Nest excavators’ learning walks in the Australian desert ant Melophorus bagoti

Nest excavators’ learning walks in the Australian desert ant Melophorus bagoti | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Nest excavators’ learning walks in the Australian desert ant Melophorus bagoti Sudhakar Deeti, Donald James McLean, Ken Cheng This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-3908727/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 23 May, 2024 Read the published version in Animal Cognition → Version 1 posted 7 You are reading this latest preprint version Abstract The Australian red honey ant, Melophorus bagoti , stands out as the most thermophilic ant in Australia, engaging in all outdoor activities during the hottest periods of the day during summer months. This species of desert ants often navigates by means of path integration and learning landmark cues around the nest. In our study, we observed the outdoor activities of M. bagoti workers engaged in nest excavation, the maintenance of the nest structure, primarily by taking excess sand out of the nest. Before undertaking nest excavation, the ants conducted a single exploratory walk. Following their initial learning expedition, these ants then engaged in nest excavation activities. Consistent with previous findings on pre-foraging learning walks, after just one learning walk, the desert ants in our study demonstrated the ability to return home from locations 2 metres away from the nest, although not from locations 4 metres away. These findings indicate that even for activities like dumping excavated sand within a range of 5–10 cm outside the nest, these ants learn and utilize the visual landmark panorama around the nest. Excavation Navigation Red honey ant Exploratory walks Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 INTRODUCTION The ecological success of social insects, notably ants, is frequently attributed to the division of labour (Hölldobler and Wilson 1990). Individuals in a colony specialize in specific tasks, such as caring for the brood, foraging, constructing nests, or defending the colony (Wilson 1987). Each task is carried out by a distinct subset of the worker population. Specialization is found even when all the workers look similar. Colonies of the Australian desert red honey ant, Melophorus bagoti , exhibit monomorphic workers, with a similar physical appearance across workers (Deeti and Cheng 2021a). Despite this uniformity, there exists specialization in tasks and roles among these monomorphic workers. Our personal observations suggest that red honey ant workers engage in three distinct outdoor activities. Firstly, nest excavation is a continuous activity that persists from the colony’s post-hibernation phase until it resumes hibernation over the winter. This activity involves the construction and maintenance of the nest structure, including taking excess sand out of the nest. Secondly, foraging activity is performed by a subset of workers. These individuals scavenge for dead arthropods and seeds in the desert terrain, transporting their findings back to the nest (Muser et al. 2005; Schultheiss and Nooten 2013). Lastly, dumping activity involves certain workers disposing of food waste outside the nest before returning to the colony (Deeti et al. 2023a). Aside from foraging behaviour, however, the other two activities, namely nest excavation and dumping, remain understudied in this desert ant. A separate work examines the paths of dumpers, both experienced and naive. Our focus here is on comprehending the navigational understanding and learning of excavation workers. Desert ants have been extensively studied for their remarkable navigational abilities, particularly within the genera Cataglyphis , Ocymyrmex , and Melophorus (Wehner 2020). These ants, known for their visual navigation skills during foraging periods outside the nest, possess a diverse navigational toolkit. This toolkit encompasses path integration, enabling ants to continuously track the direction and distance to their starting position (Collett and Collett 2000; Wehner and Srinivasan 2003). This ability enables an animal to turn and orient itself towards a starting point, such as its nest, without relying on terrestrial visual cues or knowledge of the nest’s surroundings. Ants, however, also learn the surrounding terrestrial visual cues for navigation. Characteristic ‘learning walks’ performed before heading off on foraging journeys are thought to facilitate the learning of visual cues (Knaden and Graham 2016; Wehner 2003; Wehner et al. 2004). The toolkit also includes systematic search strategies (Schultheiss et al. 2015; Wehner and Srinivasan 1981) and backtracking capabilities (Wystrach et al. 2013). The integration of these sophisticated navigational tools highlights the flexibility of desert ants in navigating their environments (Hoinville and Wehner 2018; Wehner et al. 2016). Some parts of the toolkit require some learning to set up; this is especially so for view-based navigation. Venturing beyond the nest’s entrance exposes insects such as ants to a higher risk of predation and getting lost. To mitigate the risk of getting lost, foragers engage in 3–7 learning walks around their nest before heading off on any extended journey (Deeti and Cheng 2021b; Fleischmann et al. 2016; Jayatilaka et al. 2018). These pre-foraging walks consist of loops of increasing size, covering a larger area with each successive walk. This process plays a vital role in enabling ants to familiarize themselves with the visual landscape surrounding their nest. View-based models of ant navigation propose the existence of a catchment area around the nest, where views acquired during learning walks guide foragers back to the nest (Zeil 2012; Zeil et al. 2003; Zeil et al. 2014). Studies on Cataglyphis noda and Melophorus bagoti reveal their ability to generalize views to non-visited locations up to 10 metres away after learning the nest panorama in a limited area around their nest entrance (Deeti et al. 2020; Fleischmann et al. 2018; Wystrach et al. 2012). Previous observations of M. bagoti also revealed that, after a single initial learning walk within a range of ~ 20 cm, they could generalize their views to find their way back to the nest from a distance of 2 metres but not 4 metres (Deeti and Cheng 2021b). Earlier observations on M. bagoti also highlighted the importance of learning walks during the first two days of outdoor life for nest finding (Muser et al. 2005). Despite these insights, additional research is crucial to fully comprehend the developmental aspects of learning walks for all worker castes in desert ants. In M. bagoti , it remains uncertain whether those desert ants engaged in daily nest excavation activities require learning walks for this task and to what extent they generalize views during this process. We did not expect excavating ants to do any learning walks because our companion study conducted earlier on dumpers (Deeti et al. 2024 preprint), which travel greater distances than excavators, showed that dumpers at one nest did not do any learning walks, and because the distance travelled by excavators is small, not more than 15 cm. To foreshadow, however—a point that will be obvious in the next paragraph and the methods—we were wrong. In our current study, we examined the learning walks of nest-excavating red honey ants, specifically focusing on their pre- and post-excavation activities in their natural habitat. We documented the structure and spatial distribution of learning walks and excavation activities conducted by red honey ants at the nest site. Moreover, we carried out displacement experiments following the ants’ initial excavation walk to evaluate their ability to effectively use the terrestrial panorama from a greater distance to the nest. METHODS During the Australian summer months from November 2023 to December 2023, we conducted a study on Melophorus bagoti Lubbock 1883 desert ants originating from a single nest situated in the vicinity of the Centre for Appropriate Technology, located 10 km south of Alice Springs, NT, Australia (23°45′28.12″S, 133°52′59.77″E). The prevalent vegetation in this semi-arid desert habitat includes buffel grass ( Pennisetum cenchroides ), a mosaic of Acacia bushes, and Eucalyptus trees. No specific ethical regulations regarding the study of ants are found in Australia, and all experimental procedures employed were entirely non-invasive. Animals The red honey ant Melophorus bagoti is a most thermophilic ant on the Australian continent (Christian and Morton 1992), engages in three outdoor activities during hot summer days, and forages by primarily scavenging deceased arthropods while collecting sugary plant exudates and seeds (Muser et al. 2005; Schultheiss and Nooten 2013). Besides foraging, some workers dedicate themselves to excavating the nest throughout the day while others dump waste materials from the nest. Foraging ants operate individually for short durations, covering distances up to 50 m from the nest, relying on path integration and terrestrial visual landmarks without utilizing any chemical trails (Cheng et al. 2009). In contrast, nest workers involved in excavation activities remove sand from the nest, depositing it within a 15-cm radius outside (Deeti and Cheng 2023a). Some of the ants bigger in size also guard the nest entrance and occasionally move around the nest. Our specific investigation required the examination of naive excavating ants. After completing tasks within the nest, on average, they forage outside the nest for 4.9 days (Muser et al., 2005). We thus classified ants emerging for the first time after 5 days as naive. To identify the naive ants, all workers emerging from the colony were uniformly painted with the same colour over a 6-day period to ensure that all experienced ants were marked (Deeti and Cheng 2021b). From the 7th day onward, any unpainted ants emerging from the nest were considered as newcomers with no prior experience of terrestrial landmarks around their nest. These ants were captured at the nest entrance immediately upon appearance and marked with an individually distinct colour code on the abdomen or thorax of the ant (Tamiya™) (Deeti and Cheng 2021a). We painted 40 naive ants, and following the marking process, naive ants were released back into the nest entrance. To prevent them from re-emerging, the nest entrance was securely sealed with a lid until the ant went into the nest (about 20 s) (Deeti and Cheng 2021b). Experimental procedure The vegetation surrounding the chosen nest was cleared on the initial day of the study. To enhance the visibility of the red ants against the sandy red soil background, fine white sand was spread within the recording area. This ensured a clear contrast between the ants and the background, facilitating easy distinction in the recorded footage. Out of the total 40 painted naive ants, the study focused on the 20 ants that worked on nest excavation after their initial learning walk, and displacement experiments were conducted on these same 20 marked individuals known to be naive excavating ants. As it turned out, upon their first reappearance, these newly marked (painted) naive ants that went on to excavate always exhibited a small loop around the nest. These naive walks were recorded with video a camera (Sony Handy camera (FDR-AX700), recording at a frame rate of 25 frames per second. A tripod was set at a height of 1.2 m from the ground, with the camera looking straight down at the ground, and the recording area measured 1 m by 1 m centred at the nest, and the camera boasted a resolution of 3860 by 2160 pixels. After this initial walk, in their next appearance, when, as it always turned out, they deposited sand outside and returned to the nest, they were captured before entering and displaced to four test locations at 2 or 4 m in distance, positioned to the North and East of the nest. These directions were chosen because they possessed distinctive panoramas compared to the other cardinal directions. Each ant underwent testing at each of these four locations (2mN, 2mE, 4mN, 4mE) in a random order. At the displacement site, ant trajectories were recorded using a Sony Handy camera (FDR-AX700) recording at a frame rate of 25 frames per second. Once the tests were completed, we released the test ant back into the nest. In subsequent appearances, they were observed carrying excavated sand in their mandibles and typically tossing it outside the nest within a range of 5–10 cm. We also video-recorded the excavation activity of these ants at the nest after their displacement tests. Test ants were manually captured near their nest using a wide-mouth 50-ml glass container, transported in darkness, and released at the centre of the recording area. Each ant was released one at a time at the displacement site. After leaving the recording area of the first test, the ants were captured in a test container and then released at the next site. Tracking We used the animal tracking program DLTdv8 (version 8.2.9) in MATLAB (2022B) to extract frame-by-frame coordinates of the head and thorax — specifically the tip of the head and the middle of the thorax (see Fig. S1) — for each ant in every video obtained during our displacement test recording as well as the recordings of their learning and excavation walks. These extracted frame-by-frame coordinates served as the basis for all subsequent analyses of the workers’ movements and behaviour. Data Analysis In this study, we analysed the ants’ orientation at the displacement site and path characteristics while performing the excavation. On displacement tests, the final heading direction of ants was determined by assessing the thorax–head direction vector as they exited the recording area, that is, on the last frame in which they were visible. This vector represents the orientation of each ant at that specific moment. To analyse and visualize this data, the final heading mean vector angles were divided into 24 equal wedges, each spanning a 15-degree angle. These wedges are employed in circular plotting to illustrate how directly oriented towards the nest the ants’ trajectories were at the displacement site. For analysis, we used the circular statistics package in R (version 4.2.1; R Core Team, 2020) on the exact exit headings of the ants (and not the wedge-sector headings). We analysed the maximum displacement from the nest, the duration and the area covered by the ants’ trajectory. The nest location was chosen as the origin (0, 0). These measures helped us understand how extensively the ants explored their surroundings during the learning walks and in the case of excavation walks, discern any notable differences compared to their learning walks. Maximum displacement was the thorax position at the maximum distance from the nest. We recorded the duration of each learning and excavation walk from the moment the ant left the nest until it just before they entered the nest. For the area covered during these walks, we calculated the enclosed area of the path joining the thorax positions of the ants, using the convex-hull method to encompass all the points visited by the ant. This area measurement provided a measure of the spatial extent of their exploration. To understand how quickly and directly the ants moved on each learning and excavation walk, we calculated speed of the workers along with several characteristics of the trajectory. These path characteristics were based on the positions of the thorax across frames. Speed refers to the magnitude of an ant’s velocity and was calculated as the average over the entire trajectory for each ant, excluding the stopping durations. We measure the speed at which ants change their gaze direction by observing how quickly their gaze angle changes over time. As ants walk, their gaze constantly shifts. To determine gaze direction, we draw a line from the thorax coordinates to the head coordinates. Gaze angular velocity is calculated by dividing the change in gaze angle by the change in time. This helps us understand how fast ants adjust their viewing direction. To understand path measures of the displacement-test recordings, we used three indices of straightness: path straightness, sinuosity , and \({E}_{max}^{a}\) , each of which relates to the directness of navigation towards a destination. Straightness is computed as the ratio of the straight-line distance between the release point at the displacement site and the final point in the frame before the ant moved out of the recording area to the overall length of the path. (Batschelet 