–4921. doi: 10.1523/JNEUROSCI.0332-09.2009

Heinze S, Reppert SM (2011) Sun compass integration of skylight cues in migratory monarch butterflies. Neuron 69:345–358. doi: 10.1016/j.neuron.2010.12.025

Heisenberg M (2003) Mushroom body memoir: from maps to models. Nat Rev Neurosci 4:266–275. doi: 10.1038/nrn1074

Heisenberg M, Wonneberger R, Wolf R (1978) optomotor-blind H31 - a Drosophila Mutant of the Lobula Plate Giant Neurons. J Comp Physiol A 296:287–296.

122

Herbert Z, Rauser S, Williams L, et al. (2010) Developmental expression of neuromodulators in the central complex of the grasshopper Schistocerca gregaria. J Morphol 271:1509–1526. doi: 10.1002/jmor.10895

Hikosaka O, Matsumura M, Kojima J, Gardiner T (1993) Role of basal ganglia in initiation and suppression of saccadic eye movements. In: Mano N, Hamada I, DeLong M (eds) Role Cerebellum Basal Ganglia Volunt. Mov. Elsevier, Amsterdam, pp 213–219

Hikosaka O, Takikawa Y, Kawagoe R (2000) Role of the basal ganglia in the control of purposive saccadic eye movements. Physiol Rev 80:953–78.

Homberg U (2008) Evolution of the central complex in the arthropod brain with respect to the visual system. Arthropod Struct Dev 37:347–362. doi: 10.1016/j.asd.2008.01.008

Homberg U (1991) Neuroarchitecture of the central complex in the brain of the locust Schistocerca gregaria and S. americana as revealed by serotonin immunocytochemistry. J Comp Neurol 303:245–254. doi: 10.1002/cne.903030207

Homberg U, Heinze S, Pfeiffer K, et al. (2011) Central neural coding of sky polarization in insects. Philos Trans R Soc Lond B Biol Sci 366:680–687. doi: 10.1098/rstb.2010.0199

Homberg U, Kingan T, Hildebrand JG (1987) Immunocytochemistry of GABA in the brain and suboesophogeal ganglion of Manduca sexta. Cell Tissue Res 248:1–24.

Homberg U, Vitzthum H, Müller M, Binkle U (1999) Immunocytochemistry of GABA in the central complex of the locust Schistocerca gregaria: identification of immunoreactive neurons and colocalization with neuropeptides. J Comp Neurol 409:495–507.

Homberg U, Wiirden S, Dircksen H, Rao KR (1991) Comparative anatomy of pigment- dispersing hormone-immunoreactive neurons in the brain of orthopteroid insects. Cell Tissue Res 266:343–357.

Huber F (1960) Untersuchungen über die Funktion des Zentralner-vensystems und insbesondere des Gehirns bei der Fortbewegung und Lauterzeugung der Grillen. Z Vgl Physiol 44:60–132.

Huston SJ, Krapp HG (2008) Visuomotor transformation in the fly gaze stabilization system. PLoS Biol 6:e173. doi: 10.1371/journal.pbio.0060173

Ilius M, Wolf R, Heisenberg M (1994) The central complex of Drosophila melanogaster is involved in flight control: studies on mutants and mosaics of the gene ellipsoid body open. J Neurogenet 9:189–206.

123

Iwano M, Hill ES, Mori A, et al. (2010) Neurons associated with the flip-flop activity in the lateral accessory lobe and ventral protocerebrum of the silkworm moth brain. J Comp Neurol 518:366–388. doi: 10.1002/cne.22224

Kahsai L, Carlsson MA, Winther AME, Nässel DR (2012) Distribution of metabotropic receptors of serotonin, dopamine, GABA, glutamate, and short neuropeptide F in the central complex of Drosophila. Neuroscience 208:11–26. doi: 10.1016/j.neuroscience.2012.02.007

Kahsai L, Martin J-R, Winther AME (2010) Neuropeptides in the Drosophila central complex in modulation of locomotor behavior. J Exp Biol 213:2256–2265. doi: 10.1242/jeb.043190

Kahsai L, Winther ÅME (2011) Chemical neuroanatomy of the Drosophila central complex: Distribution of multiple neuropeptides in relation to neurotransmitters. J Comp Neurol 519:290–315. doi: 10.1002/cne.22520

Kandel E, Schwartz J, Jessell T (2013) Principles of neural science, 5th ed. McGraw-Hill

Karmeier K, van Hateren JH, Kern R, Egelhaaf M (2006) Encoding of naturalistic optic flow by a population of blowfly motion-sensitive neurons. J Neurophysiol 96:1602– 14. doi: 10.1152/jn.00023.2006

Kathman ND, Kesavan M, Ritzmann RE (2014) Encoding wide-field motion and direction in the central complex of the cockroach Blaberus discoidalis. J Exp Biol 217:4079–4090. doi: 10.1242/jeb.112391

Kral K (2012) The functional significance of mantis peering behaviour. Eur J Entomol 109:295–301.

