Strategies of the honeybee Apis mellifera during visual search for vertical targets presented at various heights: a role for spatial attention?

When honeybees are presented with a colour discrimination task, they tend to choose swiftly and accurately when objects are presented in the ventral part of their frontal visual field. In contrast, poor performance is observed when objects appear in the dorsal part. Here we investigate if this asymmetry is caused by fixed search patterns or if bees can use alternative search mechanisms such as spatial attention, which allows flexible focusing on different areas of the visual field. We asked individual honeybees to choose an orange rewarded target among blue distractors. Target and distractors were presented in the ventral visual field, the dorsal field or both. Bees presented with targets in the ventral visual field consistently had the highest search efficiency, with rapid decisions, high accuracy and direct flight paths. In contrast, search performance for dorsally located targets was inaccurate and slow at the beginning of the test phase, but bees increased their search performance significantly after a few learning trials: they found the target faster, made fewer errors and flew in a straight line towards the target. However, bees needed thrice as long to improve the search for a dorsally located target when the target’s position changed randomly between the ventral and the dorsal visual field. We propose that honeybees form expectations of the location of the target’s appearance and adapt their search strategy accordingly. Different possible mechanisms of this behavioural adaptation are discussed.


Introduction
When honeybees search for targets presented on a vertical plane, they show a distinct spatial asymmetry in colour and pattern learning between the ventral and the dorsal half of their frontal visual field. They easily learn a target when its defining features are perceived by the ventro-frontal area of the eye, but are less accurate when the crucial features of the target appear in the dorso-frontal area (Baumgärtner, 1928;Chittka et al., 1988;Giurfa et al., 1999;Lehrer, 1999;Morawetz & Spaethe, 2012;Wehner, 1972). This behavioural asymmetry might in theory be explained by specialization in eye morphology. This is found in the eye of the honeybee drone, where the dorsal area is adapted to queen detection by increasing visual acuity and sensitivity, which is partly achieved by enlarged facet diameters and a reduction of the interommatidial angles (Menzel et al., 1991;Seidl, 1982;Streinzer et al., 2013;van Praagh et al., 1980). However, such regional specialisation in eye optics is not found in the worker honeybee, where the interommatidial angles and facet diameters are similar in the frontal visual field 30° below and above the horizontal plane (Seidl, 1982).
It is also possible that regional specialisation of the visual system occurs at the neuronal level. Indeed, ascending neurons of the medulla show differences in arborisation patterns between the ventral and dorsal area (Ehmer & Gronenberg, 2002), indicating that both areas of the visual field are to some extent processed separately. Furthermore, the output of these two areas becomes segregated in the anterior optic tubercle and in the collar of the calyxes of the mushroom bodies (Ehmer & Gronenberg, 2002;Mota et al., 2011). The data hint at differences in neuronal processing between these two eye regions and correspond with the behavioural evidence of a dorso-ventral differentiation (Baumgärtner, 1928;Chittka et al., 1988;Giurfa et al., 1999;Lehrer, 1999;Morawetz & Spaethe, 2012;Wehner, 1972).
An alternative explanation for the dorso-ventral asymmetry is the usage of attentional mechanisms, which focuses the visual processing capacity of the brain flexibly to the currently most important area in space (spatial attention;Carrasco & McElree, 2001;Druker & Anderson, 2010;Geng & Behrmann, 2002;Posner, 1980;Yantis & Jonides, 1984). Attention here can be thought of as a kind of 'inner' eye, focusing on a spatial subset of the information that is available from the visual sensory periphery. The existence of spatial attention is well known in vertebrates, but has only recently been described for the first time in insects, the fruit fly Drosophila melanogaster (Sareen et al., 2011) and the honeybee Apis mellifera (Paulk et al., 2014).
Spatial attention optimizes search processes in detection tasks, where the subject has an expectation of the appearance of the object, using external cueing or own experience (Posner, 1980). Search efforts can then be directed to this region which leads to faster and more accurate decisions (Carrasco & McElree, 2001). Hence, spatial attention would be a useful tool for foraging bees helping to adapt to various spatial settings of flower distribution. The dorso-ventral asymmetry observed in visual discrimination tasks could be explained by attention being focused on the ventral part of the visual field by default, but the attentional focus could be moved to other parts of the visual field, if necessary. To test if bees can employ spatial attention during foraging, we confronted honeybees with three search scenarios differing in the positioning of the target in the visual field. This approach allows to analyse the changes in search performance and flight behaviour over time and therefore to identify possible adaptations of the search behaviour to the particular target presentation.

Material and methods
The experiments were conducted between July and September in 2011 on the terraces of the Biozentrum, University of Vienna, where several hives of Apis mellifera were located. Bees were trained to an experimental box and marked individually.