1981; Deeti et al. 2023b; Islam et al. 2021; Islam et al. 2023). Straightness ranges from 0 to 1, with larger values indicating straighter paths, while smaller values indicate more curved or convoluted paths. Sinuosity is an estimate of the tortuosity in a path, calculated as \(S=2{\left[p\left(\frac{1+c}{1-c}+{b}^{2}\right)\right]}^{-0.5}\) , where \(p\) is the mean step length, \(c\) is the mean cosine of turning angles and \(b\) is the coefficient of variation of the step length. A trajectory step is the movement between the positions of the animal (thorax positions) recorded at consecutive video frames. Accordingly, step lengths are the Euclidean distances between consecutive points along a path, and turning angle refers to the change in direction between two consecutive steps. Sinuosity varies between 0 (straight) and 1 (extremely curved) (Benhamou, 2004). The maximum expected displacement of a path, \({E}_{max}^{a}=\frac{{\beta }}{1-{\beta }}\) , where \({\beta }\) is the mean cosine of turning angles, is a dimensionless value expressed as a function of number of steps, and is consistent with the intuitive meaning of straightness (Cheung et al. 2007). Larger maximum expected displacement values indicate straighter paths, hence greater displacement, while a smaller value suggests more localized or constrained movement. Paths were characterized and visualized in R (version 4.2.1; R Core Team, 2020) using the packages trajr (McLean and Skowron Volponi 2018) and Durga (Khan and McLean 2023). During learning walks, ants frequently displayed a series of stereotypical successive fixations in different head directions by rotating on the spot at one location, known as a “scanning bout” and we extracted the number of scanning bouts from each learning walk (Deeti et al. 2023c; Lionetti et al. 2023). We obtained panoramic images using Richo Theta cameras at each test location, including one directly on top of the nest. These panoramas were utilized to compute the rotational image difference function (rotIDF) between the nest panorama and each test-location panorama. The pixel differences were calculated for each 1° shift in pixels, following methodologies outlined in previous studies (Islam et al. 2022). We calculated the depth of a rotIDF as the mean discrepancy minus the minimum discrepancy. Statistical analysis We used generalized linear mixed models to investigate whether the displacement location affected the route navigation of the foragers on tests. Path characteristics were compared across the four different displacement conditions. In the model, displacement conditions were treated as an independent variable (X) that predicts the dependent variables (Y) and Ant ID was treated as a random factor. We used 4 individual models to compare the effect of the displacement location on path straightness , sinuosity , \({E}_{max}^{a}\) , and speed. Similarly, 4 other models were used to compare characteristics of learning walks and excavation activity, in speed, convex-hull area, maximum displacement, and duration. We conducted a Welch’s ANOVA (one-way) in cases of variables with significant heterogeneity of variance. The GLM was formulated using the lme4 package (version 1.1–27) and fitted using the glmer function (Bates et al., 2015). Since path straightness and sinuosity are bounded measures (0–1), we used the binomial family, whereas all of our other response variables are only bounded by zero, and so for them we applied the Gaussian family of models. For paths on tests, we employed the Tukey post hoc test (emmeans package), to perform pair-wise post hoc comparisons for each of our models. Since we tested multiple dependent variables computed from the same data sets of trajectories on tests or on walks at the nest, we adopted p = 0.01 as an alpha level to lower type 1 errors. We report all significant terms and the significance values for all key pair-wise comparisons. Statistical analyses were conducted using R (version 4.3.1). We visualized the data using summary characteristics such as median (Box plots), density and mean confidence intervals. To compare the learning walks with excavation walks, we examined the first three excavation walks together with the learning walks (one for each ant), so that this factor, the independent variable, contained four levels. The excavation walks all looked similar when we observed them, so that we limited the analysis to a small number that every ant in the study carried out. In the generalized linear-model ANOVA, we applied a priori Helmert contrasts. The learning walk was compared against all three excavation walks in the first contrast. The second contrast compared excavation walk 1 against excavation walks 2 and 3 together, and the third contrast compared excavation walks 2 vs. 3. These contrasts are independent, making up the 3 df in the numerator of the ANOVA. Our hypothesis before data analysis was that only the first Helmert contrast would reach significance. To assess the uniform distribution of headings for each test condition (P > 0.05), we performed a Rayleigh’s test. In case both distributions turned out non-uniform, we planned to compare the mean direction of the two groups using the Watson–Williams test (alpha = 0.05). If the hypothesis of a uniform distribution cannot be rejected for one or both groups, it makes no sense to run this test. Additionally, we examined if final heading orientations significantly clustered around the nest direction at 0 degrees by checking whether 0 degrees fell within the 95% confidence interval (CI) of orientations (Watson tests). V-tests were conducted, with alpha set at P = 0.05, to determine if the mean headings were notably clustered around a specified target direction. Single-sample log likelihood ratio tests were also conducted to investigate whether the heading distributions of the ants were uniform in each test condition. RESULTS After emerging from winter hibernation, in the initial days of activity in the new summer season, we observed the naive walks of nest excavators and their subsequent outbound excavating trips. All the ants in this study engaged in a distinctive pattern. They first conducted a single learning walk in the vicinity of the nest. On the same day, following the learning walk, they then actively participated in nest excavation or digging activities. During the learning walk near the nest, the ants walked in a loop in close proximity to the nest and returned to the nest (Fig. 1 A). This looping trajectory resembled the pattern observed during foragers’ first, naive learning walks (Deeti and Cheng 2021b). Subsequently, they excavated sand from inside the nest, dumping it outside by tossing the grain, much like the dumping behaviour observed in an earlier study (Deeti and Cheng 2023a). On these excavation trips, we observed the ants to come straight out, drop the sand, and then turn around and head straight back without any ado; except for tossing the sand, they did not stop, and they did not scan on any of these trips. We conducted an analysis of various path characteristics associated with the learning and three consecutive sand excavation behaviours. Ant paths covered a significantly smaller area during the excavation trip than the learning trip (Fig. 2 A). The generalized linear-model ANOVA revealed a significant difference in area between conditions with a priori Helmert contrast-1 ( F (1,76) = 4.35, p = 0.0006), between the learning walk vs all three excavation walks, but not with contrast-2 (the excavation trip 1 vs. 2 and 3) ( F (1,76) = 1.4, p = 0.14) or contrast-3 (excavation trip 2 vs. 3) ( F (1,76) = 0.06, p = 0.804) groups. In maximum displacement from the nest, the learning walks showed numerically larger values than did the excavation walks (Fig. 2 B). A Welch’s one-way ANOVA on the maximum displacement, however, revealed no significant difference between the conditions ( F (3,76) = 1.04, p = 0.06). While there was not a statistically significant difference in maximum displacement between the learning walks and excavation activity, ants spent a longer time outside during the learning walks compared to the excavation activity (Fig. 2 C). The generalized linear-model ANOVA found significant differences between the conditions in contrast-1 and contrast-2 (the learning walk vs all three excavation walks: F (1,76 ) = 4.45, p = 0.001; the excavation trip 1 vs. 2 and 3: F (1,76) = 6.01, p = 0.004) but not in contrast-3, (excavation trip 2 vs. 3: F (1,76) = 0.06, p = 0.84. During the learning walks, ants stayed longer duration outside than the excavation trips. A Welch’s one-way ANOVA on duration uncovered a significant difference between the conditions (F (1, 35 ) = 17.2, p = 0.00001). The generalized linear-model ANOVA found significant differences between the conditions in contrast-1 ( F (1,76 ) = 33.18, p = 0.00001), but not in contrast-2 (the excavation trip 1 vs. 2 and 3: F (1,76) = 1.96, p = 0.056) and in contrast-3 ( excavation trip 2 vs. 3: F (1,76) = 1.06, p = 0.3). Thus, the ants moved similarly far from the nest on learning walks and excavation trips, but learning walks were longer in duration and covered more area. We found a notable difference in mean speed between the excavation activity and the naive learning walk. The mean speed during the learning walk was significantly lower than the speed during the excavation activity (Fig. 3 A). The a priori Helmert contrast-1 revealed a significant difference ( F (1, 76) = 30.33, p = 0.0005). Other Helmert contrasts, however, revealed no significant differences between the conditions in contrast-2 ( F (1,76 ) = 3.7, p = 0.012) or contrast-3 ( F (1,76) = 1.01, p = 0.32). In the angular velocity of gaze-direction change, during the learning walks the ants showed a numerically lower angular velocity than during the excavation trips (Fig. 3 B). This difference resulted in a statistically significant difference between conditions in a priori Helmert contrast-1 ( F (1, 76) = 4.52, p = 0.005), with no significant differences in contrast-2 ( F (1, 76) = 0.43, p = 0.64) or contrast-3 ( F (1, 76) = 0.33, p = 0.56). During the learning walks ants showed scanning behaviour. Naive learners performed a minimum one scanning bout and a maximum of 3 bouts during the learning walks, but no ant scanned on any of the excavation trips (Fig. 3 C), making inferential statistics on this variable both inappropriate and moot. In order to understand the navigational knowledge of these excavators, after their first excavation trip, each ant was captured just before it entered the nest and displaced to four different locations to the North and East of their nest entrance. The ants’ final headings at each location showed that the majority of the ants in the 2 m displacements were oriented towards the nest direction of 0 deg. In contrast, ants that were displaced 4 m from the nest were not oriented towards the nest direction from the release point (Fig. 4 ). By the Rayleigh test, the foragers’ initial orientations were non-uniformly distributed in 2-m displacement tests whereas in 4-m tests they were scattered and uniformly distributed (Table 1 , Fig. 4 ). In addition, ants in 2-m displacement conditions showed significant V-test results in the nest direction, and the means of their 95% confidence interval of initial heading values include the nest direction 0 deg (Watson test, p > 0.05). The log likelihood ratio test for the 2-m tests failed to reject the hypothesis that the distribution was clustered in the home direction ( p ≥ 0 .05, k ≥ 0, χ2 ≤ 1). However, for the 4-m test, the log likelihood ratio rejected the hypothesis that the mean value of distribution was equal to the predicted value (home direction) ( p = 0.003, k = 0.56 and χ2 = 4.6), meaning that the headings were oriented in a different direction from the nest direction. Table 1 Statistical results for initial heading directions in 2m North, 2m East, 4m North and 4m East displacement tests. Mean vector 95% confidence interval Rayleigh test V test detection 0 0 Test µ Minus Plus Z P Z P 2m North 0.34° 335.67° 5.01° 7.96 < 0.001 3.91 < 0.005 2m East 11.95° 344.05° 39.85° 6.60 < 0.001 3.55 < 0.002 4m North 289.77° 232.45° 317° 1.83 0.16 0.64 0.26 4m East 260.11° 221.71° 298.51° 3.84 0.01 -0.47 0.68 We utilized rotIDF to assess the visual similarity of each displacement location to the nest panorama. Comparisons with the nest panorama revealed detectable minima for all displacement locations, with specific values 2m North location = 27.88, 4m North location = 35.73, 2m East, location = 33. 61, 4m East = 30.49 (Fig. 4 ). Interestingly, the depth of these minima showed only slight variation across displacement locations: 2m North location = 17.49, 4m North location = 19.14, 2m East, location = 12.48, 4m East = 16.04 (Fig. 5 ). This suggests that all test locations had a ‘best’ direction pointing roughly towards the nest. We checked whether ants showed any differences in their path characteristics from oriented and non-oriented displacement locations. Firstly, in tortuosity ants appear to show lower sinuosity in the 4mN tests compared with the other conditions. However, the generalized linear model ANOVA showed no significant differences in sinuosity across the four displacement conditions (Z (3, 76) = 0.64, p 1.04), (Fig. 6 A). Secondly, \({\text{E}}_{\text{m}\text{a}\text{x}}^{\text{a}}\) appeared similar across conditions. The generalized linear model ANOVA showed no significant difference in \({\text{E}}_{\text{m}\text{a}\text{x}}^{\text{a}}\) across the four displacement conditions (Z (3, 76) = 1.05, p = 0.061), (Fig. 6 B). Finally, with straightness, the 4mN displacement condition appears to be lower on average than the other conditions (Fig. 6 C). The generalized linear model ANOVA showed a significant difference in straightness across the four conditions (Z (3, 76) = 12.59, p = 0.0001), but the Tukey post-hoc comparisons, however, showed no significant differences between any pair of displacement conditions (Table 2 ). On the whole then, path characteristics at different locations were similar, or differences were at most idiosyncratic, as we did not find any pairwise significant contrasts in any variable. During the displacement test we had not found any noticeable impact on the speed and angular deviation of the ants (Fig. 7 A and 7 B). The generalized linear-model ANOVA showed no significant difference between the displacement locations in mean speed (Z (3, 76) =–0.52, p = 1.61) and in the angular velocity of gaze-direction change (Z (3, 76) = 0.8, p = 1.14). Table 2 Tukey post hoc comparisons of statistical results for straightness between the displacement locations (alpha = 0.01). Straightness Condition ratio SE Z ratio p value east2m / east4m 1.228 0.576 0.437 0.9721 east2m / north2m 1.078 0.489 0.167 0.9984 east2m / north4m 1.802 0.948 3.303 0.0078 east4m / north2m 0.878 0.419 -0.272 0.993 east4m / north4m 1.468 0.803 0.702 0.8965 north2m / north4m 1.671 0.891 2.764 0.0308 DISCUSSION In summary, our observations of nest excavators’ behaviours in Melophorus bagoti showed that before becoming a nest excavator, workers made a single learning walk around the nest. This first walk of naive excavating ants was close to the nest, short in duration, and covered only a small area, similar to the first learning walk of foragers. On the next trips, ants started excavating the nest, carrying excavated sand in their mandibles within 5–10 cm around the nest and tossing the grain in a stereotypical way before returning straight to the nest. This activity lasted a shorter time and covered a smaller area than did their learning walks, and the ants moved faster and oscillated their gaze directions faster than they did on their learning walks. Other work that we are still analysing suggest that gaze swings are common when ants navigate homebound or outbound. The faster gaze-diretion swings on excavation walks may be coupled with the faster speed of walking. After the first excavation trip, when ants were displaced 2 m and 4 m North and East from their nest, they could orient towards the nest direction from 2 m North and East but not from 4 m away, suggesting that a catchment area between 2 and 4 m had been learned in one walk in our setting, similar to results after the first, naive learning walks of foragers. These latter ants could also orient nestward from 2 m but not from 4 m (Deeti and Cheng 2021b). Despite these differences in orientational performance, path characteristics at all the test sites were on the whole similar. As discussed further below, this similarity perhaps reflects