Krapp HG (2000) Neuronal matched filters for optic flow processing in flying insects. Int Rev Neurobiol 44:93–120.

Krapp HG, Hengstenberg R, Egelhaaf M (2001) Binocular Contributions to Optic Flow Processing in the Fly Visual System. J Neurophysiol 85:724–734.

Kunst M, Pförtner R, Aschenbrenner K, Heinrich R (2011) Neurochemical architecture of the central complex related to its function in the control of grasshopper acoustic communication. PLoS One 6:e25613. doi: 10.1371/journal.pone.0025613

Labhart T, Meyer EP (1999) Detectors for polarized skylight in insects: a survey of ommatidial specializations in the dorsal rim area of the compound eye. Microsc Res Tech 47:368–379. doi: 10.1002/(SICI)1097-0029(19991215)47:6<368::AID- JEMT2>3.0.CO;2-Q

124

Lagae L, Maes H, Raiguel S, et al. (1994) Responses of Macaque STS Neurons to Optic Flow Components : A Comparison of Areas MT and MST. J. Neurophysiol. 71:

Lappe M, Bremmer F, van den Berg AV (1999) Perception of self-motion from visual flow. Trends Cogn Sci 3:329–336.

Lappe M, Grigo a (1999) How stereovision interacts with optic flow perception: neural mechanisms. Neural Netw 12:1325–1329.

Lin C-Y, Chuang C-C, Hua T-E, et al. (2013) A Comprehensive Wiring Diagram of the Protocerebral Bridge for Visual Information Processing in the Drosophila Brain. Cell Rep 3:1739–1753. doi: 10.1016/j.celrep.2013.04.022

Loesel R, Nässel DR, Strausfeld NJ (2002) Common design in a unique midline neuropil in the brains of arthropods. Arthropod Struct Dev 31:77–91. doi: 10.1016/S1467- 8039(02)00017-8

Mappes M, Homberg U (2007) Surgical lesion of the anterior optic tract abolishes polarotaxis in tethered flying locusts, Schistocerca gregaria. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 193:43–50. doi: 10.1007/s00359-006-0166-x

Martin JR, Raabe T, Heisenberg M (1999) Central complex substructures are required for the maintenance of locomotor activity in Drosophila melanogaster. J Comp Physiol A 185:277–88.

Miles FA (1997) Visual stabilization of the eyes in primates. Curr Opin Neurobiol 7:867– 871.

Misof B, Liu S, Meusemann K, et al. (2014) Phylogenomics resolves the timing and pattern of insect evolution. Science (80- ) 346:763–767. doi: 10.1126/science.1257570

Mitchiner JC, Pinto LH, Vanable JW (1976) Visually evoked eye movements in the mouse (Mus musculus). Vision Res 16:1169–1171.

Mizunami M (1993) Processing of Contrast Signals in the Insect Ocellar System. Zoolog Sci 11:175–190.

Mizunami M (1995) Functional Diversity of Neural Organization in Insect Ocellar Systems. Vision Res 35:443–452.

Mizunami M, Weibrecht JM, Strausfeld NJ (1998) Mushroom bodies of the cockroach: their participation in place memory. J Comp Neurol 402:520–37.

125

Mu L, Bacon JP, Strausfeld NJ (2012) Optic glomeruli and their inputs in Drosophila share an organizational ground pattern with antennal lobes. J Neurosci 32:6061– 6071. doi: 10.1523/JNEUROSCI.0221-12.2012.Optic

Mu L, Ritzmann RE (2005) Kinematics and motor activity during tethered walking and turning in the cockroach, Blaberus discoidalis. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 191:1037–1054. doi: 10.1007/s00359-005-0029-x

Muller D, Staffelt D, Fiala A, Menzel R (2003) Procaine impairs learning and memory consolidation in the honeybee. Brain Res 977:124–127.

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 Res 288:159–176.

Nässel DR, Homberg U (2006) Neuropeptides in interneurons of the insect brain. Cell Tissue Res 326:1–24. doi: 10.1007/s00441-006-0210-8

Neuser K, Triphan T, Mronz M, et al. (2008) Analysis of a spatial orientation memory in Drosophila. Nature 453:1244–1247. doi: 10.1038/nature07003

Ofstad T a, Zuker CS, Reiser MB (2011) Visual place learning in Drosophila melanogaster. Nature 474:204–7. doi: 10.1038/nature10131

Okada J, Toh Y (2001) Peripheral representation of antennal orientation by the scapal hair plate of the cockroach Periplaneta americana. J Exp Biol 204:4301–4309.