Experimental setup
A wooden box (30 × 54 × 40 cm) served as experimental arena ( Figure 1A; see also Morawetz & Spaethe, 2012). The bees could enter the box through a Plexiglas tube on the front and shutters in the tube allowed to control access to the box. Two video cameras (DCR-SR55, Sony, Minato, Tokyo, Japan) were placed above and at one side of the box to record the flights through a small-meshed net and a Plexiglas wall, respectively.
The back wall was divided into nine fields (3 rows by 3 columns) of which only the top and the bottom row were used for training and testing. In the centre of each field a platform provided reward for a correct choice (1 M sucrose solution ad libitum) and punishment for an incorrect choice (0.1% quinine solution w/w; Chittka et al., 2003), respectively. Stimulus discs (K + E, Stuttgart-Feuerbach, Germany) of 9 cm diameter were cut from coloured paper; the target discs were orange (HKS 7N) and distractor discs were blue (HKS 49N, for colour details see Figure 1B,C). The stimuli subtended a visual angle of >15° on the eye of the bee when sitting at the box entrance, which enabled the bee to perceive chromatic information from the beginning of the search (Dyer et al., 2008;Giurfa et al., 1996). The stimuli were presented on the back wall of the arena subtending a visual field of 82° in the horizontal and 60° in the vertical on the eye of a bee ( Figure 1A). Stimuli located in the dorsal row were thus perceived by the dorsal part of the eye, while objects in the ventral row were perceived by the ventral part of the eye, when the bee entered the arena ( Figure 1A).

Training and testing
Bees were trained to enter the experimental arena and to search for the orange target on the back wall. During this pre-training, special care was taken in presenting the target at different heights and positions in the experimental arena to avoid any position learning. After the first successful visit of the target, a training phase of 20 visits followed. The position of the target was changed in a pseudorandom pattern: on average, in 50% of the visits the target was presented in the top row, and in the other 50% in the bottom row. The target discs and feeders were exchanged with clean discs/feeders after every third visit to avoid odour contamination.
The training phase was followed by an experimental phase, which consisted of 30 flights (five blocks, six flights per block). Bees were divided in three experimental groups. Bees from all groups had to search for one orange target among two blue distractors, but the groups differed in the placement of the objects (see insets of Figure 2). Bees enter the arena through a Plexiglas tube, view the objects on the back wall and fly towards one of them. The bee's decision is counted as correct when the bee crosses the decision line (5 cm in front of the back wall) at the position of the target. The bee's flight can be observed and filmed through a small-meshed net from above and through a Plexiglas cover from the side. The number gives the angle of the target subtended on the bee's eye when viewed from the entrance of the box. B. Reflectance curves of the back wall cover (grey BG -background), the orange target (HKS 7N) and the blue distractor (HKS 49N). C. Colour hexagon showing the colour loci of the target and the distractor (colour distance of HKS 7N to background: 0.23 hexagon units; HKS 49N to background: 0.25 hexagon units; distance between the two colours: 0.27 hexagon units); calculations after Chittka (1992).

Figure 2. Search accuracy of honeybees depending on treatments.
Treatments differed in the positioning of the rewarding target and distractors which were associated with a punishment (insets: grey circle = target, black circle = distractor). In the 'dorsal group' (straight grey line, grey circle) the target was always presented in the top row, in the 'ventral group' (straight black line, black circle) it was always located in the bottom row, and in the 'mixed group' the position of the target alternated randomly between both rows (top row: dashed grey line, grey triangle; bottom row: dashed black line, black triangle). A, B. Comparison between the one row condition and the two row condition separated in trials where the target appears in the bottom row (A) and trials with the target in the top row (B). C. Comparison of the first 12 decisions of the dorsal group and the 'ventral group'. Statistics: general linear model with binomial distribution: n.s. P>0.10, * P<0.05, ** P<0.01, ***P<0.001; data points: mean ± s.e.m., N=9 individuals per group.
In the 'dorsal group', all objects were always placed in the top row; in the 'ventral group', they were only presented in the bottom row. In the 'mixed group', target and distractors could appear in the top row as well as in the bottom row. The target position was changed in a pseudo-random order assuring that the target was presented 50% of the trials in the top row and 50% of the trials in the bottom row. The distractor positions were changed randomly between the remaining five positions of the two rows. Therefore, bees of the