the similarity across all the locations in their rotational image difference functions when compared to the panoramic image at the nest. In summary, ants perform a learning walk before they start excavating the nest and they generalize the views from that single learning trip to locations 2 m away from the nest. The phenomenon of learning walks and learning flights as a precursor to the transition to foraging is well established across various hymenopteran species, including honeybees (Becker 1958; Capaldi and Dyer 1999; Capaldi et al. 2000; Lehrer 1991, 1993; Vollbehr 1975), bumblebees (Hempel de Ibarra et al. 2009), wasps (Stürzl et al. 2016; Zeil 1993a, 1993b), wood ants (Judd and Collett 1998; Nicholson et al. 1999), and desert ants (Deeti and Cheng 2021a; Fleischmann et al. 2016; Müller and Wehner 2010). Through multiple flights or walks around their nest, these insects engage in exploratory behaviour, progressively covering more distance and exploring a wider range of directions to enhance their spatial knowledge around the nest (Capaldi et al. 2000; Deeti and Cheng 2021b; Fleischmann et al. 2016). It is noteworthy that this exploration often involves walks or flights in all quadrants around the nest over successive trips. We have discovered that another outdoor performer, the excavator, also engages in a single learning walk around the nest area before undertaking their work. This seems surprising because excavating ants deposit excavated sand within a 5–10 cm area around the nest. It is likely that this learning walk still serves to facilitate the return trip on such a short journey. Possible functions include calibrating the sky compass and odometer (Wehner 2020) for path integration and also learning the visual scene surrounding the nest. Using a combination of path integration and view-based homing might expedite the journey home. Testing the function of the learning walk is not easy because our observations suggest that the workers will not excavate without first taking a learning walk. At the moment, we have no way to induce them to take an excavating trip without any learning walks. The naive learning walk of red honey ants engaged in excavation activities displays parallels with the first learning walks of desert ant ( Cataglyphis , Melophorus ) foragers (Fleischmann et al. 2016; Deeti and Cheng 2021b). Comparing Melophorus bagoti would-be excavators and would-be foragers, the durations, lasting less than a minute, maximum displacement (less than 30 cm) and area covered (~ 20 cm 2 ) are similar. Would-be foragers, however, perform more than one learning walk before setting off to forage. Given that foraging takes place at a much greater distance than excavating, the multiple learning walks make functional sense. Multiple walks are needed to return from long distances. After all, a single learning walk is good for returning only from ~ 2 m away (Deeti and Cheng 2021b). However, multiple learning walks up to 1 m distance in all directions from the nest entrance allowed them to orient nestwards from 10 m distance ( Cataglyphis : Fleischmann et al. 2018; Melophorus : Wystrach et al. 2012; Deeti et al. 2020). In essence, the multiple learning walks serve to increase the catchment range for the ants. By exploring and learning the surroundings in various directions from the nest entrance, the ants expand their spatial knowledge, allowing them to navigate and return from more extended distances with precision. This adaptive strategy ensures efficient orientation and successful homeward journeys, especially in the context of foraging activities that necessitate exploration over more significant distances. Unfamiliarity is thought to trigger ‘uncertainty’ in solitary foraging ants, although how and what is encoded about uncertainty remains unclear. Experimental manipulations have included changes in the visual panorama and conflicts between path integration and view-based navigation. These alterations manifested as a reduction in the path straightness of foragers, indicating one response to navigational uncertainty induced by visual changes (Wystrach et al. 2019, 2020; Islam et al. 2021; Deeti et al. 2023b). Another change is slower travel speeds (Buehlmann et al. 2018). In the current study, path characteristics were on the whole similar across displacement-test locations, even though the ants were homeward oriented in their initial trajectories at 2 m distance but not at 4 m distance. Our image analysis of panoramic views at the test sites perhaps provides an explanation. When compared with the panoramic view at the nest, the views at all test sites differed from the nest view by similar amounts on a pixel-by-pixel basis. A pixel-by-pixel analysis is unlikely to the be sole analysis carried out by the ants’ visual system, which is known to abstract features such as the skyline where terrestrial objects meet the sky (Graham and Cheng 2009) and the fraction position of mass, the proportion of the visual scene to the left and right of the goal direction (Lent et al. 2013). Nevertheless, the pixel-by-pixel analyses serve to give proxy measures of similarity to the panoramic view at the nest. What the image analysis does not explain is why ants were homeward oriented in initial trajectories at 2 m but not at 4 m, when all the panoramic views contained a minimum-mismatch best direction pointing roughly at the nest direction. By the image analysis, the ants should have been able to orient homewards from 4 m as well as at 2 m. It is possible that other modalities of cues are used as well, such as olfactory (Buehlmann et al. 2015), vibratory (Buehlmann et al. 2012), or magnetic cues (Buehlmann et al. 2012), and these cues are less helpful at 4 m compared with 2 m, but we have no evidence for the use of these cues in this study or for this species. It remains unexplained why our ants could not orient homewards from 4 m. In conclusion, M. bagoti ants engage in learning walks in the proximity of the nest before taking on the role of a nest excavator. These learning walks extend up to 15 cm away, allowing the ants to acquire spatial knowledge about their surrounding panorama up to at least 2 m away from the nest. Future studies should delve into a more detailed exploration of the differences and similarities in the behaviour of different worker castes of naive ants, based on their specific role in the colony, such as guarding the nest, foraging, dumping and excavation activities. Additionally, there is a need for an examination of the underlying neurological processes that support learning walks in ants. This research avenue holds the potential to enhance our understanding of the intricate mechanisms governing ant behaviour and learning strategies. Declarations ACKNOWLEDGEMENTS We acknowledge the traditional custodians of the land upon which this research was conducted, the Arrernte people. Their culture and customs have nurtured and sustained this land since the Dreamtime and continue to do so today. We pay our respects to their Elders past and present. We thank the Centre for Appropriate Technology at Alice Springs, Australia for letting us work on their property and providing some storage space, and the CSIRO Arid Zone Research at Alice Springs for administrative support. We are also thankful to Vito A. G. Lionetti and Cody Freas for helping us to take panoramic images. Funding The work was supported by the Australian Research Council [DP200102337] and by the Australian Defence [AUSMURIB000001 associated with ONR MURI grant N00014-19-1-2571]. Author contributions Experimental design: SD. Data collection: SD; Data analysis: SD and DJM; Writing: SD, and KC. Ethics standards Australia has no ethical regulations regarding work with insects. The study was non-invasive and no long-term aversive effects were found on the nests or on the individuals studied. Competing interests KC is an associate editor of this journal. The authors declare no other competing or financial interests. Data availability Supplementary videos, Excel file and R scripts are available at Open Science framework: https://osf.io/gqb8r/. References Bates D, Mächler M, Bolker B, Walker S (2015) Fitting linear mixed-effects models Usinglme4. Journal of Statistical Software, 67(1). https://doi.org/10.18637/jss.v067.i01 Batschelet E (1981) Circular statistics in biology / Edward Batschelet . Academic Press. Becker L (1958) Untersuchungen über das Heimfindevermögen der Bienen. Zeitschrift für vergleichende Physiologie, 41, 1-25. Benhamou S (2004) How to reliably estimate the tortuosity of an animal's path: straightness, sinuosity, or fractal dimension? J Theor Biol , 229(2), 209-220. https://doi.org/10.1016/j.jtbi.2004.03.016 Buehlmann C, Hansson BS, Knaden M (2012) Desert ants learn vibration and magnetic landmarks. Plos one , 7(3), e33117. https://doi.org/10.1371/journal.pone.0033117 Buehlmann C, Graham P, Hansson BS, Knaden M (2015) Desert ants use olfactory scenes for navigation. Animal Behaviour , 106, 99-105. https://doi.org/10.1016/j.anbehav.2015.04.029 Buehlmann C, Fernandes ASD, Graham P (2018) The interaction of path integration and terrestrial visual cues in navigating desert ants: what can we learn from path characteristics?. Journal of Experimental Biology , 221(1), jeb167304. https://doi.org/10.1242/jeb.167304 Capaldi EA, Dyer FC (1999) The role of orientation flights on homing performance in honeybees. Journal of Experimental Biology , 202(12), 1655-1666. Capaldi EA, Smith AD, Osborne JL, Fahrbach SE, Farris SM, Reynolds DR, Edwards AS, Martin A, Robinson GE, Poppy GM (2000) Ontogeny of orientation flight in the honeybee revealed by harmonic radar. Nature , 403(6769), 537-540. https://doi.org/10.1038/35000564 Cheng K, Narendra A, Sommer S, Wehner R (2009) Traveling in clutter: navigation in the Central Australian desert ant Melophorus bagoti . Behavioural processes , 80(3), 261-268. https://doi.org/10.1016/j.beproc.2008.10.015 Cheung A, Zhang S, Stricker C, Srinivasan MV (2007) Animal navigation: the difficulty of moving in a straight line. Biological Cybernetics , 97(1), 47-61. https://doi.org/10.1007/s00422-007-0158-0 Christian KA, Morton SR (1992) Extreme thermophilia in a central Australian ant, Melophorus bagot i. Physiological Zoology 65(5), 885-905. Collett M, Collett TS (2000) How do insects use path integration for their navigation? Biological Cybernetics , 83(3), 245-259. Deeti S, FujiiK, Cheng K (2020) The effect of spatially restricted experience on extrapolating learned views in desert ants, Melophorus bagoti . Animal Cognition , 23, 1063-1070. https://doi.org/10.1007/s10071-020-01359-2 Deeti S, Cheng K (2021a) Observations on relocation of a nest by a colony of red honey ants ( Melophorus bagoti ). Northern Territory Naturalist , 30, 100-107. Deeti S, Cheng K (2021b) Learning walks in an Australian desert ant, Melophorus bagoti . J ournal of Experimental Biology, 224(16), jeb242177. https://doi.org/10.1242/jeb.242177 Deeti S, Freas C, Murray T, Cheng K (2023a) Dumping behaviour of Australian desert ants ( Melophorus bagoti )(Hymenoptera: Formicidae). Insectes sociaux , 70(2), 225-232. https://doi.org/10.1007/s00040-023-00911-w Deeti S, Islam M, Freas C, Murray T, Cheng K (2023b) Intricacies of running a route without success in night-active bull ants ( Myrmecia midas ). Journal of Experimental Psychology: Animal Learning and Cognition , 49(2), 111. https://doi.org/10.1037/xan0000350 Deeti S, Cheng K, Graham P Wystrach A, (2023c) Scanning behaviour in ants: an interplay between random-rate processes and oscillators. Journal of Comparative Physiology A 209, 625–639. https://doi.org/10.1007/s00359-023-01628-8 Deeti S, McLean DJ, Cheng K (2024) Desert ants (Melophorus bagoti) learn from experience to improve waste dumping, 22 January 2024, PREPRINT (Version 1) available at Research Square [https://doi.org/10.21203/rs.3.rs-3872453/v1] Fleischmann PN, Christian M, Müller VL, Rössler W, Wehner R (2016) Ontogeny of learning walks and the acquisition of landmark information in desert ants, Cataglyphis fortis. Journal of Experimental Biology , 219(19), 3137-3145. https://doi.org/10.1242/jeb.140459 Fleischmann PN, Rössler W, Wehner R (2018) Early foraging life: spatial and temporal aspects of landmark learning in the ant Cataglyphis noda . Journal of Comparative Physiology A , 204(6), 579-592. https://doi.org/10.1007/s00359-018-1260-6 Hölldobler B, Wilson EO (1990) The ants . Harvard University Press. Graham P, Cheng K (2009) Ants use the panoramic skyline as a visual cue during navigation. Current Biology , 19(20), R935-R937. https://doi.org/10.1016/j.cub.2009.08.015 Islam M, Deeti S, Kamhi JF, Cheng K (2021) Minding the gap: learning and visual scanning behaviour in nocturnal bull ants. Journal of Experimental Biology , 224(14). https://doi.org/10.1242/jeb.242245 Islam M, Deeti S, Murray T, Cheng K (2022) What view information is most important in the homeward navigation of an Australian bull ant, Myrmecia midas ? Journal of Comparative Physiology A 208(5), 545-559. https://doi.org/10.1007/s00359-022-01565-y Islam M, Deeti S, Mahmudah Z, Kamhi JF, Cheng K (2023) Detouring while foraging up a tree: What bull ants (Myrmecia midas) learn and their reactions to novel sensory cues. Journal of Comparative Psychology , 137(1), 4. https://doi.org/10.1037/com0000333 Jayatilaka P, Murray T, Narendra A, Zeil J (2018) The choreography of learning walks in the Australian jack jumper ant Myrmecia croslandi . Journal of Experimental Biology , 221(20). https://doi.org/10.1242/jeb.185306 Judd S, Collett T (1998) Multiple stored views and landmark guidance in ants. Nature , 392 (6677), 710-714. Khan MK, McLean DJ (2023) Durga: An R package for effect size estimation and visualisation. bioRxiv , 2023.2002.2006.526960. https://doi.org/10.1101/2023.02.06.526960 Knaden M, Graham P (2016) The sensory ecology of ant navigation: from natural environments to neural mechanisms. Annual review of entomology , 61, 63-76. https://doi.org/10.1146/annurev-ento-010715-023703 Lehrer M (1991) Bees which turn back and look. Naturwissenschaften , 78(6), 274-276. Lehrer M (1993) Why do bees turn back and look? Journal of Comparative Physiology A , 172, 549-563. Lent D, Graham P, Collett TS (2013) Visual scene perception in navigating wood ants. Current Biology , 23(8):684–690. https://doi.org/10.1016/j.cub.2013.03.016 Lionetti VAG, Deeti S, Murray T, Cheng K (2023) Resolving conflict between aversive and appetitive learning of views: how ants shift to a new route during navigation. Learning and Behaviour , 51, 446–457 (2023). https://doi.org/10.3758/s13420-023-00595-z. McLean DJ, Skowron Volponi MA (2018) trajr: An R package for characterisation of animal trajectories. Ethology , 124(6), 440-448. https://doi.org/https://doi.org/10.1111/eth.12739 Müller M, Wehner R (2010) Path integration provides a scaffold for landmark learning in desert ants. Current Biology , 20(15), 1368-1371. https://doi.org/10.1016/j.cub.2010.06.035 Muser B, Sommer S, Wolf H, Wehner R (2005) Foraging ecology of the thermophilic Australian desert ant, Melophorus bagoti . Australian Journal of Zoology , 53(5), 301-311. https://doi.org/10.1071/ZO05023 Nicholson D, Judd S, Cartwright B, Collett T (1999) Learning walks and landmark guidance in wood ants ( Formica rufa ). Journal of Experimental Biology , 202(13), 1831-1838. Schultheiss P, Cheng K, Reynolds AM (2015) Searching behavior in social Hymenoptera. Learning and Motivation , 50, 59-67. https://doi.org/10.1016/j.lmot.2014.11.002 Schultheiss P, Nooten SS (2013) Foraging patterns and strategies in an Australian desert ant. Austral Ecology , 38(8), 942-951. https://doi.org/10.1111/aec.12037 Stürzl W, Zeil J, Boeddeker N, Hemmi JM (2016) How wasps acquire and use views for homing. Current Biology , 26(4), 470-482. https://doi.org/10.1016/j.cub.2015.12.052 Vollbehr J (1975) ZUR ORIENTIERUNG JUNGER HONIGBIENEN BEI IHREM. I. ORIENTIERUNGSFLUG. Wehner R.