Otto D (1971) Untersuchungen zur zentralnervösen Kontrolle der Lauterzeugung von Grillen. Z Vgl Physiol 74:227–271.

Pfeiffer K, Homberg U (2014) Organization and Functional Roles of the Central Complex in the Insect Brain. Annu Rev Entomol 59:165–184. doi: 10.1146/annurev- ento-011613-162031

Phillips-Portillo J (2012) The central complex of the flesh fly, Neobellieria bullata: Recordings and morphologies of protocerebral inputs and small field neurons. J Comp Neurol 520:3088–3104. doi: 10.1002/cne.23134

Pick B, Buchner E (1979) Visual movement detection under ligh- and dark- adaptation in the fly, Musca domestica. J Comp Physiol 54:45–54.

Pick S, Strauss R (2005) Goal-driven behavioral adaptations in gap-climbing Drosophila. Curr Biol 15:1473–1478. doi: 10.1016/j.cub.2005.07.022

126

Poeck B, Triphan T, Neuser K, Strauss R (2008) Locomotor control by the central complex in Drosophila-An analysis of the tay bridge mutant. Dev Neurobiol 68:1046–58. doi: 10.1002/dneu.20643

Popov a. V., Peresleni a. I, Ozerskii P V., et al. (2005) The Role of the Flabellar and Ellipsoid Bodies of the Central Complex of the Brain of Drosophila Melanogaster in the Control of Courtship Behavior and Communicative Sound Production in Males. Neurosci Behav Physiol 35:741–750. doi: 10.1007/s11055-005-0118-x

Popov A V., Sitnik NA, Savvateeva-Popova E V., et al. (2003) The role of central parts of the brain in the control of sound production during courtship in Drosophila melanogaster. Neurosci Behav Physiol 33:53–65.

Portugues R, Engert F (2009) The neural basis of visual behaviors in the larval zebrafish. Curr Opin Neurobiol 19:644–647. doi: 10.1016/j.conb.2009.10.007

Poteser M, Kral K (1995) Visual distance discrimination between stationary targets in Praying Mantis: An index of the use of motion parallax. J Exp Biol 2137:2127– 2137.

Redgrave P, Gurney K (2006) The short-latency dopamine signal: a role in discovering novel actions? 7:967–975.

Redgrave P, Rodriguez M, Smith Y, et al. (2010) Goal-directed and habitual control in the basal ganglia: implications for Parkinson’s disease. Nat Rev Neurosci 11:760– 772. doi: 10.1038/nrn2915.Goal-directed

Reisenman C, Haag J, Borst a. (2003) Adaptation of response transients in fly motion vision. I: Experiments. Vision Res 43:1293–1309. doi: 10.1016/S0042- 6989(03)00091-9

Reiser MB, Dickinson MH (2008) A modular display system for insect behavioral neuroscience. J Neurosci Methods 167:127–39. doi: 10.1016/j.jneumeth.2007.07.019

Reppert SM, Gegear RJ, Merlin C (2010) Navigation mechanisms of migrating monarch butterflies. Trends Neurosci 33:399–406. doi: 10.1016/j.tins.2010.04.004.Navigational

Ridgel AL, Alexander BE, Ritzmann RE (2007) Descending control of turning behavior in the cockroach, Blaberus discoidalis. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 193:385–402. doi: 10.1007/s00359-006-0193-7

Ridgel AL, Ritzmann RE (2005) Effects of neck and circumoesophageal connective lesions on posture and locomotion in the cockroach. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 191:559–573. doi: 10.1007/s00359-005-0621-0

127

Ritzmann RE, Büschges A (2007) Adaptive motor behavior in insects. Curr Opin Neurobiol 17:629–636. doi: 10.1016/j.conb.2008.01.001

Ritzmann RE, Eaton R (1997) Neural substrates for initiation of startle responses. In: Stein P, Grillner S, Selverston A, Stuart D (eds) Neurons, Networks Mot. Behav. MIT Press, Cambridge, MA, pp 33–44

Ritzmann RE, Harley CM, Daltorio KA, et al. (2012) Deciding which way to go: how do insects alter movements to negotiate barriers? Front Neurosci 6:97. doi: 10.3389/fnins.2012.00097

Ritzmann RE, Pollack AJ, Hudson SE, Hyvonen A (1991) Convergence of multi-modal sensory signals at thoracic interneurons of the escape system of the cockroach, Periplaneta americana. Brain Res 563:175–183.