Data analysis and statistics
Search efficiency was measured using two parameters -error rate and decision time. This allowed to check for possible speedaccuracy trade-offs (Burns & Dyer, 2008;Chittka et al., 2003). We counted a bee's choice when it crossed a decision line, which was indicated by a grid of white twine 5 cm in front of the back wall of the experimental box ( Figure 1; Morawetz & Spaethe, 2012). We counted an error when bees crossed this decision line at a location without a target. The decision time was defined as the time the bees needed from entering the box until they crossed the decision line.
We analysed the video recording using The Observer XT Version 7 (Noldus, Wageningen, The Netherlands) and reconstructed the flight paths of the bees using SkillSpector 1.3.0. (Video4coach, Svendborg, Denmark). Statistics was calculated using the statistical package R version 2.14.1 (R Development Core Team, 2011). We used mixed linear models to test the effect of treatment group, target location and learning on the error rate and on the decision time, respectively. Furthermore, the identity of the individual bees was implemented into the model as a random factor. For analysing the error data, a binomial distribution was applied using the function lmer of the package 'lme4' (Bates et al., 2011), while for the decision time, a normal distribution was applied using the same function. Significance levels were determined by calculating a type-III ANOVA using the function Anova (package 'car ', Fox & Weisberg, 2011). To check if the bees showed a general tendency to fly upwards or downwards during the different flight stages, the flight angles were tested against 0° (horizontal flight) with a Wilcoxon test. Figures were created using Sigma Plot 11.0 (Systat Software Inc., San Jose, USA) and Corel Draw X3 (Corel Corporation, Ottawa, Canada).

Error rate
During the training phase, when only the target was present, bees of all treatment groups showed similar accuracy (χ 2 (2) =1.37, P=0.502). At the end of the training, the bees made four times fewer mistakes when searching for ventrally located targets compared to targets located in the dorsal row (bottom row: 15% ± 25% s.d.; top row: 65% ± 33%; χ 2 (1) =34.32, P<0.001; see also dataset 1). During the test phase, the overall error rate was similar in all three treatment groups (χ 2 (2) =2.79, P=0.248; see also dataset 2), but the speed of improvement differed between the three groups (interaction treatment group x learning block: χ 2 (8) =19.58, P=0.012).
Bees of the group presented with targets in the ventral visual field showed a constant performance during the test with a mean error rate of 17% ± 14 s.d. (χ 2 (4) =2.72, P=0.606; Figure 2A, straight line). Conversely, when bees were presented with targets in the dorsal visual field, they made about twice as much mistakes at the beginning of the test than at the end (χ 2 (4) =12.84, P=0.012; Figure 2B, straight line). However, the overall error rate did not differ from the 'ventral group' (χ 2 (1) =0.78, P=0.377; Figure 2A). The main change in learning performance in the 'dorsal group' took place between the first and second learning block; thus a possible difference between both treatments at the beginning of the test phase may have been masked. We then analysed the first 12 decisions in more detail, subdividing the 12 visits into blocks of three ( Figure 2C). During the first three flights, bees of the 'dorsal group' made four times more mistakes than bees of the 'ventral group' (z = 2.77, P=0.006; Figure 2C). After this initial phase, both groups achieved similar levels of accuracy (all further learning blocks: P>0.05; Figure 2C).
The 'mixed group' was the only group in which the target and distractors were presented in both the ventral and the dorsal part of the visual field. Similar to the 'dorsal group', bees improved their accuracy during the test (χ 2 (4) =9.54, P=0.049). Bees of the 'mixed group' made fewer mistakes when the target was presented in the ventral row compared to the situation when it was presented in the dorsal row (χ 2 (1) =5.27, P=0.022; Figure 2A,B dashed lines). Next, we analysed both data sets separately and compared them to the correspondent single row treatment group ( Figure 2A: flights of the 'mixed group' when the target was ventrally presented compared to the 'ventral group'; Figure 2B: flights of the 'mixed group' when target was dorsally presented compared to the 'dorsal group'). The performance of bees from the 'mixed group', when searching for a ventral target, was similar to the performance of the 'ventral group' where the target was always presented ventrally (χ 2 (1) =0.35, P=0.556; Figure 2A). Furthermore, bees of the 'mixed group', when searching for ventral targets, showed a constant performance during the experiment (χ 2 (4) =3.90, P=0.420). However, they made significantly more mistakes than bees of the 'dorsal group' in finding a dorsally positioned target (χ 2 (1) =7.69, P=0.006; Figure 2B). Nonetheless, the 'mixed group' improved their performance for dorsal targets (χ 2 (4) =14.05, P=0.007), and in the fourth learning block, bees managed to achieve a similarly low error rate as the bees of the 'dorsal group' ( Figure 2B).

Decision time
All three experimental groups became faster during the experiment (χ 2 (4) =16.83, P=0.002; Figure 3), but the decision time did not differ among treatment groups (χ 2 (2) =1.04, P=0.595) or in the interaction of both factors (χ 2 (8) =10.00, P=0.265). However, bees of the 'mixed group' were slower when the target was presented in the top row than when it was presented in the bottom row (χ 2 (1) =28.6, P<0.001). Bees of the 'dorsal' and 'ventral group', on the other hand, showed similar decision times during most of the time (χ 2 (1) =0.24, P=0.624). However, during the first three flights, bees of the 'dorsal group' were significantly slower than bees of the 'ventral group' (t (13.1) =2.21, P=0.045; Figure 3C).