(2003) Desert ant navigation: how miniature brains solve complex tasks. Journal of Comparative Physiology A , 189, 579-588. https://doi.org/10.1007/s00359-003-0431-1 Wehner R (2020) Desert navigator: the journey of an ant . Harvard University Press. Wehner R, Meier C, Zollikofer C (2004) The ontogeny of foragwehaviour in desert ants, Cataglyphis bicolor . Ecological Entomology , 29(2), 240-250. https://doi.org/10.1111/j.0307-6946.2004.00591.x Wehner R, Srinivasan MV (1981) Searching behaviour of desert ants, genus Cataglyphis (Formicidae, Hymenoptera). Journal of comparative physiology , 142, 315-338. Wehner R, Srinivasan MV (2003) Path integration in insects. The neurobiology of spatial behaviour , 9-30. Wilson EO (1987) Causes of ecological success: the case of the ants. J ournal of animal Ecology , 56(1), 1-9. Wystrach A, Graham P (2012) What can we learn from studies of insect navigation? Animal Behaviour , 84(1), 13-20. https://doi.org/10.1016/j.anbehav.2012.04.017 Wystrach A, Schwarz S, Baniel A, Cheng K (2013) Backtracking behaviour in lost ants: an additional strategy in their navigational toolkit. Proceedings of the Royal Society B: Biological Sciences , 280(1769), 20131677. https://doi.org/10.1098/rspb.2013.1677 Wystrach A, Schwarz S, Graham P, Cheng K (2019) Running paths to nowhere: repetition of routes shows how navigating ants modulate online the weights accorded to cues. Animal Cognition Vol. 22, pp. 213–222. https://doi.org/10.1007/s10071-019-01236-7 Wystrach A, Buehlmann C, Schwarz S, Cheng K, Graham P (2020) Rapid aversive and memory trace learning during route navigation in desert ants. Current biology , 30(10), 1927-1933. https://doi.org/10.1016/j.cub.2020.02.082 Zeil J (1993a) Orientation flights of solitary wasps ( Cerceris Sphecidae ; Hymenoptera) I. Description of flight. Journal of Comparative Physiology A , 172, 189-205. Zeil J (1993b) Orientation flights of solitary wasps ( Cerceris Sphecidae ; Hymenoptera) II. Similarities between orientation and return flights and the use of motion parallax. Journal of Comparative Physiology A , 172, 207-222. Zeil J (2012) Visual homing: an insect perspective. Current opinion in neurobiology, 22(2), 285-293. https://doi.org/10.1016/j.conb.2011.12.008 Zeil J, Hofmann MI, Chahl JS (2003) Catchment areas of panoramic snapshots in outdoor scenes. JOSA A , 20(3), 450-469. https://doi.org/10.1364/JOSAA.20.000450 Zeil J, Narendra A, Stürzl W (2014) Looking and homing: how displaced ants decide where to go. Philosophical Transactions of the Royal Society B: Biological Sciences , 369(1636), 20130034. https://doi.org/10.1098/rstb.2013.0034 Additional Declarations Competing interest reported. KC is an associate editor of this journal. The authors declare no other competing or financial interests. Supplementary Files floatimage8.jpeg Figure S1. The annotated ants’ two body positions, used in data analysis: Front of the head and Mid of the thorax. Cite Share Download PDF Status: Published Journal Publication published 23 May, 2024 Read the published version in Animal Cognition → Version 1 posted Editorial decision: Revision requested 09 Apr, 2024 Reviews received at journal 20 Feb, 2024 Reviewers agreed at journal 03 Feb, 2024 Reviewers invited by journal 02 Feb, 2024 Editor assigned by journal 02 Feb, 2024 Submission checks completed at journal 31 Jan, 2024 First submitted to journal 29 Jan, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-3908727","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":270219205,"identity":"5922284f-7cc0-49df-9e6b-976f2e317b9a","order_by":0,"name":"Sudhakar Deeti","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA40lEQVRIiWNgGAWjYBAC+QYGMwYGNiCLmfkAkJSQIajF4ABUCw8zWwJICw9hLQwwLQw8BiABIrRIJG978KPMRs6enefzqxs1FjwM7IePbsDrlxlp5YY959KMeZh5t1nnHAM6jCct7QZea27kmEnwth1O7AFqMc5hA2qR4DEjqEXyb9v/+h5mnmfGOf+I1CLN23YggYeZh/lxbhsRWgzOPCuTljmXbNhzmM2MObdPgoeNkF/k25O3Sb4ps5Nn7z/8+HPOtzo5fvbDx/A7DAmwSYBJYpWDAPMHUlSPglEwCkbByAEAORZBCyHSnJgAAAAASUVORK5CYII=","orcid":"","institution":"Macquarie University","correspondingAuthor":true,"prefix":"","firstName":"Sudhakar","middleName":"","lastName":"Deeti","suffix":""},{"id":270219206,"identity":"02b4bf58-818a-4b98-b2dd-fdefe2ec7031","order_by":1,"name":"Donald James McLean","email":"","orcid":"","institution":"Macquarie University","correspondingAuthor":false,"prefix":"","firstName":"Donald","middleName":"James","lastName":"McLean","suffix":""},{"id":270219207,"identity":"797e572b-19cb-4500-8d3d-ce8c7debd7f2","order_by":2,"name":"Ken Cheng","email":"","orcid":"","institution":"Macquarie University","correspondingAuthor":false,"prefix":"","firstName":"Ken","middleName":"","lastName":"Cheng","suffix":""}],"badges":[],"createdAt":"2024-01-29 10:49:26","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-3908727/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-3908727/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s10071-024-01877-3","type":"published","date":"2024-05-24T00:33:43+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":50583137,"identity":"25da069d-b640-45d0-b78d-a1a5ee7db3c5","added_by":"auto","created_at":"2024-02-02 20:25:19","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":118455,"visible":true,"origin":"","legend":"\u003cp\u003eNaive learning walk and nest excavating ants’ paths. The line graphs show the naïve learning walks of excavators (A), each ant represented by a different colour, and same ants’ first (B), second (C) and third (D) round-trips during the sand-excavation. The nest is located at (0, 0).\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-3908727/v1/62c970ec9f7e0b5279f2c037.jpeg"},{"id":50582999,"identity":"5a237966-b3b3-4ccd-b303-e2e43277037b","added_by":"auto","created_at":"2024-02-02 20:17:19","extension":"jpeg","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":253136,"visible":true,"origin":"","legend":"\u003cp\u003eCharacteristics of naive learning walks and three consecutive excavation activity of ants at the nest. A) Convex hull area, B) maximum displacement from the nest, C) duration of the walk. The boxes indicate the median and quartiles, while the whiskers show extreme values excluding outliers. Each point represents a single trajectory.\u003c/p\u003e","description":"","filename":"floatimage2.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-3908727/v1/27a316203f0816fb06200fea.jpeg"},{"id":50582996,"identity":"e96d663b-f4a1-4eb1-8992-be9391b70312","added_by":"auto","created_at":"2024-02-02 20:17:19","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":116168,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of speed, gaze angular velocity and scanning bouts of ants during the learning and three consecutive excavation trips. The violin plot shows the mean speed of ants across their entire trajectory (A). The violin or half violin plot shows the distribution of bootstrapped differences of mean gaze angular velocity of ants during the learning and consecutive excavation trips (B). Number of scanning bouts of naïve ants during the learning walk and next three consecutive excavation trips (C). \u0026nbsp;The solid dot shows mean, while the vertical bar shows 95% confidence interval of the mean.\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-3908727/v1/b80a627c0416b208e0c9a33f.png"},{"id":50583138,"identity":"b59840e2-222f-4bda-b80e-077809270b90","added_by":"auto","created_at":"2024-02-02 20:25:19","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":90304,"visible":true,"origin":"","legend":"\u003cp\u003eCircular histograms of initial headings of foragers during the displacement test on North 2m (A), East 2m (B), North 4m (C) and East 4m (D) tests. In the histograms, the nest direction is set at 0°. The arrows denote the length and direction of the mean vector.\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-3908727/v1/898938d6696fe73e073ea2dc.png"},{"id":50582997,"identity":"bc72779b-94da-4f3b-8864-b2ef380cf318","added_by":"auto","created_at":"2024-02-02 20:17:19","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":623293,"visible":true,"origin":"","legend":"\u003cp\u003eA to E). Panoramic views at various tested locations, all aligned towards the nest direction: at A) the Nest, B) 2m North, C) 4m North, D) 2m East, and E) 4m East. F). The rotation image difference function for each location when compared with the nest panorama facing the direction of the nest. Table in the inset shows the minimum, median, mean and mean depth at each test location when compared with the nest panorama facing the direction of the nest.\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-3908727/v1/bc08ae57921d7f0727082aa9.jpeg"},{"id":50582995,"identity":"2b8fbc48-90ec-46e2-8117-84ceca219023","added_by":"auto","created_at":"2024-02-02 20:17:19","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":105199,"visible":true,"origin":"","legend":"\u003cp\u003ePath characteristics of ants at different displacement locations: A) \u003cem\u003esinuosity\u003c/em\u003e, B) \u0026nbsp;\u003cem\u003eE\u003c/em\u003e\u003csup\u003e\u003cem\u003ea\u003c/em\u003e\u003c/sup\u003e\u003csub\u003e\u003cem\u003emax\u003c/em\u003e\u003c/sub\u003e and C) \u003cem\u003estraightness\u003c/em\u003e. Box plots display the median (line inside the box), interquartile range (box), and extreme values excluding outliers (whiskers). Individual data points are shown as dots.\u003c/p\u003e","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-3908727/v1/7d5bbb47006b7cf3f694f690.jpeg"},{"id":50583139,"identity":"2fd612c9-9833-423d-a383-72d0cfacbb2d","added_by":"auto","created_at":"2024-02-02 20:25:19","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":141263,"visible":true,"origin":"","legend":"\u003cp\u003eComparison of speed and gaze angular velocity of ants at different displacement locations. The violin plot shows the mean speed of ants across their entire trajectory at the displacement location (A). The violon or half violin plot shows the distribution of bootstrapped differences of mean gaze angular velocity ants at the various displacement locations (B). In (A) and (B), the solid dot shows mean, while the vertical bar shows 95% confidence interval of the mean.\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-3908727/v1/18d281fd05144526a4772a8d.png"},{"id":57115331,"identity":"040cb01b-a19f-41f1-b3e0-e89d4d3022a1","added_by":"auto","created_at":"2024-05-25 00:33:49","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1957453,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-3908727/v1/7ea95b24-773d-47c3-b01f-3175a4c9dfe1.pdf"},{"id":50582993,"identity":"1f8c2973-fb34-49ec-9254-6017cd900b87","added_by":"auto","created_at":"2024-02-02 20:17:19","extension":"jpeg","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":382558,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eFigure S1.\u003c/strong\u003e The annotated ants’ two body positions, used in data analysis: Front of the head and Mid of the thorax.\u003c/p\u003e","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-3908727/v1/9865168b00e37d86f2f3ae2b.jpeg"}],"financialInterests":"Competing interest reported. KC is an associate editor of this journal. The authors declare no other competing or financial interests.","formattedTitle":"Nest excavators’ learning walks in the Australian desert ant Melophorus bagoti","fulltext":[{"header":"INTRODUCTION","content":"\u003cp\u003eThe ecological success of social insects, notably ants, is frequently attributed to the division of labour (H\u0026ouml;lldobler and Wilson 1990). Individuals in a colony specialize in specific tasks, such as caring for the brood, foraging, constructing nests, or defending the colony (Wilson 1987). Each task is carried out by a distinct subset of the worker population. Specialization is found even when all the workers look similar. Colonies of the Australian desert red honey ant, \u003cem\u003eMelophorus bagoti\u003c/em\u003e, exhibit monomorphic workers, with a similar physical appearance across workers (Deeti and Cheng 2021a). Despite this uniformity, there exists specialization in tasks and roles among these monomorphic workers. Our personal observations suggest that red honey ant workers engage in three distinct outdoor activities. Firstly, nest excavation is a continuous activity that persists from the colony\u0026rsquo;s post-hibernation phase until it resumes hibernation over the winter. This activity involves the construction and maintenance of the nest structure, including taking excess sand out of the nest. Secondly, foraging activity is performed by a subset of workers. These individuals scavenge for dead arthropods and seeds in the desert terrain, transporting their findings back to the nest (Muser et al. 2005; Schultheiss and Nooten 2013). Lastly, dumping activity involves certain workers disposing of food waste outside the nest before returning to the colony (Deeti et al. 2023a). Aside from foraging behaviour, however, the other two activities, namely nest excavation and dumping, remain understudied in this desert ant. A separate work examines the paths of dumpers, both experienced and naive. Our focus here is on comprehending the navigational understanding and learning of excavation workers.\u003c/p\u003e \u003cp\u003eDesert ants have been extensively studied for their remarkable navigational abilities, particularly within the genera \u003cem\u003eCataglyphis\u003c/em\u003e, \u003cem\u003eOcymyrmex\u003c/em\u003e, and \u003cem\u003eMelophorus\u003c/em\u003e (Wehner 2020). These ants, known for their visual navigation skills during foraging periods outside the nest, possess a diverse navigational toolkit. This toolkit encompasses path integration, enabling ants to continuously track the direction and distance to their starting position (Collett and Collett 2000; Wehner and Srinivasan 2003). This ability enables an animal to turn and orient itself towards a starting point, such as its nest, without relying on terrestrial visual cues or knowledge of the nest\u0026rsquo;s surroundings. Ants, however, also learn the surrounding terrestrial visual cues for navigation. Characteristic \u0026lsquo;learning walks\u0026rsquo; performed before heading off on foraging journeys are thought to facilitate the learning of visual cues (Knaden and Graham 2016; Wehner 2003; Wehner et al. 2004). The toolkit also includes systematic search strategies (Schultheiss et al. 2015; Wehner and Srinivasan 1981) and backtracking capabilities (Wystrach et al. 2013). The integration of these sophisticated navigational tools highlights the flexibility of desert ants in navigating their environments (Hoinville and Wehner 2018; Wehner et al. 2016). Some parts of the toolkit require some learning to set up; this is especially so for view-based navigation.\u003c/p\u003e \u003cp\u003eVenturing beyond the nest\u0026rsquo;s entrance exposes insects such as ants to a higher risk of predation and getting lost. To mitigate the risk of getting lost, foragers engage in 3\u0026ndash;7 learning walks around their nest before heading off on any extended journey (Deeti and Cheng 2021b; Fleischmann et al. 2016; Jayatilaka et al. 2018). These pre-foraging walks consist of loops of increasing size, covering a larger area with each successive walk. This process plays a vital role in enabling ants to familiarize themselves with the visual landscape surrounding their nest. View-based models of ant navigation propose the existence of a catchment area around the nest, where views acquired during learning walks guide foragers back to the nest (Zeil 2012; Zeil et al. 2003; Zeil et al. 2014). Studies on \u003cem\u003eCataglyphis noda\u003c/em\u003e and \u003cem\u003eMelophorus bagoti\u003c/em\u003e reveal their ability to generalize views to non-visited locations up to 10 metres away after learning the nest panorama in a limited area around their nest entrance (Deeti et al. 2020; Fleischmann et al. 2018; Wystrach et al. 2012). Previous observations of \u003cem\u003eM. bagoti\u003c/em\u003e also revealed that, after a single initial learning walk within a range of ~\u0026thinsp;20 cm, they could generalize their views to find their way back to the nest from a distance of 2 metres but not 4 metres (Deeti and Cheng 2021b). Earlier observations on \u003cem\u003eM. bagoti\u003c/em\u003e also highlighted the importance of learning walks during the first two days of outdoor life for nest finding (Muser et al. 2005). Despite these insights, additional research is crucial to fully comprehend the developmental aspects of learning walks for all worker castes in desert ants. In \u003cem\u003eM. bagoti\u003c/em\u003e, it remains uncertain whether those desert ants engaged in daily nest excavation activities require learning walks for this task and to what extent they generalize views during this process. We did not expect excavating ants to do any learning walks because our companion study conducted earlier on dumpers (Deeti et al. 2024 preprint), which travel greater distances than excavators, showed that dumpers at one nest did not do any learning walks, and because the distance travelled by excavators is small, not more than 15 cm. To foreshadow, however\u0026mdash;a point that will be obvious in the next paragraph and the methods\u0026mdash;we were wrong.