Ritzmann RE, Ridgel AL, Pollack AJ (2008) Multi-unit recording of antennal mechano- sensitive units in the central complex of the cockroach, Blaberus discoidalis. J Comp Physiol A Neuroethol Sens Neural Behav Physiol 194:341–360. doi: 10.1007/s00359-007-0310-2

Rosner R, Homberg U (2013) Widespread sensitivity to looming stimuli and small moving objects in the central complex of an insect brain. J Neurosci 33:8122–8133. doi: 10.1523/JNEUROSCI.5390-12.2013

Rossel S (1980) Foveal Fixation and Tracking in the Praying Mantis. J Comp Physiol A 331:307–331.

Sakai T, Kitamoto T (2006) Differential Roles of Two Major Brain Structures , Mushroom Bodies and Central Complex , for Drosophila Male Courtship Behavior. doi: 10.1002/neu

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. J Neurophysiol 99:667–682. doi: 10.1152/jn.00784.2007

Sanger TD (2003) Neural population codes. Curr Opin Neurobiol 13:238–249. doi: 10.1016/S0959-4388(03)00034-5

Schaefer PL (1994) Motion analysis of escape movements evoked by tactile stimulation in the cockroach Periplaneta americana. J Exp Biol 190:287–294.

Schaefer PL, Ritzmann RE (2001) Descending Influences on Escape Behavior and Motor Pattern in the Cockroach. J Neurobiol 49:9–28.

Seelig JD, Jayaraman V (2013) Feature detection and orientation tuning in the Drosophila central complex. Nature 503:262–266. doi: 10.1038/nature12601

128

Srinivasan M V, Poteser M, Kral K (1999) Motion detection in insect orientation and navigation. Vision Res 39:2749–66.

Steinman RM, Collewijn H (1980) Binocular retinal image motion during active head rotation. Vision Res 20:415–429.

Strausfeld NJ (1999) A brain region in insects that supervises walking. Prog Brain Res 123:273–284.

Strausfeld NJ (2012) Arthropod Brains: Evolution, Functional Elegance, and Historical Significance. Harvard University Press, Cambridge, MA

Strausfeld NJ (1976) Atlas of the Insect Brain. Springer-Verlag, New York

Strausfeld NJ, Bassemir UK (1985) Lobula plate and ocellar interneurons converge onto a cluster of descending neurons leading to neck and leg motor neuropil in Calliphora erythrocephala. Cell Tissue Res 240:617–640.

Strausfeld NJ, Hirth F (2013) Deep homology of arthropod central complex and vertebrate basal ganglia. Science 340:157–61. doi: 10.1126/science.1231828

Strausfeld NJ, Seyan HS (1985) Convergence of visual, haltere, and prosternal inputs at neck motor neurons of Calliphora erythrocephala. Cell Tissue Res 240:601–615.

Strauss R (2002) The central complex and the genetic dissection of locomotor behaviour. Curr Opin Neurobiol 12:633–638. doi: 10.1016/S0959-4388(02)00385-9

Strauss R, Berg C (2010) The central control of oriented locomotion in insects - towards a neurobiological model. 2010 Int Jt Conf Neural Networks 1–8. doi: 10.1109/IJCNN.2010.5596461

Strauss R, Heisenberg M (1993) A higher control center of locomotor behavior in the Drosophila brain. J Neurosci 13:1852–1861.

Suthers RA (1966) Optomotor Responses by Echolocating Bats. Science (80- ) 152:1102–1104. doi: 10.1126/science.152.3725.1102

Szczecinski NS, Brown AE, Bender JA, et al. (2014) A neuromechanical simulation of insect walking and transition to turning of the cockroach Blaberus discoidalis. Biol Cybern 108:1–21. doi: 10.1007/s00422-013-0573-3

Taylor GK, Krapp HG (2007) Sensory Systems and Flight Stability: What do Insects Measure and Why? Adv In Insect Phys. doi: 10.1016/S0065-2806(07)34005-8

Triphan T, Poeck B, Neuser K, Strauss R (2010) Visual targeting of motor actions in climbing Drosophila. Curr Biol 20:663–668. doi: 10.1016/j.cub.2010.02.055

129

Tryba AK, Ritzmann RE (2000) Multi-joint coordination during walking and foothold searching in the Blaberus cockroach. I. Kinematics and electromyograms. J Neurophysiol 83:3323–3336.

Wehner R (2001) Polarized vision - a uniform sensory capacity? J Exp Biol 2596:2589– 2596.

Weir PT, Schnell B, Dickinson MH (2014) Central complex neurons exhibit behaviorally gated responses to visual motion in Drosophila. J Neurophysiol 111:62–71. doi: 10.1152/jn.00593.2013

Willis MA, Avondet JL, Zheng E (2011) The role of vision in odor-plume tracking by walking and flying insects. J Exp Biol 214:4121–32. doi: 10.1242/jeb.036954

Wilson M (1978) The Functional Organisation of Locust Ocelli. J Comp Physiol A 316:297–316.

130