Flight path
When analysing the flight structure of the first flights in detail, it became evident that bees of the 'dorsal group' increased the steepness of their departure angle (flight angle between 0 and 5 cm distance from the entrance) significantly from the first to the third flight ( Figure 4; Wilcoxon test between first and third flight: P<0.05). During their first flight, they started horizontally into the box (median flight angle 5.9°, Figure 4), but from the third flight on they flew upwards when taking off (median flight angle third flight: 12.1° upwards; fifth flight: 10.0° upwards). This change in flight  . For all four search situations the first (black arrows), the third (blue arrows) and the fifth flight (green arrows) are shown (N=7). Please note that in the mixed group the flight number refers to number of trials absolved with the particular target situation and not to the total number of trials the bee has absolved. Arrows: median flight angles; grey area: first to third quartile; statistics: one-sample Wilcoxon test against 0° (flight without significant trend of changing flight height). structure indicates that the bees decided earlier to fly upwards in their third flight compared to their first flight. All other experimental groups showed median departure angles between 7° downwards and 5° upwards -no significant trend in flying up-or downwards during the departure was found.
When the target was located in the bottom tier, the bees' flight was composed of downward angles during most of their flight path, suggesting a continuous downward flight (Figure 4, Figure 5). Bees searching for a dorsally located target, however, differed in their flight structure between treatment groups: bees of the 'mixed group' flew in a steep angle downwards during the first flight period (maximum flight angle 40° downwards) and showed a significant upwards direction between 20 and 25 cm distance from the entrance (median flight angle between 33° and 58° upwards, Figure 4). During the fifth flight of the experiment with the target in the top row, these bees lost their tendency to fly downwards at the beginning of their flight and flew straight forward for the first 20 cm and ascended steeply afterwards. Bees of the 'dorsal group' flew horizontally between the first 5 to 20 cm, with some animals already beginning to ascend.   Between 20 and 25 cm they ascended in a steep angle of around 25° similar to the animals of the 'mixed group'. Therefore, all bees searching for a top row target approached the target from below. Flight paths of individual bees demonstrate the described differences of the three different groups ( Figure 5).
In conclusion, the accuracy, decision time and flight path of bees searching for a vertically presented target was partly dependent on the target's location and partly on the bees' possibility to predict the target's location. Searching for ventrally located targets was an easy task -all bees showed high accuracy, short decision times and a direct flight path. The predictability became important in situations in which bees had to detect the target with the dorsal part of their eyes. When the target was always in the top row ('dorsal group'), bees adapted their search strategy rapidly to the situation -after three search flights they detected the target swiftly, decided accurately and approached the target directly. When the target was positioned randomly in the top or the bottom row ('mixed group'), the bees showed a more downward directed flight pattern when being confronted with dorsally located targets. As a consequence, these bees took longer in making their decisions and made more mistakes.

Discussion
We simulated three foraging situations that a bee might encounter in her natural environment: approaching flowers from above (flowers perceived by the ventral part of the eye), foraging within a 3D floral environment, either in a tree or a meadow with flowers at various heights (flowers may appear in the dorsal and ventral part of the visual field) and approaching a flower from below when foraging, for example, upwards on vertical inflorescences (flowers perceived by the dorsal area of the eye). We demonstrated that bees were able to adapt their detection strategy to all three foraging situations, although the course of performance improvement differs among the experimental conditions. In the following paragraphs, we discuss the possible proximate and ultimate mechanisms of this behavioural flexibility.

'Ventral group'
Bees of this group were presented with one target and two distractors in the bottom row. They decided accurately and swiftly and had a downwards directed flight vector. It is therefore possible that bees, by default, focus on the ventral part of the frontal visual field when being confronted with vertically presented objects, a behaviour which might explain the dorso-ventral asymmetries in visual discrimination observed by several independent studies (Baumgärtner, 1928;Chittka et al., 1988;Giurfa et al., 1999;Lehrer, 1999;Morawetz & Spaethe, 2012;Wehner, 1972).

'Dorsal group'
In the 'dorsal group', both target and distractors were presented in the top row. During the initial flights, bees' search was inaccurate and slow compared to the 'ventral group'. The bees flew straight ahead for the first 20 cm in the flight arena and then turned upwards in a steep angle during the final approach. After the first three flights, bees showed two major changes in their behaviour: (1) the error rate dropped from 40 to 20% ( Figure 2C) and (2) they changed their departure angle from horizontal to, on average, 11° upwards (Figure 4), a behaviour which was exclusively shown by the 'dorsal group'.
The reason for the bees' improvement might be explained by two independent but not exclusive mechanisms. Firstly, a top-down attentional mechanism could have accelerated information processing from the dorsal part of the frontal visual field by shifting the attentional focus to this area. In this case, bees might have detected the target stimulus just after entering the experimental arena and started an upward flight towards the top row. Alternatively, bees could have learned to adjust their flight path to the area where they expected the target: they learned to initiate an upward flight immediately after entering the arena before detecting the target. By flying upwards they were able to analyse the top row with the ventral part of their eye and therefore increase their accuracy. Evidence for both mechanisms comes from earlier studies: bees can learn to perform directional changes and changing flight routes to optimize foraging success (Cheng & Wignall, 2006;Lihoreau et al., 2010;Perry & Barron, 2013), as well as use attentional processes to selectively react to visual stimuli (Giurfa et al., 1999;Paulk et al., 2014;Spaethe et al., 2006). However, our results do not allow discriminating between both mechanisms.