\u003c/p\u003e \u003cp\u003eIn our current study, we examined the learning walks of nest-excavating red honey ants, specifically focusing on their pre- and post-excavation activities in their natural habitat. We documented the structure and spatial distribution of learning walks and excavation activities conducted by red honey ants at the nest site. Moreover, we carried out displacement experiments following the ants\u0026rsquo; initial excavation walk to evaluate their ability to effectively use the terrestrial panorama from a greater distance to the nest.\u003c/p\u003e"},{"header":"METHODS","content":"\u003cp\u003eDuring the Australian summer months from November 2023 to December 2023, we conducted a study on \u003cem\u003eMelophorus bagoti\u003c/em\u003e Lubbock 1883 desert ants originating from a single nest situated in the vicinity of the Centre for Appropriate Technology, located 10 km south of Alice Springs, NT, Australia (23\u0026deg;45\u0026prime;28.12\u0026Prime;S, 133\u0026deg;52\u0026prime;59.77\u0026Prime;E). The prevalent vegetation in this semi-arid desert habitat includes buffel grass (\u003cem\u003ePennisetum cenchroides\u003c/em\u003e), a mosaic of \u003cem\u003eAcacia\u003c/em\u003e bushes, and \u003cem\u003eEucalyptus\u003c/em\u003e trees. No specific ethical regulations regarding the study of ants are found in Australia, and all experimental procedures employed were entirely non-invasive.\u003c/p\u003e \u003cdiv id=\"Sec3\" class=\"Section2\"\u003e \u003ch2\u003eAnimals\u003c/h2\u003e \u003cp\u003eThe red honey ant \u003cem\u003eMelophorus bagoti\u003c/em\u003e is a most thermophilic ant on the Australian continent (Christian and Morton 1992), engages in three outdoor activities during hot summer days, and forages by primarily scavenging deceased arthropods while collecting sugary plant exudates and seeds (Muser et al. 2005; Schultheiss and Nooten 2013). Besides foraging, some workers dedicate themselves to excavating the nest throughout the day while others dump waste materials from the nest. Foraging ants operate individually for short durations, covering distances up to 50 m from the nest, relying on path integration and terrestrial visual landmarks without utilizing any chemical trails (Cheng et al. 2009). In contrast, nest workers involved in excavation activities remove sand from the nest, depositing it within a 15-cm radius outside (Deeti and Cheng 2023a). Some of the ants bigger in size also guard the nest entrance and occasionally move around the nest.\u003c/p\u003e \u003cp\u003eOur specific investigation required the examination of naive excavating ants. After completing tasks within the nest, on average, they forage outside the nest for 4.9 days (Muser et al., 2005). We thus classified ants emerging for the first time after 5 days as naive. To identify the naive ants, all workers emerging from the colony were uniformly painted with the same colour over a 6-day period to ensure that all experienced ants were marked (Deeti and Cheng 2021b). From the 7th day onward, any unpainted ants emerging from the nest were considered as newcomers with no prior experience of terrestrial landmarks around their nest. These ants were captured at the nest entrance immediately upon appearance and marked with an individually distinct colour code on the abdomen or thorax of the ant (Tamiya\u0026trade;) (Deeti and Cheng 2021a). We painted 40 naive ants, and following the marking process, naive ants were released back into the nest entrance. To prevent them from re-emerging, the nest entrance was securely sealed with a lid until the ant went into the nest (about 20 s) (Deeti and Cheng 2021b).\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec4\" class=\"Section2\"\u003e \u003ch2\u003eExperimental procedure\u003c/h2\u003e \u003cp\u003eThe vegetation surrounding the chosen nest was cleared on the initial day of the study. To enhance the visibility of the red ants against the sandy red soil background, fine white sand was spread within the recording area. This ensured a clear contrast between the ants and the background, facilitating easy distinction in the recorded footage. Out of the total 40 painted naive ants, the study focused on the 20 ants that worked on nest excavation after their initial learning walk, and displacement experiments were conducted on these same 20 marked individuals known to be naive excavating ants. As it turned out, upon their first reappearance, these newly marked (painted) naive ants that went on to excavate always exhibited a small loop around the nest. These naive walks were recorded with video a camera (Sony Handy camera (FDR-AX700), recording at a frame rate of 25 frames per second. A tripod was set at a height of 1.2 m from the ground, with the camera looking straight down at the ground, and the recording area measured 1 m by 1 m centred at the nest, and the camera boasted a resolution of 3860 by 2160 pixels. After this initial walk, in their next appearance, when, as it always turned out, they deposited sand outside and returned to the nest, they were captured before entering and displaced to four test locations at 2 or 4 m in distance, positioned to the North and East of the nest. These directions were chosen because they possessed distinctive panoramas compared to the other cardinal directions. Each ant underwent testing at each of these four locations (2mN, 2mE, 4mN, 4mE) in a random order. At the displacement site, ant trajectories were recorded using a Sony Handy camera (FDR-AX700) recording at a frame rate of 25 frames per second. Once the tests were completed, we released the test ant back into the nest. In subsequent appearances, they were observed carrying excavated sand in their mandibles and typically tossing it outside the nest within a range of 5\u0026ndash;10 cm. We also video-recorded the excavation activity of these ants at the nest after their displacement tests.\u003c/p\u003e \u003cp\u003eTest ants were manually captured near their nest using a wide-mouth 50-ml glass container, transported in darkness, and released at the centre of the recording area. Each ant was released one at a time at the displacement site. After leaving the recording area of the first test, the ants were captured in a test container and then released at the next site.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec5\" class=\"Section2\"\u003e \u003ch2\u003eTracking\u003c/h2\u003e \u003cp\u003eWe used the animal tracking program DLTdv8 (version 8.2.9) in MATLAB (2022B) to extract frame-by-frame coordinates of the head and thorax \u0026mdash; specifically the tip of the head and the middle of the thorax (see Fig. S1) \u0026mdash; for each ant in every video obtained during our displacement test recording as well as the recordings of their learning and excavation walks. These extracted frame-by-frame coordinates served as the basis for all subsequent analyses of the workers\u0026rsquo; movements and behaviour.\u003c/p\u003e \u003cp\u003e \u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec6\" class=\"Section2\"\u003e \u003ch2\u003eData Analysis\u003c/h2\u003e \u003cp\u003eIn this study, we analysed the ants\u0026rsquo; orientation at the displacement site and path characteristics while performing the excavation. On displacement tests, the final heading direction of ants was determined by assessing the thorax\u0026ndash;head direction vector as they exited the recording area, that is, on the last frame in which they were visible. This vector represents the orientation of each ant at that specific moment. To analyse and visualize this data, the final heading mean vector angles were divided into 24 equal wedges, each spanning a 15-degree angle. These wedges are employed in circular plotting to illustrate how directly oriented towards the nest the ants\u0026rsquo; trajectories were at the displacement site. For analysis, we used the circular statistics package in R (version 4.2.1; R Core Team, 2020) on the exact exit headings of the ants (and not the wedge-sector headings). We analysed the maximum displacement from the nest, the duration and the area covered by the ants\u0026rsquo; trajectory. The nest location was chosen as the origin (0, 0). These measures helped us understand how extensively the ants explored their surroundings during the learning walks and in the case of excavation walks, discern any notable differences compared to their learning walks. Maximum displacement was the thorax position at the maximum distance from the nest. We recorded the duration of each learning and excavation walk from the moment the ant left the nest until it just before they entered the nest. For the area covered during these walks, we calculated the enclosed area of the path joining the thorax positions of the ants, using the convex-hull method to encompass all the points visited by the ant. This area measurement provided a measure of the spatial extent of their exploration.\u003c/p\u003e \u003cp\u003eTo understand how quickly and directly the ants moved on each learning and excavation walk, we calculated speed of the workers along with several characteristics of the trajectory. These path characteristics were based on the positions of the thorax across frames. Speed refers to the magnitude of an ant\u0026rsquo;s velocity and was calculated as the average over the entire trajectory for each ant, excluding the stopping durations. We measure the speed at which ants change their gaze direction by observing how quickly their gaze angle changes over time. As ants walk, their gaze constantly shifts. To determine gaze direction, we draw a line from the thorax coordinates to the head coordinates. Gaze angular velocity is calculated by dividing the change in gaze angle by the change in time. This helps us understand how fast ants adjust their viewing direction.\u003c/p\u003e \u003cp\u003eTo understand path measures of the displacement-test recordings, we used three indices of straightness: \u003cem\u003epath straightness, sinuosity\u003c/em\u003e, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({E}_{max}^{a}\\)\u003c/span\u003e\u003c/span\u003e, each of which relates to the directness of navigation towards a destination. \u003cem\u003eStraightness\u003c/em\u003e is computed as the ratio of the straight-line distance between the release point at the displacement site and the final point in the frame before the ant moved out of the recording area to the overall length of the path. (Batschelet 1981; Deeti et al. 2023b; Islam et al. 2021; Islam et al. 2023). \u003cem\u003eStraightness\u003c/em\u003e ranges from 0 to 1, with larger values indicating straighter paths, while smaller values indicate more curved or convoluted paths. \u003cem\u003eSinuosity\u003c/em\u003e is an estimate of the tortuosity in a path, calculated as \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(S=2{\\left[p\\left(\\frac{1+c}{1-c}+{b}^{2}\\right)\\right]}^{-0.5}\\)\u003c/span\u003e\u003c/span\u003e, where \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(p\\)\u003c/span\u003e\u003c/span\u003e is the mean step length, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(c\\)\u003c/span\u003e\u003c/span\u003e is the mean cosine of turning angles and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(b\\)\u003c/span\u003e\u003c/span\u003e is the coefficient of variation of the step length. A trajectory \u003cem\u003estep\u003c/em\u003e is the movement between the positions of the animal (thorax positions) recorded at consecutive video frames. Accordingly, step lengths are the Euclidean distances between consecutive points along a path, and turning angle refers to the change in direction between two consecutive steps. Sinuosity varies between 0 (straight) and 1 (extremely curved) (Benhamou, 2004). The maximum expected displacement of a path, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({E}_{max}^{a}=\\frac{{\\beta }}{1-{\\beta }}\\)\u003c/span\u003e\u003c/span\u003e, where \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\beta }\\)\u003c/span\u003e\u003c/span\u003e is the mean cosine of turning angles, is a dimensionless value expressed as a function of number of steps, and is consistent with the intuitive meaning of straightness (Cheung et al. 2007). Larger maximum expected displacement values indicate straighter paths, hence greater displacement, while a smaller value suggests more localized or constrained movement. Paths were characterized and visualized in R (version 4.2.1; R Core Team, 2020) using the packages trajr (McLean and Skowron Volponi 2018) and Durga (Khan and McLean 2023). During learning walks, ants frequently displayed a series of stereotypical successive fixations in different head directions by rotating on the spot at one location, known as a \u0026ldquo;scanning bout\u0026rdquo; and we extracted the number of scanning bouts from each learning walk (Deeti et al. 2023c; Lionetti et al. 2023).\u003c/p\u003e \u003cp\u003eWe obtained panoramic images using Richo Theta cameras at each test location, including one directly on top of the nest. These panoramas were utilized to compute the rotational image difference function (rotIDF) between the nest panorama and each test-location panorama. The pixel differences were calculated for each 1\u0026deg; shift in pixels, following methodologies outlined in previous studies (Islam et al. 2022). We calculated the depth of a rotIDF as the mean discrepancy minus the minimum discrepancy.\u003c/p\u003e \u003c/div\u003e \u003cdiv id=\"Sec7\" class=\"Section2\"\u003e \u003ch2\u003eStatistical analysis\u003c/h2\u003e \u003cp\u003eWe used generalized linear mixed models to investigate whether the displacement location affected the route navigation of the foragers on tests. Path characteristics were compared across the four different displacement conditions. In the model, displacement conditions were treated as an independent variable (X) that predicts the dependent variables (Y) and Ant ID was treated as a random factor. We used 4 individual models to compare the effect of the displacement location on \u003cem\u003epath straightness\u003c/em\u003e, \u003cem\u003esinuosity\u003c/em\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({E}_{max}^{a}\\)\u003c/span\u003e\u003c/span\u003e, and speed. Similarly, 4 other models were used to compare characteristics of learning walks and excavation activity, in speed, convex-hull area, maximum displacement, and duration. We conducted a Welch\u0026rsquo;s ANOVA (one-way) in cases of variables with significant heterogeneity of variance. The GLM was formulated using the lme4 package (version 1.1\u0026ndash;27) and fitted using the glmer function (Bates et al., 2015). Since \u003cem\u003epath straightness\u003c/em\u003e and \u003cem\u003esinuosity\u003c/em\u003e are bounded measures (0\u0026ndash;1), we used the binomial family, whereas all of our other response variables are only bounded by zero, and so for them we applied the Gaussian family of models. For paths on tests, we employed the Tukey post hoc test (emmeans package), to perform pair-wise post hoc comparisons for each of our models. Since we tested multiple dependent variables computed from the same data sets of trajectories on tests or on walks at the nest, we adopted p\u0026thinsp;=\u0026thinsp;0.01 as an alpha level to lower type 1 errors. We report all significant terms and the significance values for all key pair-wise comparisons. Statistical analyses were conducted using R (version 4.3.1). We visualized the data using summary characteristics such as median (Box plots), density and mean confidence intervals.\u003c/p\u003e \u003cp\u003eTo compare the learning walks with excavation walks, we examined the first three excavation walks together with the learning walks (one for each ant), so that this factor, the independent variable, contained four levels. The excavation walks all looked similar when we observed them, so that we limited the analysis to a small number that every ant in the study carried out. In the generalized linear-model ANOVA, we applied a priori Helmert contrasts. The learning walk was compared against all three excavation walks in the first contrast. The second contrast compared excavation walk 1 against excavation walks 2 and 3 together, and the third contrast compared excavation walks 2 vs. 3. These contrasts are independent, making up the 3 df in the numerator of the ANOVA. Our hypothesis before data analysis was that only the first Helmert contrast would reach significance.\u003c/p\u003e \u003cp\u003eTo assess the uniform distribution of headings for each test condition (P\u0026thinsp;\u0026gt;\u0026thinsp;0.05), we performed a Rayleigh\u0026rsquo;s test. In case both distributions turned out non-uniform, we planned to compare the mean direction of the two groups using the Watson\u0026ndash;Williams test (alpha\u0026thinsp;=\u0026thinsp;0.05). If the hypothesis of a uniform distribution cannot be rejected for one or both groups, it makes no sense to run this test. Additionally, we examined if final heading orientations significantly clustered around the nest direction at 0 degrees by checking whether 0 degrees fell within the 95% confidence interval (CI) of orientations (Watson tests). V-tests were conducted, with alpha set at P\u0026thinsp;=\u0026thinsp;0.05, to determine if the mean headings were notably clustered around a specified target direction. Single-sample log likelihood ratio tests were also conducted to investigate whether the heading distributions of the ants were uniform in each test condition.\u003c/p\u003e \u003c/div\u003e"},{"header":"RESULTS","content":"\u003cp\u003eAfter emerging from winter hibernation, in the initial days of activity in the new summer season, we observed the naive walks of nest excavators and their subsequent outbound excavating trips. All the ants in this study engaged in a distinctive pattern. They first conducted a single learning walk in the vicinity of the nest. On the same day, following the learning walk, they then actively participated in nest excavation or digging activities. During the learning walk near the nest, the ants walked in a loop in close proximity to the nest and returned to the nest (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003eA). This looping trajectory resembled the pattern observed during foragers\u0026rsquo; first, naive learning walks (Deeti and Cheng 2021b). Subsequently, they excavated sand from inside the nest, dumping it outside by tossing the grain, much like the dumping behaviour observed in an earlier study (Deeti and Cheng 2023a). On these excavation trips, we observed the ants to come straight out, drop the sand, and then turn around and head straight back without any ado; except for tossing the sand, they did not stop, and they did not scan on any of these trips.\u003c/p\u003e\n\u003cp\u003eWe conducted an analysis of various path characteristics associated with the learning and three consecutive sand excavation behaviours. Ant paths covered a significantly smaller area during the excavation trip than the learning trip (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eA). The generalized linear-model ANOVA revealed a significant difference in area between conditions with a priori Helmert contrast-1 (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;4.35, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0006), between the learning walk vs all three excavation walks, but not with contrast-2 (the excavation trip 1 vs. 2 and 3) (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;1.4, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.14) or contrast-3 (excavation trip 2 vs. 3) (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.06, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.804) groups. In maximum displacement from the nest, the learning walks showed numerically larger values than did the excavation walks (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eB). A Welch\u0026rsquo;s one-way ANOVA on the maximum displacement, however, revealed no significant difference between the conditions (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(3,76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;1.04, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.06). While there was not a statistically significant difference in maximum displacement between the learning walks and excavation activity, ants spent a longer time outside during the learning walks compared to the excavation activity (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003eC). The generalized linear-model ANOVA found significant differences between the conditions in contrast-1 and contrast-2 (the learning walk vs all three excavation walks: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,76\u003c/sub\u003e)\u0026thinsp;=\u0026thinsp;4.45, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.001; the excavation trip 1 vs. 2 and 3: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;6.01, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.004) but not in contrast-3, (excavation trip 2 vs. 3: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.06, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.84. During the learning walks, ants stayed longer duration outside than the excavation trips. A Welch\u0026rsquo;s one-way ANOVA on duration uncovered a significant difference between the conditions (F\u003csub\u003e(1, 35\u003c/sub\u003e)\u0026thinsp;=\u0026thinsp;17.2, p\u0026thinsp;=\u0026thinsp;0.00001). The generalized linear-model ANOVA found significant differences between the conditions in contrast-1 (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,76\u003c/sub\u003e)\u0026thinsp;=\u0026thinsp;33.18, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.00001), but not in contrast-2 (the excavation trip 1 vs. 2 and 3: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;1.96, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.056) and in contrast-3 ( excavation trip 2 vs. 3: \u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;1.06, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.3). Thus, the ants moved similarly far from the nest on learning walks and excavation trips, but learning walks were longer in duration and covered more area.\u003c/p\u003e\n\u003cp\u003eWe found a notable difference in mean speed between the excavation activity and the naive learning walk. The mean speed during the learning walk was significantly lower than the speed during the excavation activity (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eA). The a priori Helmert contrast-1 revealed a significant difference (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1, 76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;30.33, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0005). Other Helmert contrasts, however, revealed no significant differences between the conditions in contrast-2 (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,76\u003c/sub\u003e)\u0026thinsp;=\u0026thinsp;3.7, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.012) or contrast-3 (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1,76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;1.01, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.32). In the angular velocity of gaze-direction change, during the learning walks the ants showed a numerically lower angular velocity than during the excavation trips (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eB). This difference resulted in a statistically significant difference between conditions in a priori Helmert contrast-1 (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1, 76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;4.52, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.005), with no significant differences in contrast-2 (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1, 76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.43, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.64) or contrast-3 (\u003cem\u003eF\u003c/em\u003e\u003csub\u003e(1, 76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.33, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.56). During the learning walks ants showed scanning behaviour. Naive learners performed a minimum one scanning bout and a maximum of 3 bouts during the learning walks, but no ant scanned on any of the excavation trips (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003eC), making inferential statistics on this variable both inappropriate and moot.\u003c/p\u003e\n\u003cp\u003eIn order to understand the navigational knowledge of these excavators, after their first excavation trip, each ant was captured just before it entered the nest and displaced to four different locations to the North and East of their nest entrance. The ants\u0026rsquo; final headings at each location showed that the majority of the ants in the 2 m displacements were oriented towards the nest direction of 0 deg. In contrast, ants that were displaced 4 m from the nest were not oriented towards the nest direction from the release point (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e). By the Rayleigh test, the foragers\u0026rsquo; initial orientations were non-uniformly distributed in 2-m displacement tests whereas in 4-m tests they were scattered and uniformly distributed (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e, Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e). In addition, ants in 2-m displacement conditions showed significant V-test results in the nest direction, and the means of their 95% confidence interval of initial heading values include the nest direction 0 deg (Watson test, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026gt;\u0026thinsp;0.05). The log likelihood ratio test for the 2-m tests failed to reject the hypothesis that the distribution was clustered in the home direction (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;\u0026ge;\u0026thinsp;0 .05, k\u0026thinsp;\u0026ge;\u0026thinsp;0, \u0026chi;2\u0026thinsp;\u0026le;\u0026thinsp;1). However, for the 4-m test, the log likelihood ratio rejected the hypothesis that the mean value of distribution was equal to the predicted value (home direction) (\u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.003, k\u0026thinsp;=\u0026thinsp;0.56 and \u0026chi;2\u0026thinsp;=\u0026thinsp;4.6), meaning that the headings were oriented in a different direction from the nest direction.\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n\u003ctable id=\"Tab1\" border=\"1\"\u003e\u003ccaption\u003e\n\u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n\u003cdiv class=\"CaptionContent\"\u003e\n\u003cp\u003eStatistical results for initial heading directions in 2m North, 2m East, 4m North and 4m East displacement tests.\u003c/p\u003e\n\u003c/div\u003e\n\u003c/caption\u003e\n\u003cthead\u003e\n\u003ctr\u003e\n\u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n\u003cth align=\"left\"\u003e\n\u003cp\u003e\u003cspan class=\"Underline\"\u003eMean vector\u003c/span\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e\u003cspan class=\"Underline\"\u003e95% confidence interval\u003c/span\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e\u003cspan class=\"Underline\"\u003eRayleigh test\u003c/span\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003cth colspan=\"2\" align=\"left\"\u003e\n\u003cp\u003e\u003cspan class=\"Underline\"\u003eV test detection 0\u003c/span\u003e\u003csup\u003e\u003cspan class=\"Underline\"\u003e0\u003c/span\u003e\u003c/sup\u003e\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eTest\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003e\u0026micro;\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eMinus\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003ePlus\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eZ\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eP\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eZ\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u003cstrong\u003eP\u003c/strong\u003e\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2m North\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.34\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e335.67\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e5.01\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e7.96\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.91\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;0.005\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e2m East\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e11.95\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e344.05\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e39.85\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e6.60\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.55\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e\u0026lt;\u0026thinsp;0.002\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e4m North\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e289.77\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e232.45\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e317\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e1.83\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.16\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.64\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.26\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e4m East\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e260.11\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e221.71\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e298.51\u0026deg;\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e3.84\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.01\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e-0.47\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd align=\"left\"\u003e\n\u003cp\u003e0.68\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003eWe utilized rotIDF to assess the visual similarity of each displacement location to the nest panorama. Comparisons with the nest panorama revealed detectable minima for all displacement locations, with specific values 2m North location\u0026thinsp;=\u0026thinsp;27.88, 4m North location\u0026thinsp;=\u0026thinsp;35.73, 2m East, location\u0026thinsp;=\u0026thinsp;33. 61, 4m East\u0026thinsp;=\u0026thinsp;30.49 (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e4\u003c/span\u003e). Interestingly, the depth of these minima showed only slight variation across displacement locations: 2m North location\u0026thinsp;=\u0026thinsp;17.49, 4m North location\u0026thinsp;=\u0026thinsp;19.14, 2m East, location\u0026thinsp;=\u0026thinsp;12.48, 4m East\u0026thinsp;=\u0026thinsp;16.04 (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e5\u003c/span\u003e). This suggests that all test locations had a \u0026lsquo;best\u0026rsquo; direction pointing roughly towards the nest.\u003c/p\u003e\n\u003cp\u003eWe checked whether ants showed any differences in their path characteristics from oriented and non-oriented displacement locations. Firstly, in tortuosity ants appear to show lower \u003cem\u003esinuosity\u003c/em\u003e in the 4mN tests compared with the other conditions. However, the generalized linear model ANOVA showed no significant differences in \u003cem\u003esinuosity\u003c/em\u003e across the four displacement conditions (Z\u003csub\u003e(3, 76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.64, \u003cem\u003ep\u003c/em\u003e 1.04), (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eA). Secondly, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\text{E}}_{\\text{m}\\text{a}\\text{x}}^{\\text{a}}\\)\u003c/span\u003e\u003c/span\u003e appeared similar across conditions. The generalized linear model ANOVA showed no significant difference in \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\({\\text{E}}_{\\text{m}\\text{a}\\text{x}}^{\\text{a}}\\)\u003c/span\u003e\u003c/span\u003e across the four displacement conditions (Z\u003csub\u003e(3, 76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;1.05, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.061), (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eB). Finally, with straightness, the 4mN displacement condition appears to be lower on average than the other conditions (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e6\u003c/span\u003eC). The generalized linear model ANOVA showed a significant difference in \u003cem\u003estraightness\u003c/em\u003e across the four conditions (Z\u003csub\u003e(3, 76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;12.59, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;0.0001), but the Tukey post-hoc comparisons, however, showed no significant differences between any pair of displacement conditions (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). On the whole then, path characteristics at different locations were similar, or differences were at most idiosyncratic, as we did not find any pairwise significant contrasts in any variable. During the displacement test we had not found any noticeable impact on the speed and angular deviation of the ants (Fig.\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003eA and \u003cspan class=\"InternalRef\"\u003e7\u003c/span\u003eB). The generalized linear-model ANOVA showed no significant difference between the displacement locations in mean speed (Z \u003csub\u003e(3, 76)\u003c/sub\u003e =\u0026ndash;0.52, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.61) and in the angular velocity of gaze-direction change (Z \u003csub\u003e(3, 76)\u003c/sub\u003e\u0026thinsp;=\u0026thinsp;0.8, \u003cem\u003ep\u003c/em\u003e\u0026thinsp;=\u0026thinsp;1.14).\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cdiv class=\"gridtable\"\u003e\n\u003ctable id=\"Tab2\" border=\"1\"\u003e\u003ccaption\u003e\n\u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n\u003cdiv class=\"CaptionContent\"\u003e\n\u003cp\u003eTukey post hoc comparisons of statistical results for straightness between the displacement locations (alpha\u0026thinsp;=\u0026thinsp;0.01).\u003c/p\u003e\n\u003c/div\u003e\n\u003c/caption\u003e\n\u003cthead\u003e\n\u003ctr style=\"height: 35px;\"\u003e\n\u003cth style=\"height: 35px;\" colspan=\"5\" align=\"left\"\u003e\n\u003cp\u003eStraightness\u003c/p\u003e\n\u003c/th\u003e\n\u003c/tr\u003e\n\u003c/thead\u003e\n\u003ctbody\u003e\n\u003ctr style=\"height: 35px;\"\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003eCondition\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003eratio\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003eSE\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003eZ ratio\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e\u003cem\u003ep\u003c/em\u003e value\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr style=\"height: 35px;\"\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003eeast2m / east4m\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e1.228\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.576\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.437\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.9721\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr style=\"height: 35px;\"\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003eeast2m / north2m\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e1.078\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.489\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.167\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.9984\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr style=\"height: 35px;\"\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003eeast2m / north4m\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e1.802\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.948\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e3.303\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.0078\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr style=\"height: 35px;\"\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003eeast4m / north2m\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.878\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.419\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e-0.272\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.993\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr style=\"height: 35px;\"\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003eeast4m / north4m\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e1.468\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.803\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.702\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35px;\" align=\"left\"\u003e\n\u003cp\u003e0.8965\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003ctr style=\"height: 35.2247px;\"\u003e\n\u003ctd style=\"height: 35.2247px;\" align=\"left\"\u003e\n\u003cp\u003enorth2m / north4m\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35.2247px;\" align=\"left\"\u003e\n\u003cp\u003e1.671\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35.2247px;\" align=\"left\"\u003e\n\u003cp\u003e0.891\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35.2247px;\" align=\"left\"\u003e\n\u003cp\u003e2.764\u003c/p\u003e\n\u003c/td\u003e\n\u003ctd style=\"height: 35.2247px;\" align=\"left\"\u003e\n\u003cp\u003e0.0308\u003c/p\u003e\n\u003c/td\u003e\n\u003c/tr\u003e\n\u003c/tbody\u003e\n\u003c/table\u003e\n\u003c/div\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e"},{"header":"DISCUSSION","content":"\u003cp\u003eIn summary, our observations of nest excavators\u0026rsquo; behaviours in \u003cem\u003eMelophorus bagoti\u003c/em\u003e showed that before becoming a nest excavator, workers made a single learning walk around the nest. This first walk of naive excavating ants was close to the nest, short in duration, and covered only a small area, similar to the first learning walk of foragers. On the next trips, ants started excavating the nest, carrying excavated sand in their mandibles within 5\u0026ndash;10 cm around the nest and tossing the grain in a stereotypical way before returning straight to the nest. This activity lasted a shorter time and covered a smaller area than did their learning walks, and the ants moved faster and oscillated their gaze directions faster than they did on their learning walks. Other work that we are still analysing suggest that gaze swings are common when ants navigate homebound or outbound. The faster gaze-diretion swings on excavation walks may be coupled with the faster speed of walking. After the first excavation trip, when ants were displaced 2 m and 4 m North and East from their nest, they could orient towards the nest direction from 2 m North and East but not from 4 m away, suggesting that a catchment area between 2 and 4 m had been learned in one walk in our setting, similar to results after the first, naive learning walks of foragers. These latter ants could also orient nestward from 2 m but not from 4 m (Deeti and Cheng 2021b). Despite these differences in orientational performance, path characteristics at all the test sites were on the whole similar. As discussed further below, this similarity perhaps reflects the similarity across all the locations in their rotational image difference functions when compared to the panoramic image at the nest. In summary, ants perform a learning walk before they start excavating the nest and they generalize the views from that single learning trip to locations 2 m away from the nest.\u003c/p\u003e \u003cp\u003eThe phenomenon of learning walks and learning flights as a precursor to the transition to foraging is well established across various hymenopteran species, including honeybees (Becker 1958; Capaldi and Dyer 1999; Capaldi et al. 2000; Lehrer 1991, 1993; Vollbehr 1975), bumblebees (Hempel de Ibarra et al. 2009), wasps (St\u0026uuml;rzl et al. 2016; Zeil 1993a, 1993b), wood ants (Judd and Collett 1998; Nicholson et al. 1999), and desert ants (Deeti and Cheng 2021a; Fleischmann et al. 2016; M\u0026uuml;ller and Wehner 2010). Through multiple flights or walks around their nest, these insects engage in exploratory behaviour, progressively covering more distance and exploring a wider range of directions to enhance their spatial knowledge around the nest (Capaldi et al. 2000; Deeti and Cheng 2021b; Fleischmann et al. 2016). It is noteworthy that this exploration often involves walks or flights in all quadrants around the nest over successive trips. We have discovered that another outdoor performer, the excavator, also engages in a single learning walk around the nest area before undertaking their work. This seems surprising because excavating ants deposit excavated sand within a 5\u0026ndash;10 cm area around the nest. It is likely that this learning walk still serves to facilitate the return trip on such a short journey. Possible functions include calibrating the sky compass and odometer (Wehner 2020) for path integration and also learning the visual scene surrounding the nest. Using a combination of path integration and view-based homing might expedite the journey home. Testing the function of the learning walk is not easy because our observations suggest that the workers will not excavate without first taking a learning walk. At the moment, we have no way to induce them to take an excavating trip without any learning walks.\u003c/p\u003e \u003cp\u003eThe naive learning walk of red honey ants engaged in excavation activities displays parallels with the first learning walks of desert ant (\u003cem\u003eCataglyphis\u003c/em\u003e, \u003cem\u003eMelophorus\u003c/em\u003e) foragers (Fleischmann et al. 2016; Deeti and Cheng 2021b). Comparing \u003cem\u003eMelophorus bagoti\u003c/em\u003e would-be excavators and would-be foragers, the durations, lasting less than a minute, maximum displacement (less than 30 cm) and area covered (~\u0026thinsp;20 cm\u003csup\u003e2\u003c/sup\u003e) are similar. Would-be foragers, however, perform more than one learning walk before setting off to forage. Given that foraging takes place at a much greater distance than excavating, the multiple learning walks make functional sense. Multiple walks are needed to return from long distances. After all, a single learning walk is good for returning only from ~\u0026thinsp;2 m away (Deeti and Cheng 2021b). However, multiple learning walks up to 1 m distance in all directions from the nest entrance allowed them to orient nestwards from 10 m distance (\u003cem\u003eCataglyphis\u003c/em\u003e: Fleischmann et al. 2018; \u003cem\u003eMelophorus\u003c/em\u003e: Wystrach et al. 2012; Deeti et al. 2020). In essence, the multiple learning walks serve to increase the catchment range for the ants. By exploring and learning the surroundings in various directions from the nest entrance, the ants expand their spatial knowledge, allowing them to navigate and return from more extended distances with precision. This adaptive strategy ensures efficient orientation and successful homeward journeys, especially in the context of foraging activities that necessitate exploration over more significant distances.\u003c/p\u003e \u003cp\u003eUnfamiliarity is thought to trigger \u0026lsquo;uncertainty\u0026rsquo; in solitary foraging ants, although how and what is encoded about uncertainty remains unclear. Experimental manipulations have included changes in the visual panorama and conflicts between path integration and view-based navigation. These alterations manifested as a reduction in the path straightness of foragers, indicating one response to navigational uncertainty induced by visual changes (Wystrach et al. 2019, 2020; Islam et al. 2021; Deeti et al. 2023b). Another change is slower travel speeds (Buehlmann et al. 2018). In the current study, path characteristics were on the whole similar across displacement-test locations, even though the ants were homeward oriented in their initial trajectories at 2 m distance but not at 4 m distance. Our image analysis of panoramic views at the test sites perhaps provides an explanation. When compared with the panoramic view at the nest, the views at all test sites differed from the nest view by similar amounts on a pixel-by-pixel basis. A pixel-by-pixel analysis is unlikely to the be sole analysis carried out by the ants\u0026rsquo; visual system, which is known to abstract features such as the skyline where terrestrial objects meet the sky (Graham and Cheng 2009) and the fraction position of mass, the proportion of the visual scene to the left and right of the goal direction (Lent et al. 2013). Nevertheless, the pixel-by-pixel analyses serve to give proxy measures of similarity to the panoramic view at the nest.\u003c/p\u003e \u003cp\u003eWhat the image analysis does not explain is why ants were homeward oriented in initial trajectories at 2 m but not at 4 m, when all the panoramic views contained a minimum-mismatch best direction pointing roughly at the nest direction. By the image analysis, the ants should have been able to orient homewards from 4 m as well as at 2 m. It is possible that other modalities of cues are used as well, such as olfactory (Buehlmann et al. 2015), vibratory (Buehlmann et al. 2012), or magnetic cues (Buehlmann et al. 2012), and these cues are less helpful at 4 m compared with 2 m, but we have no evidence for the use of these cues in this study or for this species. It remains unexplained why our ants could not orient homewards from 4 m.\u003c/p\u003e \u003cp\u003eIn conclusion, \u003cem\u003eM. bagoti\u003c/em\u003e ants engage in learning walks in the proximity of the nest before taking on the role of a nest excavator. These learning walks extend up to 15 cm away, allowing the ants to acquire spatial knowledge about their surrounding panorama up to at least 2 m away from the nest. Future studies should delve into a more detailed exploration of the differences and similarities in the behaviour of different worker castes of naive ants, based on their specific role in the colony, such as guarding the nest, foraging, dumping and excavation activities. Additionally, there is a need for an examination of the underlying neurological processes that support learning walks in ants. This research avenue holds the potential to enhance our understanding of the intricate mechanisms governing ant behaviour and learning strategies.\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eACKNOWLEDGEMENTS\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe acknowledge the traditional custodians of the land upon which this research was conducted, the Arrernte people. Their culture and customs have nurtured and sustained this land since the Dreamtime and continue to do so today. We pay our respects to their Elders past and present. We thank the Centre for Appropriate Technology at Alice Springs, Australia for letting us work on their property and providing some storage space, and the CSIRO Arid Zone Research at Alice Springs for administrative support. We are also thankful to Vito A. G. Lionetti and Cody Freas for helping us to take panoramic images.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFunding\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe work was supported by the Australian Research Council [DP200102337] and by the Australian Defence [AUSMURIB000001 associated with ONR MURI grant N00014-19-1-2571].\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eExperimental design: SD. Data collection: SD; Data analysis: SD and DJM; Writing: SD, and KC.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics standards\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAustralia has no ethical regulations regarding work with insects. The study was non-invasive and no long-term aversive effects were found on the nests or on the individuals studied.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eCompeting interests\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eKC is an associate editor of this journal. The authors declare no other competing or financial interests.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eSupplementary videos, Excel file and R scripts are available at Open Science framework:\u0026nbsp;\u0026nbsp;https://osf.io/gqb8r/.\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n\u003cli\u003eBates D, M\u0026auml;chler M, Bolker B, Walker S (2015) Fitting linear mixed-effects models Usinglme4. Journal of Statistical Software, 67(1). https://doi.org/10.18637/jss.v067.i01\u003c/li\u003e\n\u003cli\u003eBatschelet E (1981) \u003cem\u003eCircular statistics in biology / Edward Batschelet\u003c/em\u003e. Academic Press.\u003c/li\u003e\n\u003cli\u003eBecker L (1958) Untersuchungen \u0026uuml;ber das Heimfindeverm\u0026ouml;gen der Bienen. Zeitschrift f\u0026uuml;r vergleichende Physiologie, 41, 1-25.\u003c/li\u003e\n\u003cli\u003eBenhamou S (2004) How to reliably estimate the tortuosity of an animal's path: straightness, sinuosity, or fractal dimension? \u003cem\u003eJ Theor Biol\u003c/em\u003e, 229(2), 209-220. https://doi.org/10.1016/j.jtbi.2004.03.016\u003c/li\u003e\n\u003cli\u003eBuehlmann C, Hansson BS, Knaden M (2012) Desert ants learn vibration and magnetic landmarks. \u003cem\u003ePlos one\u003c/em\u003e, 7(3), e33117. https://doi.org/10.1371/journal.pone.0033117\u003c/li\u003e\n\u003cli\u003eBuehlmann C, Graham P, Hansson BS, Knaden M (2015) Desert ants use olfactory scenes for navigation. \u003cem\u003eAnimal Behaviour\u003c/em\u003e, 106, 99-105. https://doi.org/10.1016/j.anbehav.2015.04.029\u003c/li\u003e\n\u003cli\u003eBuehlmann C, Fernandes ASD, Graham P (2018) The interaction of path integration and terrestrial visual cues in navigating desert ants: what can we learn from path characteristics?. \u003cem\u003eJournal of Experimental Biology\u003c/em\u003e, 221(1), jeb167304. https://doi.org/10.1242/jeb.167304\u003c/li\u003e\n\u003cli\u003eCapaldi EA, Dyer FC (1999) The role of orientation flights on homing performance in honeybees. \u003cem\u003eJournal of Experimental Biology\u003c/em\u003e, 202(12), 1655-1666.\u003c/li\u003e\n\u003cli\u003eCapaldi EA, Smith AD, Osborne JL, Fahrbach SE, Farris SM, Reynolds DR, Edwards AS, Martin A, Robinson GE, Poppy GM (2000) Ontogeny of orientation flight in the honeybee revealed by harmonic radar. \u003cem\u003eNature\u003c/em\u003e, 403(6769), 537-540. https://doi.org/10.1038/35000564\u003c/li\u003e\n\u003cli\u003eCheng K, Narendra A, Sommer S, Wehner R (2009) Traveling in clutter: navigation in the Central Australian desert ant \u003cem\u003eMelophorus bagoti\u003c/em\u003e. \u003cem\u003eBehavioural processes\u003c/em\u003e, 80(3), 261-268. https://doi.org/10.1016/j.beproc.2008.10.015\u003c/li\u003e\n\u003cli\u003eCheung A, Zhang S, Stricker C, \u0026nbsp;Srinivasan MV (2007) Animal navigation: the difficulty of moving in a straight line. \u003cem\u003eBiological Cybernetics\u003c/em\u003e, 97(1), 47-61. https://doi.org/10.1007/s00422-007-0158-0\u003c/li\u003e\n\u003cli\u003eChristian KA, Morton SR\u0026nbsp; (1992)\u0026nbsp; Extreme thermophilia in a central Australian ant, \u003cem\u003eMelophorus bagot\u003c/em\u003ei. \u003cem\u003ePhysiological Zoology\u003c/em\u003e 65(5), 885-905.\u003c/li\u003e\n\u003cli\u003eCollett M, Collett TS (2000) How do insects use path integration for their navigation? \u003cem\u003eBiological Cybernetics\u003c/em\u003e, 83(3), 245-259.\u003c/li\u003e\n\u003cli\u003eDeeti S, FujiiK, Cheng K (2020) The effect of spatially restricted experience on extrapolating learned views in desert ants, \u003cem\u003eMelophorus bagoti\u003c/em\u003e. \u003cem\u003eAnimal Cognition\u003c/em\u003e, 23, 1063-1070. https://doi.org/10.1007/s10071-020-01359-2\u003c/li\u003e\n\u003cli\u003eDeeti S, Cheng K (2021a) Observations on relocation of a nest by a colony of red honey ants (\u003cem\u003eMelophorus bagoti\u003c/em\u003e). \u003cem\u003eNorthern Territory Naturalist\u003c/em\u003e, 30, 100-107.\u003c/li\u003e\n\u003cli\u003eDeeti S, Cheng K (2021b) Learning walks in an Australian desert ant,\u003cem\u003e Melophorus bagoti\u003c/em\u003e. J\u003cem\u003eournal of Experimental Biology,\u003c/em\u003e 224(16), jeb242177. https://doi.org/10.1242/jeb.242177\u003c/li\u003e\n\u003cli\u003eDeeti S, Freas C, Murray T, Cheng K (2023a) Dumping behaviour of Australian desert ants (\u003cem\u003eMelophorus bagoti\u003c/em\u003e)(Hymenoptera: Formicidae). \u003cem\u003eInsectes sociaux\u003c/em\u003e, 70(2), 225-232. https://doi.org/10.1007/s00040-023-00911-w\u003c/li\u003e\n\u003cli\u003eDeeti S, Islam M, Freas C, Murray T, Cheng K (2023b) Intricacies of running a route without success in night-active bull ants (\u003cem\u003eMyrmecia midas\u003c/em\u003e). \u003cem\u003eJournal of Experimental Psychology: Animal Learning and Cognition\u003c/em\u003e, 49(2), 111. https://doi.org/10.1037/xan0000350\u003c/li\u003e\n\u003cli\u003eDeeti S, Cheng K, Graham P Wystrach A, (2023c)\u0026nbsp; Scanning behaviour in ants: an interplay between random-rate processes and oscillators. \u003cem\u003eJournal of\u003c/em\u003e \u003cem\u003eComparative Physiology A\u003c/em\u003e 209, 625\u0026ndash;639. https://doi.org/10.1007/s00359-023-01628-8\u003c/li\u003e\n\u003cli\u003eDeeti S, McLean DJ, Cheng K (2024) Desert ants (Melophorus bagoti) learn from experience to improve waste dumping, 22 January 2024, PREPRINT (Version 1) available at Research Square [https://doi.org/10.21203/rs.3.rs-3872453/v1]\u003c/li\u003e\n\u003cli\u003eFleischmann PN, Christian M, M\u0026uuml;ller VL, R\u0026ouml;ssler W, Wehner R (2016) Ontogeny of learning walks and the acquisition of landmark information in desert ants, Cataglyphis fortis. \u003cem\u003eJournal of Experimental Biology\u003c/em\u003e, 219(19), 3137-3145. https://doi.org/10.1242/jeb.140459\u003c/li\u003e\n\u003cli\u003eFleischmann PN, R\u0026ouml;ssler W, Wehner R (2018) Early foraging life: spatial and temporal aspects of landmark learning in the ant \u003cem\u003eCataglyphis noda\u003c/em\u003e. \u003cem\u003eJournal of Comparative Physiology A\u003c/em\u003e, 204(6), 579-592. https://doi.org/10.1007/s00359-018-1260-6\u003c/li\u003e\n\u003cli\u003eH\u0026ouml;lldobler B, Wilson EO (1990) \u003cem\u003eThe ants\u003c/em\u003e. Harvard University Press.\u003c/li\u003e\n\u003cli\u003eGraham P, Cheng K (2009) Ants use the panoramic skyline as a visual cue during navigation. \u003cem\u003eCurrent Biology\u003c/em\u003e, 19(20), R935-R937. https://doi.org/10.1016/j.cub.2009.08.015\u003c/li\u003e\n\u003cli\u003eIslam M, Deeti S, Kamhi JF, Cheng K (2021) Minding the gap: learning and visual scanning behaviour in nocturnal bull ants. \u003cem\u003eJournal of Experimental Biology\u003c/em\u003e, 224(14). https://doi.org/10.1242/jeb.242245\u003c/li\u003e\n\u003cli\u003eIslam M, Deeti S, Murray T, Cheng K\u0026nbsp; (2022)\u0026nbsp; What view information is most important in the homeward navigation of an Australian bull ant, \u003cem\u003eMyrmecia midas\u003c/em\u003e? \u003cem\u003eJournal of Comparative Physiology A\u003c/em\u003e 208(5), 545-559. https://doi.org/10.1007/s00359-022-01565-y\u003c/li\u003e\n\u003cli\u003eIslam M, Deeti S, Mahmudah Z, Kamhi JF, Cheng K (2023) Detouring while foraging up a tree: What bull ants (Myrmecia midas) learn and their reactions to novel sensory cues. \u003cem\u003eJournal of Comparative Psychology\u003c/em\u003e, 137(1), 4. https://doi.org/10.1037/com0000333\u003c/li\u003e\n\u003cli\u003eJayatilaka P, Murray T, Narendra A, Zeil J (2018) The choreography of learning walks in the Australian jack jumper ant \u003cem\u003eMyrmecia croslandi\u003c/em\u003e. \u003cem\u003eJournal of Experimental Biology\u003c/em\u003e, 221(20). https://doi.org/10.1242/jeb.185306\u003c/li\u003e\n\u003cli\u003eJudd S, Collett T (1998) Multiple stored views and landmark guidance in ants. \u003cem\u003eNature\u003c/em\u003e,\u003cem\u003e 392\u003c/em\u003e(6677), 710-714.\u003c/li\u003e\n\u003cli\u003eKhan MK, McLean DJ (2023) Durga: An R package for effect size estimation and visualisation. \u003cem\u003ebioRxiv\u003c/em\u003e, 2023.2002.2006.526960. https://doi.org/10.1101/2023.02.06.526960\u003c/li\u003e\n\u003cli\u003eKnaden M, Graham P (2016) The sensory ecology of ant navigation: from natural environments to neural mechanisms. \u003cem\u003eAnnual review of entomology\u003c/em\u003e, 61, 63-76. https://doi.org/10.1146/annurev-ento-010715-023703\u003c/li\u003e\n\u003cli\u003eLehrer M (1991) Bees which turn back and look. \u003cem\u003eNaturwissenschaften\u003c/em\u003e, 78(6), 274-276.\u003c/li\u003e\n\u003cli\u003eLehrer M (1993) Why do bees turn back and look?\u003cem\u003e Journal of Comparative Physiology A\u003c/em\u003e, 172, 549-563.\u003c/li\u003e\n\u003cli\u003eLent D, Graham P, Collett TS (2013) Visual scene perception in navigating wood ants. \u003cem\u003eCurrent Biology\u003c/em\u003e, 23(8):684\u0026ndash;690. https://doi.org/10.1016/j.cub.2013.03.016\u003c/li\u003e\n\u003cli\u003eLionetti VAG, Deeti S, Murray T, Cheng K (2023)\u0026nbsp; Resolving conflict between aversive and appetitive learning of views: how ants shift to a new route during navigation. \u003cem\u003eLearning and Behaviour\u003c/em\u003e, 51, 446\u0026ndash;457 (2023). https://doi.org/10.3758/s13420-023-00595-z.\u003c/li\u003e\n\u003cli\u003eMcLean DJ, Skowron Volponi MA (2018) trajr: An R package for characterisation of animal trajectories. \u003cem\u003eEthology\u003c/em\u003e, 124(6), 440-448. https://doi.org/https://doi.org/10.1111/eth.12739\u003c/li\u003e\n\u003cli\u003eM\u0026uuml;ller M, Wehner R (2010) Path integration provides a scaffold for landmark learning in desert ants. \u003cem\u003eCurrent Biology\u003c/em\u003e, 20(15), 1368-1371. https://doi.org/10.1016/j.cub.2010.06.035\u003c/li\u003e\n\u003cli\u003eMuser B, Sommer S, Wolf H, Wehner R (2005) Foraging ecology of the thermophilic Australian desert ant, \u003cem\u003eMelophorus bagoti\u003c/em\u003e. \u003cem\u003eAustralian Journal of Zoology\u003c/em\u003e, 53(5), 301-311. https://doi.org/10.1071/ZO05023\u003c/li\u003e\n\u003cli\u003eNicholson D, Judd S, Cartwright B, Collett T (1999) Learning walks and landmark guidance in wood ants (\u003cem\u003eFormica rufa\u003c/em\u003e). \u003cem\u003eJournal of Experimental Biology\u003c/em\u003e, 202(13), 1831-1838.\u003c/li\u003e\n\u003cli\u003eSchultheiss P, Cheng K, Reynolds AM (2015) Searching behavior in social Hymenoptera. \u003cem\u003eLearning and Motivation\u003c/em\u003e, 50, 59-67. https://doi.org/10.1016/j.lmot.2014.11.002\u003c/li\u003e\n\u003cli\u003eSchultheiss P, Nooten SS (2013) Foraging patterns and strategies in an Australian desert ant. \u003cem\u003eAustral Ecology\u003c/em\u003e, 38(8), 942-951. https://doi.org/10.1111/aec.12037\u003c/li\u003e\n\u003cli\u003eSt\u0026uuml;rzl W, Zeil J, Boeddeker N, Hemmi JM (2016) How wasps acquire and use views for homing. \u003cem\u003eCurrent Biology\u003c/em\u003e, 26(4), 470-482. https://doi.org/10.1016/j.cub.2015.12.052\u003c/li\u003e\n\u003cli\u003eVollbehr J (1975) ZUR ORIENTIERUNG JUNGER HONIGBIENEN BEI IHREM. I. ORIENTIERUNGSFLUG.\u003c/li\u003e\n\u003cli\u003eWehner R.(2003) Desert ant navigation: how miniature brains solve complex tasks. \u003cem\u003eJournal of Comparative Physiology A\u003c/em\u003e, 189, 579-588. https://doi.org/10.1007/s00359-003-0431-1\u003c/li\u003e\n\u003cli\u003eWehner R (2020) \u003cem\u003eDesert navigator: the journey of an ant\u003c/em\u003e. Harvard University Press.\u003c/li\u003e\n\u003cli\u003eWehner R, Meier C, Zollikofer C (2004) The ontogeny of foragwehaviour in desert ants, \u003cem\u003eCataglyphis bicolor\u003c/em\u003e. \u003cem\u003eEcological Entomology\u003c/em\u003e, 29(2), 240-250. https://doi.org/10.1111/j.0307-6946.2004.00591.x\u003c/li\u003e\n\u003cli\u003eWehner R, Srinivasan MV (1981) Searching behaviour of desert ants, genus \u003cem\u003eCataglyphis\u003c/em\u003e (Formicidae, Hymenoptera). \u003cem\u003eJournal of comparative physiology\u003c/em\u003e, 142, 315-338.\u003c/li\u003e\n\u003cli\u003eWehner R, Srinivasan MV (2003) Path integration in insects. \u003cem\u003eThe neurobiology of spatial behaviour\u003c/em\u003e, 9-30.\u003c/li\u003e\n\u003cli\u003eWilson EO (1987) Causes of ecological success: the case of the ants. J\u003cem\u003eournal of animal Ecology\u003c/em\u003e, 56(1), 1-9.\u003c/li\u003e\n\u003cli\u003eWystrach A, Graham P (2012) What can we learn from studies of insect navigation? \u003cem\u003eAnimal Behaviour\u003c/em\u003e, 84(1), 13-20. https://doi.org/10.1016/j.anbehav.2012.04.017\u003c/li\u003e\n\u003cli\u003eWystrach A, Schwarz S, Baniel A, Cheng K (2013) Backtracking behaviour in lost ants: an additional strategy in their navigational toolkit. \u003cem\u003eProceedings of the Royal Society B: Biological Sciences\u003c/em\u003e, 280(1769), 20131677. https://doi.org/10.1098/rspb.2013.1677\u003c/li\u003e\n\u003cli\u003eWystrach A, Schwarz S, Graham P, Cheng K (2019) Running paths to nowhere: repetition of routes shows how navigating ants modulate online the weights accorded to cues. \u003cem\u003eAnimal Cognition\u003c/em\u003e Vol. 22, pp. 213\u0026ndash;222. https://doi.org/10.1007/s10071-019-01236-7\u003c/li\u003e\n\u003cli\u003eWystrach A, Buehlmann C, Schwarz S, Cheng K, Graham P (2020) Rapid aversive and memory trace learning during route navigation in desert ants.\u0026nbsp;\u003cem\u003eCurrent biology\u003c/em\u003e,\u0026nbsp;30(10), 1927-1933. https://doi.org/10.1016/j.cub.2020.02.082\u003c/li\u003e\n\u003cli\u003eZeil J (1993a) Orientation flights of solitary wasps (\u003cem\u003eCerceris Sphecidae\u003c/em\u003e; Hymenoptera) I. Description of flight. \u003cem\u003eJournal of Comparative Physiology A\u003c/em\u003e, 172, 189-205.\u003c/li\u003e\n\u003cli\u003eZeil J (1993b) Orientation flights of solitary wasps (\u003cem\u003eCerceris Sphecidae\u003c/em\u003e; Hymenoptera) II. Similarities between orientation and return flights and the use of motion parallax. \u003cem\u003eJournal of Comparative Physiology A\u003c/em\u003e, 172, 207-222.\u003c/li\u003e\n\u003cli\u003eZeil J (2012) Visual homing: an insect perspective. Current opinion in neurobiology, 22(2), 285-293. \u0026nbsp;https://doi.org/10.1016/j.conb.2011.12.008\u003c/li\u003e\n\u003cli\u003eZeil J, Hofmann MI, Chahl JS (2003) Catchment areas of panoramic snapshots in outdoor scenes. \u003cem\u003eJOSA A\u003c/em\u003e, 20(3), 450-469. https://doi.org/10.1364/JOSAA.20.000450\u003c/li\u003e\n\u003cli\u003eZeil J, Narendra A, St\u0026uuml;rzl W (2014) Looking and homing: how displaced ants decide where to go. \u003cem\u003ePhilosophical Transactions of the Royal Society B: Biological Sciences\u003c/em\u003e, 369(1636), 20130034. https://doi.org/10.1098/rstb.2013.0034\u003c/li\u003e\n\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"animal-cognition","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"anco","sideBox":"Learn more about [Animal Cognition](http://link.springer.com/journal/10071)","snPcode":"10071","submissionUrl":"https://submission.nature.com/new-submission/10071/3","title":"Animal Cognition","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Excavation, Navigation, Red honey ant, Exploratory walks","lastPublishedDoi":"10.21203/rs.3.rs-3908727/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-3908727/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe Australian red honey ant, \u003cem\u003eMelophorus bagoti\u003c/em\u003e, stands out as the most thermophilic ant in Australia, engaging in all outdoor activities during the hottest periods of the day during summer months. This species of desert ants often navigates by means of path integration and learning landmark cues around the nest. In our study, we observed the outdoor activities of \u003cem\u003eM. bagoti\u003c/em\u003e workers engaged in nest excavation, the maintenance of the nest structure, primarily by taking excess sand out of the nest. Before undertaking nest excavation, the ants conducted a single exploratory walk. Following their initial learning expedition, these ants then engaged in nest excavation activities. Consistent with previous findings on pre-foraging learning walks, after just one learning walk, the desert ants in our study demonstrated the ability to return home from locations 2 metres away from the nest, although not from locations 4 metres away. These findings indicate that even for activities like dumping excavated sand within a range of 5\u0026ndash;10 cm outside the nest, these ants learn and utilize the visual landmark panorama around the nest.\u003c/p\u003e","manuscriptTitle":"Nest excavators’ learning walks in the Australian desert ant Melophorus bagoti","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-02-02 20:17:14","doi":"10.21203/rs.3.rs-3908727/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-04-09T20:42:33+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-02-20T21:30:09+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"4c6a89e8-ac1e-4c94-b42f-21d3c63bd13d","date":"2024-02-03T17:53:12+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-02-02T21:29:42+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-02-02T21:22:20+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-01-31T10:02:30+00:00","index":"","fulltext":""},{"type":"submitted","content":"Animal Cognition","date":"2024-01-29T10:32:06+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"animal-cognition","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"anco","sideBox":"Learn more about [Animal Cognition](http://link.springer.com/journal/10071)","snPcode":"10071","submissionUrl":"https://submission.nature.com/new-submission/10071/3","title":"Animal Cognition","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"e8b7a76d-2493-4f8e-ae55-135ca54cec26","owner":[],"postedDate":"February 2nd, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2024-05-25T00:33:43+00:00","versionOfRecord":{"articleIdentity":"rs-3908727","link":"https://doi.org/10.1007/s10071-024-01877-3","journal":{"identity":"animal-cognition","isVorOnly":false,"title":"Animal Cognition"},"publishedOn":"2024-05-24 00:33:43","publishedOnDateReadable":"May 24th, 2024"},"versionCreatedAt":"2024-02-02 20:17:14","video":"","vorDoi":"10.1007/s10071-024-01877-3","vorDoiUrl":"https://doi.org/10.1007/s10071-024-01877-3","workflowStages":[]},"version":"v1","identity":"rs-3908727","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-3908727","identity":"rs-3908727","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}