'Mixed group'
Bees of the mixed group were confronted with one target and two distractors, which were randomly distributed between the bottom and top row. During most trials, objects were located in both the bottom and the top row. Hence, the search strategy of the 'mixed group' was fundamentally different from the two other groups. Bees were not simply able to fly towards a group of objects, but had to decide which group of objects was relevant for them, i.e. contained the target. At the beginning of the experiment, bees flew downwards regardless of the target position -probably executing their 'default' behavioural pattern and flying towards the lowermost objects. This behaviour resulted in a low error rate when the target showed up in the bottom row, but in a high error rate and long decision times when the target appeared in the top row. Furthermore, our data indicate that bees detected ventrally located stimuli first, but analysed the chromatic features of the stimuli only later during flight.
Bees of the mixed group had to change their flight pattern to improve the search for targets located in the top row. A downwards orientated flight route can hamper the detection of a dorsally located target due to an unfavourable angle of view for objects in the top row. Additionally, it makes complex course corrections necessary (demonstrated by the flight curves in Figure 5). Indeed, bees of the 'mixed group' stopped flying downwards in the first part of their flight after their tenth trial ( Figure 4) and instead flew straight before deciding for an upwards or downwards movement.
Additionally, bees of the mixed group may have also applied mechanisms of visual attention to increase the detection performance of targets located in the top row. For example, bees could have increased the area of their attentional focus to a size large enough to cover both rows. This would have allowed the bees to process all presented objects simultaneously (Eriksen & St. James, 1986;Pashler, 1987). In humans, an increase in the size of processing focus is normally accompanied by a decrease of processing quality (Castiello & Umiltà, 1990;Eriksen & St. James, 1986). We did not find a decrease of performance level in the search for ventrally located target in the 'mixed group', which makes this explanation unlikely. An alternative search mechanism might be that bees learn to move their attentional focus from the ventral part of the visual field to the dorsal part after they had not detected a target in the ventral part. This explanation fits with the constant high search performance for targets in the bottom row and with longer decision times and late upward directed flight vectors, when searching for a target in the top row. The high error rate in the first part of the experiment probably owes to the learning process, in which the bees (1) learned to avoid flying downwards immediately after entering the arena and (2) learned to continue searching the top row when the target could not be detected in the bottom row.

Ecological relevance
When searching for a flower patch, bees often approach the meadow or tree from above (Kevan, 1990;Lehrer, 1999) -a situation represented by the conditions of the 'ventral group'. Bees seem especially capable to deal with this type of search, as they detect ventrally positioned targets more easily (Figure 2A; Lehrer, 1999;Morawetz & Spaethe, 2012;Skorupski et al., 2006). However, when bees are foraging in a meadow with flowers at various heights, flowers can also appear in the upper part of the visual field and information received from the dorsal part of the eyes become important. For example, bees which visit raceme inflorescences tend to begin collecting nectar at the bottom of the inflorescence and ascend vertically step by step (Fisogni et al., 2011;Ishii et al., 2008;Pyke, 1978;Valtueña et al., 2013;Waddington & Heinrich, 1979). In this situation the flowers below the just probed ones, which are perceived by the ventral visual field, typically contain no nectar, because the bee had just visited them (Heinrich, 1975;Ishii et al., 2008). To optimize foraging efficiency, bees must focus visual processing on the dorsal part of the visual field or adopt an upward directed motor pattern. We demonstrated that bees can adapt to the described situation, improving their detection ability of dorsally located targets within three visits only ( Figure 2C).
When foraging in a blooming tree, a bee can expect a rewarding flower appearing in the ventral and the dorsal visual field with a similar probability. Flowers are densely distributed in a tree and bees normally see several flowers simultaneously. Interestingly, large bees as Bombus and Xylocopa reveal a complex pattern of upwards and downward movements inside a tree: when moving between neighbouring flowers, they move upwards, similar to the movement pattern on a vertical inflorescence (Kevan, 1990). However, when moving to an inflorescence at distance larger than 20 cm they tended to fly downwards (Kevan, 1990). This behaviour matches our observation that bees tend to fly to the lowermost object when approaching from a distance. However, we also demonstrated that bees can overcome this motor pattern within 10 learning trials, when it hinders the bees' foraging success. Likewise, Kevan (1990) described that movement patterns of bees differed between tree species with different flower distributions and suggested that bees optimize their foraging strategy to the particular resource distribution.
In conclusion, we showed that bees flexibly adapt to a given foraging situation by focusing their detection and discrimination effort to the appearance of the object within their visual field. They probably use both attentional mechanisms and behavioural strategies to optimize their foraging success, although more data are needed to clearly separate between these mechanisms. This flexibility provides the ability to choose among different search strategies and to quickly adapt to various foraging environments. Author contributions LM and JS conceived the study and designed the experiments. LM carried out the research and performed the data analysis. LM, LC and JS were involved in data interpretation and writing the manuscript. All authors were involved in the revision of the draft manuscript and have agreed to the final content.

Competing interests
No competing interests were disclosed. This paper analyzes visual target recognition in bees and focuses on previously reported asymmetries between the dorsal and the ventral visual fields in terms of color/pattern recognition and discrimination. The authors studied whether a pattern detection/discrimination task varies when patterns are displayed either in the ventral or in the dorsal visual fields and discuss different possible mechanisms accounting for this variation. Specifically, bees were trained to choose a rewarded orange target which was presented among blue distracters either in the ventral visual field ('ventral' group' of bees), the dorsal field ('dorsal' group') or in both fields ('mixed group'), and choice errors, decision time, and flight pathways upon entering into the experimental arena were recorded.

Grant information
The authors show in a that bees presented with targets in the ventral visual field were convincing way most efficient in detecting the rewarded target as they made rapid decisions with high accuracy and direct flight paths. In contrast, bees perceiving the patterns in the dorsal visual field made more errors and took more time to detect the pattern. Yet these 'dorsal bees' improved significantly their performance after a few learning trials as they decreased both the number of errors and decision time and flew in a straight line towards the target. Bees of the 'mixed group' needed considerably more time to improve the search of the target whose position changed randomly between the ventral and the dorsal visual field.
The conclusions drawn by the authors focus on the notion of expectation as it is said that "bees form expectations of the location of the target's appearance and adapt their search strategy accordingly". Furthermore, the authors insist on the notion of attention (more specifically, spatial attention), which is certainly attractive in the light of the results obtained, but as the authors themselves acknowledge, is . probably not the unique explanation for their data As a consequence, a more suitable title for the paper, given the results presented, would be something like: "Honey bees exhibit flexible (plastic) visual search strategies for targets presented ". It seems to me that the main finding of the paper is precisely the flexibility vertically at various heights and adaptability of the bees' visual search strategies to the various experimental situations rather than the proof of attentional processes guiding these strategies.
A raised by this paper relates to the criteria (unclear to me) used to define the position of main concern the so-called in the setup. Clearly such a line was very close (5 cm) to the targets so that "decision line" decision counting may have under/over estimated real decisions performed (and perhaps) modified at further distances from the targets. It would be therefore interesting to contrast the decisions counted and further distances from the targets. It would be therefore interesting to contrast the decisions counted and established following the authors' method with a different one, which, for instance, may be based on quantifying angular deviations of trajectories along the flight path; one can establish a criterion for a decision based on an angular deviation from the entrance vector and quantify decision which may occur before reaching the decision line defined by the authors.
A relates to incomplete citations in the paper. In particular, given that the focus is on minor concern ventral/dorsal asymmetry of visual performances on bees, it would be worth considering and mentioning the work on ventral vs. dorsal target detection in bees ( ) . Additionally, the statement on Giurfa ., 1999 et al the relative recency of studies on attention in insects is questionable. Although it is not explicitly saidprobably because of constraints imposed by reviewers -the work by van Swinderen and Greenspan constitutes a study on visual attention. Also, if one focuses on bees, the use of this concept of (2003) attention dates to 2004 ( ) when it was argued that discrimination between perceptually closer Giurfa, 2004 stimuli was possible after differential conditioning but not after absolute conditioning due to differences in attention inculcated by these two training procedures. The idea was further developed in later papers, so that referring the use of this concept in insects to 2011 seems a bit unfair to prior contributions which also focused on the same idea.
All in all, the manuscript is sound and well-written. The results are quite convincing (but see remark on the appropriateness of the decision line criterion) and constitute an important contribution to field of insect vision.

I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
No competing interests were disclosed. The amount of information that is perceived by the peripheral visual system is thought to exceed the amount of information that can be processed by subsequent stages of the brain. Based on a host of experimental data mainly obtained with humans, both overt and covert strategies were concluded to facilitate the extraction of the behaviourally relevant information and to cope with the problem that the entire visual information cannot be processed at once. In previous studies by the same authors (or a subgroup of them) it was convincingly shown that honeybees (in contrast to bumblebees) employ mainly an overt serial strategy to search for previously learnt coloured target objects, if these are embedded between distracting cues. In analogy to similar results obtained with human observers, these conclusions were interpreted as being based on a kind of serial shift of the animals' focus of attention. The analyses were based on the search performance (i.e. measuring both error rate and search time) while detecting a target object embedded in a variable number of distracting differently coloured objects.

Martin Egelhaaf
The present study is based on a similar experimental design as employed in these previous studies and asks the highly interesting question of whether the visual system of honeybees is functionally regionalised The present study is based on a similar experimental design as employed in these previous studies and asks the highly interesting question of whether the visual system of honeybees is functionally regionalised with regard to a pattern detection/discrimination task. Honeybees were trained to detect a coloured target either in the upper or bottom row of a 3x3 matrix of possible object positions. In the test phase distracting objects were added in three different constellations to the stimulus matrix. Both target and distractor objects were placed either in the top row ('dorsal group') or only in the bottom row ('ventral group'); in a 'mixed group', target and distractors could appear in the top row as well as in the bottom row. From the way the search performance improved under the different stimulus constellations, the authors concluded that honeybees flexibly adapt to a given foraging situation by focusing their search strategy to the location of the object within their visual field. The accuracy, decision time and flight paths of honeybees were concluded to depend on both the target's location in the visual field and on the bees' possibility to predict the target location. It is discussed whether honeybees use attentional mechanisms to increase their success rate in the detection task.
Although I am sympathetic with the overall goal of this project and agree with the authors' overall conclusion, I have to admit that I am somewhat confused by parts of the experimental design of the study. Some further explanation of the rationale behind the analysis is likely to largely raise the potential of the study.

Major issues
After the training phase the bees could not have any expectation with regard to target position, a priori since target position was pseudo-randomized during the training procedure. Interestingly, a clear dorso-ventral asymmetry of the error rate of target detection was found even at the end of the training period. For the test phase the bees were subdivided into the three experimental groups that were tested with just one of the stimulus constellations. The rationale behind this design is not well explained and not immediately obvious. At least at first sight, one might have expected that the different constellations were tested again with all bees in a pseudo-random fashion.
One reason that this has not been done might have been the extreme dorso-ventral performance gradient after the training phase. Even without any distractor the coloured target could be detected with only a very high median error rate of 65% when presented in the top row. Surprisingly, no strong point is made that the detection performance is not much above chance level (this issue should have been tested statistically!), when the target was presented in the top row. Theoretically, this poor performance should not increase, if in the test phase -i.e. after adding two distractor objects -the target were also presented randomly in the top and bottom row of the stimulus arrangement. Hence, it might have been scientifically meaningful to ask whether the detection performance in the dorsal part of the visual field might increase, if the bee had the chance to build up an expectation that the target is not located ventrally. This might be a rationale for subdividing the bees into the three groups. Irrespective of whether or not this supposition is correct, an explanation for the experimental design should be given.
In any case, it can be concluded from the development of error rates, decision times, but also the trajectory data that the performance of the 'dorsal group' improves over time and that this finding is consistent with the interpretation that also dorsally located targets can be readily detected, if the bee has knowledge that the target is likely to be found in an otherwise 'suboptimal' part of the visual field. a priori Again, it might be helpful if this kind of conclusion were drawn explicitly in the paper. Whether this finding is just a consequence of a learning process or might be also linked to attentional mechanisms can then be discussed. However, it might be hard to conceptualise how attentional mechanisms can be pinpointed experimentally in the type of experiments that were performed in this study.
The hardest issue that remains to be explained are the data obtained with the 'mixed group' of bees.

F1000Research
The hardest issue that remains to be explained are the data obtained with the 'mixed group' of bees. Although here no expectation can develop with regard of the location of the target, the performance improves over stimulus presentations to values that are much better than those obtained at the end of the training phase -even without any distractors (!). This highly surprising issue definitely needs to be discussed. This somehow apparent contradiction raises also the question whether -given a tremendous variability between bees but also between different flights of individual bees -the data base is sufficiently large to allow for drawing strong conclusions.

Minor issues
Title of the paper: Since the paper does not focus on the topic of attention, I would omit in the title of the paper the addendum ": a role for spatial attention?" Attention is just one possibility, in addition to several others, to explain the experimental results. Putting too much emphasis on attention already in the title may be somewhat misleading. : Abstract "Here we investigate if this asymmetry is caused by fixed search patterns or if bees can use alternative search mechanisms such as spatial attention, which allows flexible focusing on different areas of the visual field." I do not see that the investigation of this alternative has been the focus of the experimental analysis of the study. At least the results do not allow answering this question in an unambiguous way.

Introduction:
: It appears to be surprising that the paper starts by considering whether the distinct spatial 1 paragraph asymmetry in colour and pattern learning of honeybees as observed in previous studies might be a consequence of specialisations in eye morphology, as has been proposed for detecting the queen by the drone bee eye. It should be noted that this sort of detection task is conceptually very different: in the drone bee case we are dealing with detection thresholds where -for obvious reasons -spatial acuity and sensitivity, i.e. anatomical and physiological retinal parameters, are the relevant factors, because they determine at which distance a small target (i.e. in this case the queen) can be seen. In the learning experiments the targets were usually sufficiently large, and no spatial resolution or sensitivity problem existed for them to be detected.

Methods:
What was the frame rate and resolution of the camera system? "The stimuli subtended a visual angle of …" (and following sentence): The authors probably mean the individual discs. "stimuli" not specific enough. It should be explained why a virtual decision line was chosen. It might have been more intuitive to monitor the landing on the different targets. Would the error rate change, if the landings were recorded rather than the crossing of a virtual decision line? What happened to the bees after the individual training and test flights? Do they return to the entrance hole of the flight arena or were they caught and 'manually' bought back to their hive? Were the individual bees labelled and could be identified? This might have been a precondition to be able to monitor changes in detection performance. What about the orientation of body and head in 3D space? How well can these parameters be resolved? This is important information, especially if the retinal position of the objects needs to be assessed. The term 'flight angle' should be explained. "We used mixed linear models to test the effect of treatment group, target location and learning on the error rate and on the decision time, respectively. Furthermore, the identity of the individual bees was implemented into the model as a random factor." It is not entirely clear what these sentences st was implemented into the model as a random factor." It is not entirely clear what these sentences really mean. 'Mixed linear models' and 'implementing individual bees into a model as a random factor' should be explained.

Results:
What does 'overall error rate' mean? Figure 3: If I understand the explanation of the figure correctly, the data point '6' 'N° of flights' shown in A and B should be the mean of the corresponding data points '3' and '6' 'N° of flights' in C. Since in C both 'dorsal group' data points have a larger value than the corresponding data point in B, this interpretation cannot be correct -or something went wrong with the data. Please check. Legend of figure 4: "Please note that in the mixed group the flight number refers to number of trials absolved with the particular target situation and not to the total number of trials the bee has absolved". This statement is somewhat cryptic. Please explain.

Discussion:
Most issues concerning the Discussion have already been mentioned above. Here just some minor, though relevant details: During the flights towards the target bees change their height. How is this done? Do they change the pitch angle of their body and, in particular, of the head when changing height as compared to the pitch angle during horizontal flights? This is important, since it pertains to the retinal area with which the target is seen. , for instance, changes the pitch angle of its body when Drosophila changing height, whereas it keeps the head orientation in space relatively constant (by changing the head angle relative to the body). "… bees could have increased the area of their attentional focus to a size large enough to cover both rows". Obviously, by assuming such an attentional process almost every change in detection performance can be explained. However, from an epistemological point of view such a hypothesis makes only sense, if an experiment could be designed by which this hypothesis can be tested.
I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard, however I have significant reservations, as outlined above.
No competing interests were disclosed. Competing Interests: In this study Morawetz, Chittka, and Spaethe present results from a behavioral paradigm for honeybees where they examine spatial attention. Bees were trained to select a color, using classical conditioning with sugar versus quinine. Trained bees were then presented with various scenarios whereby the targets were presented together with distracters, either in the upper or lower visual field. The authors find that targets in the ventral field led to a more efficient search strategy. Importantly, it appears that honeybees form expectations about where their target will be, in repeated flights.

Bruno van Swinderen
These are interesting results, which help consolidate the view that insects display a selective attention. In particular, the data supporting the view that there is an expectation already formed early in the flight angle is very intriguing. It is nice to see a paper where the behavior itself is examined closely, rather than only the proportion of choice outcomes. This helps to better show how attention might be operating in these simple animals.
The paper is close to indexation quality. There are a couple questions that might need to be attended to, to further improve the quality of the work.
First, it is quite unclear how the position of the "decision line" was arrived at. While it is evidently depending on previously published work, one concern is that varying the position of this decision line might substantially change the results and conclusions. The decision line is currently 5cm away from the end wall, and both metrics (error rate and decision time) depend on this distance. Clearly, a bee flying downwards is exploiting gravity to get to the decision line, while a bee flying upwards is fighting gravity. This simple factor could account for the different flight dynamics seen for either situation (as is evident in the informative single examples shown in Figure 5). Can the authors exclude a non-trivial explanation here for the effects of gravity? With regard to the error rate for example, it is not clear whether a "mistake" is meaningful here. Maybe bees fly more erratically upwards (they do). More generally, it would be good to have a bit more of a rationale for why this 5cm point was chosen. How different would the results have been at 7cm or 10cm?
Second, the error rate metric does not seem to account for proximity. Thus, a bee crossing the decision line close to the target is as equally wrong as a bee crossing a whole square away? This seems strange. Should the error rate be weighted in some way? How might that change the results? Minor: Sareen et al (2011) showed that fruit fly attention is mostly ventral. This might be discussed a bit more, in connection to the current findings in the bee.
I have read this submission. I believe that I have an appropriate level of expertise to confirm that it is of an acceptable scientific standard.
No competing interests were disclosed. Competing Interests: