Aggressive spiders make the wrong decision in a difficult task

Aggressive spiders make the wrong decision in a difficult task Abstract Accurate and timely decisions are critical for foraging, predator avoidance, and reproductive success. However, there is often a trade-off between speed and accuracy in decision-making, where individuals that make decisions more quickly make more mistakes. An individual’s personality may influence its decision-making style (i.e., whether it errs more in the speed or accuracy of a decision) and this relationship may change depending on contexts. Despite growing research on invertebrate personality, how personality correlates with decision-making style is still largely unknown and little research has assessed these relationships across tasks of varying difficulty. Here, we test the relationship between aggressiveness and decision-making style in Portia labiata, a specialized spider-eating jumping spider, in both a simple and a difficult task. We found that aggressive spiders made fewer directional changes before completing the tasks, regardless of task difficulty. However, decision accuracy was jointly determined by both aggressiveness and task difficulty. Aggressive spiders made more accurate decisions in the simple task, while docile spiders made more accurate decisions in the difficult task. We conclude that the relationship between personality and decision-making style in P. labiata is context dependent. We also discuss how the association between aggressiveness and decision-making style may serve important functions in maintaining behavioral variation in a natural population. INTRODUCTION The speed and accuracy of decision-making can be critical in the context of foraging, predator avoidance, and reproductive success (Ings and Chittka 2008; Raine and Chittka 2008; Chittka et al. 2009; Cole et al. 2012; Morand‐Ferron et al. 2016). Making an accurate decision requires sufficient information, but processing this information takes time, and most decisions need to be made quickly for the decision-maker to benefit (Bogacz et al. 2010). Therefore, there is the potential for a trade-off between these attributes, where an individual may make either a fast, error-prone decision or a slow, accurate decision (Chittka et al. 2009; Bogacz et al. 2010). Individuals are generally consistent in decision-making styles (i.e., whether they more often err on the speed side of the trade-off or the accuracy side; Sih and Del Giudice 2012). Personality describes consistent interindividual differences in behavior in a population. An individual’s personality is known to influence its decision-making, but whether and how personality influences decision-making across contexts has rarely been studied explicitly, especially in invertebrates (Sih et al. 2004; Carere and Locurto 2011; Jandt et al. 2014; Ducatez et al. 2015; Wang et al. 2015; Nawroth et al. 2017). Personality can influence individual cognitive task performance (Sih and Del Giudice 2012; Griffin et al. 2015; Morand‐Ferron et al. 2016). However, the relationship between personality and performance may vary across tasks (Carere and Locurto 2011; Sih and Del Giudice 2012; Griffin et al. 2015). For example, the relationship may be influenced by the associated levels of risk and reward (Sih and Del Giudice 2012). Proactive (bold, aggressive, or exploratory) individuals tend to accept a higher level of risk to get a reward, while reactive (shy, docile, or timid) individuals are more likely to “play it safe” (Sih and Del Giudice 2012). As a result, proactive animals are generally faster in learning how to obtain a reward than reactive individuals, while reactive individuals are faster to learn to avoid risks (Sih and Del Giudice 2012). Apart from associated levels of risk and reward, when trained to associate a certain stimulus with a reward, reactive individuals generally perform less well than proactive individuals when they must discriminate the rewarding stimuli from another, but are faster than proactive individuals to learn the reverse rule when the stimulus associated with the reward is reversed (Carere and Locurto 2011). For instance, proactive black-capped chickadees (Poecile atricapillus) perform better at discrimination tasks than reactive chickadees, but the reactive chickadees perform better in reversal learning tasks (Guillette et al. 2009; Guillette et al. 2011). As with learning, decision-making speed, accuracy and the relationship between speed and accuracy can be influenced by task difficulty (Gold and Shadlen 2007; Chittka et al. 2009). For instance, bumblebees (Bombus terrestris) trained to associate a reward with a certain flower color decide quickly and accurately when presented with two dissimilar flower colors, but their speed decreases significantly when deciding between two similarly colored flowers (Dyer and Chittka 2004). This speed and accuracy trade-off increases with task difficulty probably because it takes longer to acquire and process all the necessary information for an accurate decision, and so individuals make quick decisions with less than sufficient information (Gold and Shadlen 2007; Chittka et al. 2009; Mamuneas et al. 2015). For a simple task, the trade-off is less severe as less information is required. Given this, the relationship between an individual’s personality type and decision-making style may also vary with task difficulty. Research has shown proactive three-spined stickleback fish (Gasterosteus aculeatus), for example, make fast and accurate decisions in a simple T-shaped maze (Mamuneas et al. 2015), while proactive guppies (Poecilia reticulata) sacrifice accuracy for quick decision making in a more complicated maze set-up (Burns and Rodd 2008). Despite a significant body of research on the correlation between personality and decision-making style across taxa, the relationship between personality and decision-making style across tasks of varying difficulty has not yet been studied (Chittka et al. 2009; Mamuneas et al. 2015). Jumping spiders (Araneae: Salticidae) of the genus Portia make many decisions in the context of foraging, such as prey-type choice and route choice to reach prey, and they have multifaceted foraging strategies, regularly making complex or long-distance detours that include shifting away from the prey, breaking visual contact, and choosing between multiple routes (Jackson and Wilcox 1993; Tarsitano and Jackson 1994; Tarsitano and Jackson 1997; Tarsitano 2006). This is considered a cognitively challenging task and indicates that Portia forage in scenarios that likely vary significantly in difficulty. Therefore, Portia is an excellent animal model to study the relationship between personality and decision-making across various task difficulties. In our recent study on P. labiata, we looked solely at the relationship between consistent individual differences in aggressiveness and prey-choice discrimination decisions (Chang et al. 2017a). We found that more aggressive P. labiata made faster prey-choice discrimination decisions than less aggressive individuals without the cost of accuracy. However, this may have been due to the simplicity of the task (discriminating between very small and very large prey items). Here, we investigated the relationship between personality and decision-making in P. labiata, but across two tasks that differed in difficulty: a relatively simple task and a more difficult task. Both tasks involved an arena with two vertical poles—one pole was the correct route and led to the prey item, while the other pole (the dead-end pole) was the incorrect route (Figure 1). While in the simple task, the spider could move as it pleased, in the difficult task the spider started on a tower above the walkways leading to the poles, so that spider could not approach prey directly but had to climb down and move away from the prey before being able to approach it. In the simple task, P. labiata could see the prey and approach it with a fixed orientation. However, in the difficult task, P. labiata would break visual contact with prey and choose between two indirect paths that move away from the prey before proceeding toward it. This break in visual contact and choice between two routes (one correct and one incorrect) made the “difficult” task more cognitively demanding than the simple task. We predicted that more aggressive P. labiata would complete both the simple and difficult tasks faster than less aggressive P. labiata because aggressive spiders tend to make faster decisions, but that, in line with the speed-accuracy trade-off hypothesis, aggressive P. labiata would be more likely to make an incorrect decision than docile spiders in the difficult task, but not in the simple task. MATERIALS AND METHODS Study subjects and maintenance Adult P. labiata females (N = 38) were collected from Malaysia in December 2016. Spiders were housed individually in plastic containers (60 × 50 × 50 mm), in laboratory conditions at 25 ± 1 °C, 80 ± 5% RH, on a 12:12 h light:dark cycle and fed 3–5 fruit flies (Drosophila melanogaster) twice a week. Experiments were conducted between December 2016 and February 2017, during active daylight hours (0900 to 1700 h). Spiders were photographed using a digital SLR camera (NIKON D800) with a microlens (MICRO NIKKOR 105mm 1:2.8G ED) and close-up ring flash (SIGMA EM-140DG). The carapace width and body length of each P. labiata were then analyzed using ImageJ 1.51 software to the nearest 0.01 mm. Individual body mass was measured using an electronic balance to the nearest 0.01 mg. Aggressiveness assay The aggressiveness of each P. labiata individual was assessed using a mirror-image stimulation test following the procedure described in Chang et al. 2017b. After the lid of the housing container was removed, the housing container was placed horizontally in front of a mirror (140 × 65 mm). The test spider orientated toward the mirror image and performed an agonistic display. After a maximum of 20-min interaction, we measured the shortest distance reached between the test spider and the mirror. The shorter the distance, the more aggressive the spider. The actual distance from the mirror was subtracted from 50 mm (the maximum distance from the mirror in the container), so that higher values indicated greater levels of aggression. To ensure the consistency of the responses, each individual was tested 5 times with an interval of 1–3 days between trials, and the mean of the 5 trials was taken as that individual’s aggressiveness score. Decision-making assay The apparatus design of the simple task followed Tarsitano (2006), and the apparatus for the difficult task was modified from apparatus design of the simple task. The prey in each task was a dead salticid Cosmophasis umbratica (a natural prey item for P. labiata; Bulbert et al. 2015) mounted on the pole in lifelike posture (Cross and Jackson 2016). The C. umbratica individual (carapace width: 2.04 mm; body length: 6.92 mm) was killed with carbon dioxide and preserved in 70% ethanol. A high-speed video camera (CASIO Exilim EX-100) was placed approximately 700 mm away from the setup, elevated and to the side of the arena, for video-recoding. The apparatus for the simple task is shown in Figure 1a. The base was made of a 300 mm × 450 mm transparent plastic board stacked on a white foam board. The lure was mounted on the prey pole at the 50 mm mark. The prey pole was connected to a horizontal 125 mm rampway, and the rampway was elevated by an “access pole,” in one corner of the arena. Climbing the access pole and moving across the rampway was the only possible way to reach the prey pole and C. umbratica lure. A secondary, isolated “dead-end” pole was placed in the opposite corner of the same end of the platform. It gave no access to the prey. In attempting to reach the prey, the spider could not jump from one pole to the other pole, which means it had to choose between them. Figure 1 View largeDownload slide Diagrams of (a) the simple-task apparatus and (b) the difficult-task apparatus (not to scale). The difficult-task apparatus was surrounded by water to a depth of 30 mm in a large tray. The walkways were 50 mm above the water surface. The wood poles were 9 mm × 9 mm in cross section. The tower was 60 mm away from the wall of the enclosure. Figure 1 View largeDownload slide Diagrams of (a) the simple-task apparatus and (b) the difficult-task apparatus (not to scale). The difficult-task apparatus was surrounded by water to a depth of 30 mm in a large tray. The walkways were 50 mm above the water surface. The wood poles were 9 mm × 9 mm in cross section. The tower was 60 mm away from the wall of the enclosure. At the beginning of each test, the spider was placed in a plastic transparent petri dish (30 mm diameter, 10 mm deep), 150 mm from the prey pole. To prevent the spider from being distracted by the external surrounding, white plastic board (430 mm × 590 mm) was used to surround the 4 sides of the apparatus. The wooden poles (9 mm × 9 mm in cross section) were layered with a transparent plastic that was cleaned with 70% ethanol before each trial to minimize any chemical cues from previous test spiders influencing directional decisions. The difficult-task apparatus matched the design of the simple-task apparatus with the access pole, rampway, prey pole, and dead-end pole (Figure 1b). Again, only the access pole allowed the spider to reach the prey. There are 3 important differences from the simple-task apparatus: 1) elevated starting point, 2) the apparatus was surrounded by water which limited the spider’s movement to along the 2 walkways, and 3) the walkways extended in different directions providing indirect routes to their different destinations. The elevated starting point is the most important modification to increase the difficulty of this task, which requires the spider to turn away from the prey, break visual contact with the prey item, and climb down to walk on the walkways. In addition, the apparatus was placed in a tray (internal size: 620 mm × 330 mm, external size: 660 mm × 370 mm) filled with water to a depth of 30 mm and had only 2 walkways on which to walk. The walkways were 50 mm above the water surface, and spiders could walk along the sides of the walkway. The starting point was 230 mm away from the prey pole and 60 mm away from the wall of the enclosure nearest to the tower. Portia labiata are known to avoid contacting water (Cross and Jackson 2016), and therefore the set-up ensured the spider had to travel along a fixed walkway to reach the access pole or the dead-end pole. In the simple-task set-up, the test spiders could move without any restriction so long as they stayed on the platform, and thus they could maintain their directions the whole way to either the access pole or dead-end pole. However, in the difficult-task apparatus, the test spiders could not move freely or jump from the starting tower to the prey directly, but they had to turn away from the prey and then follow the walkway to approach the prey. The walkways did not provide a straight path to the prey, so the spiders could pause and reorientate to keep the prey in sight before continuing on the path. If spiders made a wrong decision once leaving from the starting point, they had to double back along the same path and they could climb up or climb around the tower. Spiders were not exposed to the apparatus prior to their trials and only one successful trial was conducted for each P. labiata individual for both tasks. Although the correct pole (having access pole, rampway, and prey pole) is structurally different from incorrect pole (dead-end pole), spiders were likely to be motivated to capture prey, and so their goals were likely to reach the prey item rather than reach the access pole. Besides C. umbratica is a natural prey for P. labiata and spiders are their preferred prey, each spider was starved for 4–7 days before the trial ensuring that they were likely to be highly motivated to capture prey. Spiders were first given 15 min to acclimatize at the starting point under the transparent petri dish. All spiders oriented and gazed at the prey item when they were under the petri dish, prior to the commencement of the trial. The trial commenced when the petri dish was removed, allowing the spider to move as it wished, and ended when P. labiata either: 1) reached an ending point (walked on access pole, dead-end pole, arena space under lure, or under rampway in the simple task; either access pole or dead-end pole in the difficult task); 2) walked or jumped out of the apparatus; 3) stayed motionless for 30 min; or 4) 2.5 h elapsed without scenarios 1–3 occurring. Spiders that failed to reach either the access pole or the dead-end pole were tested again on the following day, but spiders were only allowed to be retested once. For successful trials, scenario 1, once spiders reached either the access pole or the dead-end pole, they were removed from the apparatus and were not allowed to touch the prey lure. In the difficult task, only 3 spiders jumped on the wall of the enclosure next to the starting tower (60 mm distance) and left from the apparatus, and these 3 spiders were retested on the next day, while most spiders followed the walkway and oriented toward the prey during the course of trials. This suggests that in the difficult task spiders were still motivated for prey capture and not trying to escape from the environments surrounded by water. To control for directional bias, test spiders were assigned at random to either a setup with the access pole and rampway on the left side or on the right side of the apparatus. In between the trials the apparatus and petri-dish were wiped down with 70% ethanol to minimize chemical cues that could provide directional cues to the next spider. Decision accuracy was recorded as either correct (reaching the access pole) or incorrect (reaching the dead-end pole). We also recorded the time taken by the spiders to reach either pole and the number of directional changes en route. In the simple task, the directional change was when the spider paused, reoriented itself, and changed direction of movement. In the difficult task, the directional change was when a spider initially chose one side of the elevated walkway, but then turned back on itself and moved to the other side before reaching either pole. A total of 38 adult female P. labiata were tested in the simple task, and 36 adult females in the difficult task. Statistical analysis All analyses were performed in R 3.3.2 (R Core Team 2015). A linear mixed effect model, in the lme4 package (Bates et al. 2015), was used to quantify the consistency of the aggressiveness. Body size (carapace width) and trial number were included as fixed effects, and spider identity (ID) was coded as a random effect. Repeatability was calculated as the between-individual variance divided by the sum of the between-individual variance and the residual variance (Nakagawa and Schielzeth 2010). To determine whether there was significantly more between-individual difference than within-individual difference, we used a likelihood ratio test in the lmtest package (Zeileis and Hothorn 2002) to compare the full model to a model excluding spider ID. To test the effect of aggressiveness and task difficulty on decision accuracy (either access pole or dead-end pole) and the number of directional changes taken en route, we used generalized linear mixed effect models (lme4 package) with a binomial error structure for decision accuracy and a Poisson error structure for the number of directional changes. Six clear outliers (more than 2 standard deviations above the mean) were removed from the data set prior to analysis. The results of analysis with all data points are provided in the electronic supplementary materials (Supplementary Table S1 and Supplementary Figure S1). Furthermore, in the simple task, 8 individuals were excluded from the analysis as they did not make a clear decision between the access pole and the dead-end pole (either ending the trial underneath the lure, underneath the rampway, or jumping straight from the base to the prey pole). All individuals made clear decisions in the difficult task. We also used a linear mixed effect model (nlme, Pinheiro et al. 2017) to evaluate the effect of aggressiveness level and task difficulty on the time taken to reach either the access pole or dead-end pole. Carapace width, position of the access pole (right or left), task difficulty (simple or difficult), aggressiveness level, and the interaction between task difficulty and the aggressiveness were included as fixed effects for all models. Spider ID was also included as a random effect for all models. The fixed effect variables were centered on their means if the initial model was unable to converge. The significance of the explanatory variable was then tested using Wald Z-test, where the least significant variables were removed in a stepwise simplification. Ethics note All experiments on Portia labiata in this study comply with the current legal requirements of Singapore in which the research was conducted, and with all National University of Singapore guidelines (OSHM/PI/13/FOS-289). RESULTS Adult P. labiata females showed consistent interindividual differences in aggressiveness (ICC = 0.37, 95% CI = 0.25–0.45, P < 0.001). Neither carapace width (χ2 = 0.03, P = 0.87) nor trial number (χ2 = 3.07, P = 0.08) had a significant effect on the aggressiveness. In the simple task, 23 spiders reached access pole, and 7 spiders reached dead-end pole. Eight spiders did not make clear choice between access pole and dead-end pole, and they were excluded from the statistical analysis. Spiders spent 4.8 ± 1.19 min to reach either pole and changed direction 3.13 ± 0.72 times. In the difficult task, 3 spiders jumped on the enclosure wall from the starting tower, and they were retested on the following day. In total, 17 spiders reached access pole, and 19 spiders reached dead-end pole. Spiders spent 21.26 ± 3.07 min to reach either pole and changed direction 4.33 ± 1.00 times. Decision accuracy was determined by the interaction between aggressiveness and task difficulty (Table 1, Figure 2a). In the simple task, aggressive spiders more often made correct decisions than docile spiders, but in the difficult task docile spiders more often than aggressive spiders to make accurate decisions. Table 1 Results of generalized linear mixed effect model of the effect of aggressiveness, the position of the access pole (positive values indicate it was positioned on the right), carapace width, task difficulty (positive values indicate the difficult task), and the interaction between aggressiveness and task difficulty on the accuracy of the decision-making Predictor of accuracy Estimate SE Z P Intercept 1.22 0.46 2.68 0.007 Position of the access pole −0.06 0.58 −0.11 0.92 Carapace width −10.92 13.37 −0.82 0.41 Task difficulty 1.51 1.45 1.04 0.30 Aggressiveness 0.28 0.26 1.10 0.27 Task × Aggressiveness −0.43 0.17 −2.47 0.01 Predictor of accuracy Estimate SE Z P Intercept 1.22 0.46 2.68 0.007 Position of the access pole −0.06 0.58 −0.11 0.92 Carapace width −10.92 13.37 −0.82 0.41 Task difficulty 1.51 1.45 1.04 0.30 Aggressiveness 0.28 0.26 1.10 0.27 Task × Aggressiveness −0.43 0.17 −2.47 0.01 Significant effects are shown in bold. View Large Table 1 Results of generalized linear mixed effect model of the effect of aggressiveness, the position of the access pole (positive values indicate it was positioned on the right), carapace width, task difficulty (positive values indicate the difficult task), and the interaction between aggressiveness and task difficulty on the accuracy of the decision-making Predictor of accuracy Estimate SE Z P Intercept 1.22 0.46 2.68 0.007 Position of the access pole −0.06 0.58 −0.11 0.92 Carapace width −10.92 13.37 −0.82 0.41 Task difficulty 1.51 1.45 1.04 0.30 Aggressiveness 0.28 0.26 1.10 0.27 Task × Aggressiveness −0.43 0.17 −2.47 0.01 Predictor of accuracy Estimate SE Z P Intercept 1.22 0.46 2.68 0.007 Position of the access pole −0.06 0.58 −0.11 0.92 Carapace width −10.92 13.37 −0.82 0.41 Task difficulty 1.51 1.45 1.04 0.30 Aggressiveness 0.28 0.26 1.10 0.27 Task × Aggressiveness −0.43 0.17 −2.47 0.01 Significant effects are shown in bold. View Large Figure 2 View largeDownload slide The relationship between aggressiveness and (a) accuracy of decision-making, (b) the time taken to complete each task, and (c) the number of directional changes adult female Portia labiata made before completing each task (excluding outliers). Open circles and the dashed line represent the data and fitted line for the simple task. Solid circles and the solid line represent the data and fitted line for the difficult task. Figure 2 View largeDownload slide The relationship between aggressiveness and (a) accuracy of decision-making, (b) the time taken to complete each task, and (c) the number of directional changes adult female Portia labiata made before completing each task (excluding outliers). Open circles and the dashed line represent the data and fitted line for the simple task. Solid circles and the solid line represent the data and fitted line for the difficult task. Task difficulty, but not aggressiveness, predicted the time taken by the spider to complete the task (Table 2, Figure 2b), with all spiders taking longer to complete the difficult task than the simple task. On the contrary, aggressiveness predicted the number of directional changes, regardless of the difficulty of the task (Figure 2c). Aggressive P. labiata made fewer directional changes than docile P. labiata. The position of the access pole also influenced the number of directional changes. When the access pole was in right side of the apparatus, P. labiata tended to make fewer directional changes (Table 2). Table 2 Results of linear mixed effect model of the effect of aggressiveness, position of the access pole (positive values indicate it was positioned on the right), carapace width, task difficulty (positive values indicate the difficult task), and the interaction between aggressiveness and task difficulty on the time taken to complete the task and generalized linear mixed effect model for the number of directional changes made by the spiders (excluding outliers) Response variable Predictor Estimate SE t/Z P Time taken to complete task Intercept 287.63 156.77 1.83 0.08 Position of the access pole −336.29 212.72 −1.58 0.13 Carapace width 4500.94 4956.26 0.91 0.37 Task difficulty 987.95 212.27 4.65 0.0001 Aggressiveness −61.55 82.34 −0.75 0.46 Task × Aggressiveness 13.16 165.26 0.08 0.94 Number of directional changes Intercept 1.55 0.21 7.30 <0.0001 Position of the access pole −0.38 0.18 −2.10 0.04 Carapace width −3.78 3.94 −0.96 0.34 Task difficulty 0.04 0.18 0.25 0.80 Aggressiveness −0.20 0.07 −3.04 0.002 Task × Aggressiveness −0.10 0.13 −0.77 0.44 Response variable Predictor Estimate SE t/Z P Time taken to complete task Intercept 287.63 156.77 1.83 0.08 Position of the access pole −336.29 212.72 −1.58 0.13 Carapace width 4500.94 4956.26 0.91 0.37 Task difficulty 987.95 212.27 4.65 0.0001 Aggressiveness −61.55 82.34 −0.75 0.46 Task × Aggressiveness 13.16 165.26 0.08 0.94 Number of directional changes Intercept 1.55 0.21 7.30 <0.0001 Position of the access pole −0.38 0.18 −2.10 0.04 Carapace width −3.78 3.94 −0.96 0.34 Task difficulty 0.04 0.18 0.25 0.80 Aggressiveness −0.20 0.07 −3.04 0.002 Task × Aggressiveness −0.10 0.13 −0.77 0.44 Significant effects are shown in bold. View Large Table 2 Results of linear mixed effect model of the effect of aggressiveness, position of the access pole (positive values indicate it was positioned on the right), carapace width, task difficulty (positive values indicate the difficult task), and the interaction between aggressiveness and task difficulty on the time taken to complete the task and generalized linear mixed effect model for the number of directional changes made by the spiders (excluding outliers) Response variable Predictor Estimate SE t/Z P Time taken to complete task Intercept 287.63 156.77 1.83 0.08 Position of the access pole −336.29 212.72 −1.58 0.13 Carapace width 4500.94 4956.26 0.91 0.37 Task difficulty 987.95 212.27 4.65 0.0001 Aggressiveness −61.55 82.34 −0.75 0.46 Task × Aggressiveness 13.16 165.26 0.08 0.94 Number of directional changes Intercept 1.55 0.21 7.30 <0.0001 Position of the access pole −0.38 0.18 −2.10 0.04 Carapace width −3.78 3.94 −0.96 0.34 Task difficulty 0.04 0.18 0.25 0.80 Aggressiveness −0.20 0.07 −3.04 0.002 Task × Aggressiveness −0.10 0.13 −0.77 0.44 Response variable Predictor Estimate SE t/Z P Time taken to complete task Intercept 287.63 156.77 1.83 0.08 Position of the access pole −336.29 212.72 −1.58 0.13 Carapace width 4500.94 4956.26 0.91 0.37 Task difficulty 987.95 212.27 4.65 0.0001 Aggressiveness −61.55 82.34 −0.75 0.46 Task × Aggressiveness 13.16 165.26 0.08 0.94 Number of directional changes Intercept 1.55 0.21 7.30 <0.0001 Position of the access pole −0.38 0.18 −2.10 0.04 Carapace width −3.78 3.94 −0.96 0.34 Task difficulty 0.04 0.18 0.25 0.80 Aggressiveness −0.20 0.07 −3.04 0.002 Task × Aggressiveness −0.10 0.13 −0.77 0.44 Significant effects are shown in bold. View Large DISCUSSION Both task difficulty and aggressiveness jointly determine the accuracy of decision-making in P. labiata. More aggressive P. labiata perform simple tasks more accurately, whereas less aggressive spiders complete difficult tasks more accurately. This indicates that the relationship between personality and accuracy in the decision making is dependent on the difficulty of the task. Task difficulty was also the best predictor for the time taken to complete the task. In the simple task, spiders could move freely and were allowed to approach prey with a fixed direction. However, in the difficult task, spiders took longer to reach the prey, emphasizing how breaking visual contact and restricting spiders from direct approach increases the difficulty of the task. We did not find a relationship between aggressiveness and the time spiders took to complete either the simple or difficult task, despite the results of our previous study (Chang et al. 2017a), where aggressive spiders chose a more rewarding prey type in a shorter time than docile spiders. This is likely because the tasks in this study involved deciding on the best route (choosing which path leads toward the prey), rather than just the close-range direct attack required in the previous study. While aggressive spiders took less time than docile spiders in the attack aspects of prey capture tasks, in these new experiments, where prey capture involved route choice, spiders may adopt multiple secondary goals along the way (at the beginning of the walkway, after descending from the tower, for example), and after arriving at the secondary goal spiders may reorient toward the prey again before making the next decision (Hill 1979; Tarsitano 2006). That may be the reason why both aggressive and docile spiders took similar amount of time. As the correct pole leading to the prey differed structurally from the incorrect pole, a secondary question may be whether spiders were motivated by prey or by the 3-part structure of the correct pole; however, as test spiders were starved before the experiments and spiders are their preferred prey item, it is more likely that they were motivated by the prey. As the potential routes became more complex with breaking visual contact and restricting spiders from direct approach, the docile spiders showed greater accuracy than the aggressive spiders. Taken together, these studies suggest that there may be a trade-off between speed and accuracy in a multifaceted and difficult task and that the role personality plays in this trade-off is context dependent (Chittka et al. 2009). Though aggressive P. labiata did not take less time to reach the access or dead-end pole, they did make fewer directional changes in both tasks compared to docile P. labiata. This could be because docile individuals pause and change direction more often as they process the information they have gathered more slowly, while aggressive spiders change direction less as they may have processed that same information faster. However, our results do not appear to support this hypothesis as we would therefore expect to see increasing numbers of directional changes with increasing task difficulty, which we did not. Instead, it may be that aggressive individuals rely on less spatial information across a larger area while exploring, whereas docile individuals tend to gather more information in the environments with multiple locational cues per unit area (Sih and Del Giudice 2012). When spiders make directional changes, they may be exploring their environments and various potential routes to reach the prey. Docile spiders may explore the environments more thoroughly before making decisions, or they may tend to question the decisions that they already made and then collect more information to confirm the decisions. In the field, the viewing space of the prey may be hindered by many obstructions (Jackson and Wilcox 1993; Tarsitano and Jackson 1997), and scanning the environment from a stationary position is likely a less efficient method of gathering information than exploring the environment to decide on the best route. However, this exploratory behavior may not be essential because P. labiata is known to be able to execute planned routes correctly to reach the prey and without continuously orienting toward the prey along the way (Cross and Jackson 2016). Proactive animals tend to have better spatial cognition than reactive individuals (Sih and Del Giudice 2012), especially for active hunting animals that may be required to actively gather information from the environments. This may explain why the aggressive spiders showed higher accuracy than docile spiders in the simple task. However, in the difficult task, a better understanding of the potential routes is important to make an accurate decision. That is perhaps why the docile spiders perform more accurately (though at a slower speed with more detailed exploration) than aggressive spiders. Context-dependent fitness consequences may be one possible mechanism maintaining the between-individual variation in aggressiveness and decision-making speed in a population. For example, when fast-moving, nondangerous prey is most prevalent in the spider’s environment, aggressive individuals may benefit more from the speed of their decision-making in foraging decisions, as prey flee as soon as they detect the presence of predators. Fast decision-making predators may have a higher chance to capture the prey than slow decision makers. In a cognitive task with a low cost of making errors, the best strategy may be to “guess” the solution quickly to avoid taking too long to decide (Burns 2005; Kay et al. 2006; Burns and Dyer 2008). For example, it is known that fast-but-inaccurate decision-making in bumblebees (Bombus terrestris) results in the collection of more nectar than slow-but-accurate decision-making (Burns 2005). Conversely, when a harmful prey type is most prevalent, the cost of making an error may be high, such as death or injury (Chittka and Osorio 2007; Chittka et al. 2009). In this instance, foraging accuracy will matter more to success than speed, and so docile, slow decision-making spiders may have higher foraging success. Research on how personality is related to decision-making when the cost of making an error is high would shed new light on the fitness consequences of these phenotypes. Portia labiata potentially live and forage in different environments with varying topological complexities (Tarsitano and Jackson 1994). Some habitats may frequently require animals to break visual contact with the prey item while proceeding toward it. In these environments, the accuracy of decision-making would again play a large role in determining foraging success, and we might predict docile spiders will perform better than aggressive spiders. On the other hand, aggressive spiders with their fast decision-making may perform better at foraging in simpler environments than the docile spiders. It has been shown that environmental complexity plays an important role in predator-prey interactions for a salticid (Keiser et al. 2018), which supports the hypothesis that habitat-dependent foraging success may be key to maintain within-population variation in aggressiveness and decision-making style (Réale et al. 2007; Guillette et al. 2011). Habitat environmental condition may also influence the development of both personality type and decision-making style. For example, in the salticid Marpissa muscosa, individuals reared in more complex environments at the early stages tend to become more exploratory than those reared in simpler environments (Liedtke et al. 2015), suggesting that in complex environments it is necessary to acquire and process more information for all tasks, than in simpler environments. Therefore, spiders from different local habitats may differ in decision-making style because spiders from complex environmental habitats may tend to invest more in information gathering than spiders from simpler habitats, and as a result, make decisions more slowly. In summary, this study is the first to show for an invertebrate that the accuracy with which decisions are made is influenced by both personality and task difficulty. More aggressive spiders made fewer directional changes in both the simple and difficult tasks, but its potential cost (a reduction in accuracy) only became clear in the difficult task. In nature, animals face various tasks involving various costs of making incorrect or slow decisions, and whether it is better to focus on accuracy or speed depends on the decision-making context. Further research on how aggressive and docile individuals use and process information from their environments across contexts, and the fitness outcomes of these differences, will provide a clearer understanding of the processes which maintain behavioral diversity within populations. SUPPLEMENTARY MATERIAL Supplementary data are available at Behavioral Ecology online. FUNDING This study was supported by the Singapore Ministry of Education AcRF grants to D.L. (R-154-000-621-112 and R-154-000-638-112) and University of Malaya Research grant to Y.N-R (UMRG-RG379-17AFR). The authors thank Erick Yusuf Kencana for assisting in the spider collection and Poh Moi Goh for providing fruit flies for spider feeding. The authors also thank Robert R. Jackson and 1 anonymous referee for their insightful comments on the manuscript. Data accessibility: Analyses reported in this article can be reproduced using the data provided by Chang et al. (2018). REFERENCES Bates D , Machler M , Bolker B , Walker S . 2015 . Fitting linear mixed-effects models using lme4 . J Stat Softw . 67 : 1 – 48 . Google Scholar CrossRef Search ADS Bogacz R , Wagenmakers E-J , Forstmann BU , Nieuwenhuis S . 2010 . The neural basis of the speed–accuracy tradeoff . Trends Neurosci . 33 : 10 – 16 . Google Scholar CrossRef Search ADS PubMed Bulbert MW , O’Hanlon JC , Zappettini S , Zhang S , Li D . 2015 . Sexually selected UV signals in the tropical ornate jumping spider, Cosmophasis umbratica may incur costs from predation . Ecol Evol . 5 : 914 – 920 . Google Scholar CrossRef Search ADS PubMed Burns JG . 2005 . Impulsive bees forage better: the advantage of quick, sometimes inaccurate foraging decisions . Anim Behav . 70 : e1 – e5 . Google Scholar CrossRef Search ADS Burns JG , Dyer AG . 2008 . Diversity of speed-accuracy strategies benefits social insects . Curr Biol . 18 : R953 – R954 . Google Scholar CrossRef Search ADS PubMed Burns JG , Rodd FH . 2008 . Hastiness, brain size and predation regime affect the performance of wild guppies in a spatial memory task . Anim Behav . 76 : 911 – 922 . Google Scholar CrossRef Search ADS Carere C , Locurto C . 2011 . Interaction between animal personality and animal cognition . Curr Zool . 57 : 491 – 498 . Google Scholar CrossRef Search ADS Chang C-c , Lim ZY , Klomp DA , Norma-Rashid Y , Li D . 2018 . Data from: aggressive spiders make the wrong decision in a difficult task . Dryad Digital Repository . http://dx.doi.org/10.5061/dryad.0t20pq0 Chang C-c , Ng PJ , Li D . 2017a . Aggressive jumping spiders make quicker decisions for preferred prey but not at the cost of accuracy . Behav Ecol . 28 : 479 – 484 . Chang C-c , Teo HY , Norma-Rashid Y , Li D . 2017b . Predator personality and prey behavioural predictability jointly determine foraging performance . Sci Rep . 7 : 40734 . Google Scholar CrossRef Search ADS Chittka L , Osorio D . 2007 . Cognitive dimensions of predator responses to imperfect mimicry . PLoS Biol . 5 : e339 . Google Scholar CrossRef Search ADS PubMed Chittka L , Skorupski P , Raine NE . 2009 . Speed-accuracy tradeoffs in animal decision making . Trends Ecol Evol . 24 : 400 – 407 . Google Scholar CrossRef Search ADS PubMed Cole EF , Morand-Ferron J , Hinks AE , Quinn JL . 2012 . Cognitive ability influences reproductive life history variation in the wild . Curr Biol . 22 : 1808 – 1812 . Google Scholar CrossRef Search ADS PubMed Cross FR , Jackson RR . 2016 . The execution of planned detours by spider-eating predators . J Exp Anal Behav . 105 : 194 – 210 . Google Scholar CrossRef Search ADS PubMed Ducatez S , Audet JN , Lefebvre L . 2015 . Problem-solving and learning in Carib grackles: individuals show a consistent speed-accuracy trade-off . Anim Cogn . 18 : 485 – 496 . Google Scholar CrossRef Search ADS PubMed Dyer AG , Chittka L . 2004 . Bumblebees (Bombus terrestris) sacrifice foraging speed to solve difficult colour discrimination tasks . J Comp Physiol A Neuroethol Sens Neural Behav Physiol . 190 : 759 – 763 . Google Scholar PubMed Gold JI , Shadlen MN . 2007 . The neural basis of decision making . Annu Rev Neurosci . 30 : 535 – 574 . Google Scholar CrossRef Search ADS PubMed Griffin AS , Guillette LM , Healy SD . 2015 . Cognition and personality: an analysis of an emerging field . Trends Ecol Evol . 30 : 207 – 214 . Google Scholar CrossRef Search ADS PubMed Guillette LM , Reddon AR , Hoeschele M , Sturdy CB . 2011 . Sometimes slower is better: slow-exploring birds are more sensitive to changes in a vocal discrimination task . Proc Biol Sci . 278 : 767 – 773 . Google Scholar CrossRef Search ADS PubMed Guillette LM , Reddon AR , Hurd PL , Sturdy CB . 2009 . Exploration of a novel space is associated with individual differences in learning speed in black-capped chickadees, Poecile atricapillus . Behav Processes . 82 : 265 – 270 . Google Scholar CrossRef Search ADS PubMed Hill DE . 1979 . Orientation by jumping spiders of the genus Phidippus (Araneae: Salticidae) during the pursuit of prey . Behav Ecol Sociobiol . 5 : 301 – 322 . Google Scholar CrossRef Search ADS Ings TC , Chittka L . 2008 . Speed-accuracy tradeoffs and false alarms in bee responses to cryptic predators . Curr Biol . 18 : 1520 – 1524 . Google Scholar CrossRef Search ADS PubMed Jackson RR , Wilcox RS . 1993 . Observations in nature of detouring behaviour by Portia fimbriata, a web‐invading aggressive mimic jumping spider from Queensland . J Zool . 230 : 135 – 139 . Google Scholar CrossRef Search ADS Jandt JM , Bengston S , Pinter-Wollman N , Pruitt JN , Raine NE , Dornhaus A , Sih A . 2014 . Behavioural syndromes and social insects: personality at multiple levels . Biol Rev Camb Philos Soc . 89 : 48 – 67 . Google Scholar CrossRef Search ADS PubMed Kay LM , Beshel J , Martin C . 2006 . When good enough is best . Neuron . 51 : 277 – 278 . Google Scholar CrossRef Search ADS PubMed Keiser CN , Ingley SJ , Toscano BJ , Scharf I , Pruitt JN . 2018 . Habitat complexity dampens selection on prey activity level . Ethology . 124:25–32. Liedtke J , Redekop D , Schneider JM , Schuett W . 2015 . Early environmental conditions shape personality types in a jumping spider . Front Ecol Evol . 3 : 134 . Google Scholar CrossRef Search ADS Mamuneas D , Spence AJ , Manica A , King AJ . 2015 . Bolder stickleback fish make faster decisions, but they are not less accurate . Behav Ecol . 26 : 91 – 96 . Google Scholar CrossRef Search ADS Morand‐Ferron J , Cole EF , Quinn JL . 2016 . Studying the evolutionary ecology of cognition in the wild: a review of practical and conceptual challenges . Biol Rev . 91 : 367 – 389 . Google Scholar CrossRef Search ADS PubMed Nakagawa S , Schielzeth H . 2010 . Repeatability for Gaussian and non-Gaussian data: a practical guide for biologists . Biol Rev Camb Philos Soc . 85 : 935 – 956 . Google Scholar PubMed Nawroth C , Prentice PM , McElligott AG . 2017 . Individual personality differences in goats predict their performance in visual learning and non-associative cognitive tasks . Behav Processes . 134 : 43 – 53 . Google Scholar CrossRef Search ADS PubMed Pinheiro J , Bates D , DebRoy S , Sarkar D , R Core Team . 2017 . nlme: linear and nonlinear mixed effects models. R package version 3.1-131. Available from: https://CRAN.R-project.org/package=nlme . R Core Team . 2015 . R: a language and environment for statistical computing . Vienna (Austria) : R Foundation for Statistical Computing . Raine NE , Chittka L . 2008 . The correlation of learning speed and natural foraging success in bumble-bees . Proc Biol Sci . 275 : 803 – 808 . Google Scholar CrossRef Search ADS PubMed Réale D , Reader SM , Sol D , McDougall PT , Dingemanse NJ . 2007 . Integrating animal temperament within ecology and evolution . Biol Rev . 82 : 291 – 318 . Google Scholar CrossRef Search ADS PubMed Sih A , Bell A , Johnson JC . 2004 . Behavioral syndromes: an ecological and evolutionary overview . Trends Ecol Evol . 19 : 372 – 378 . Google Scholar CrossRef Search ADS PubMed Sih A , Del Giudice M . 2012 . Linking behavioural syndromes and cognition: a behavioural ecology perspective . Philos Trans R Soc Lond B Biol Sci . 367 : 2762 – 2772 . Google Scholar CrossRef Search ADS PubMed Tarsitano M . 2006 . Route selection by a jumping spider (Portia labiata) during the locomotory phase of a detour . Anim Behav . 72 : 1437 – 1442 . Google Scholar CrossRef Search ADS Tarsitano M , Jackson RR . 1994 . Jumping spiders make predatory detours requiring movement away from prey . Behaviour . 131 : 65 – 73 . Google Scholar CrossRef Search ADS Tarsitano MS , Jackson RR . 1997 . Araneophagic jumping spiders discriminate between detour routes that do and do not lead to prey . Anim Behav . 53 : 257 – 266 . Google Scholar CrossRef Search ADS Wang M-Y , Brennan CH , Lachlan RF , Chittka L . 2015 . Speed–accuracy trade-offs and individually consistent decision making by individuals and dyads of zebrafish in a colour discrimination task . Anim Behav . 103 : 277 – 283 . Google Scholar CrossRef Search ADS Zeileis A , Hothorn T . 2002 . Diagnostic checking in regression relationships . R News . 2 : 7 – 10 . © The Author(s) 2018. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Behavioral Ecology Oxford University Press

Aggressive spiders make the wrong decision in a difficult task

Loading next page...
 
/lp/ou_press/aggressive-spiders-make-the-wrong-decision-in-a-difficult-task-DYWwaaljJk
Publisher
Oxford University Press
Copyright
© The Author(s) 2018. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com
ISSN
1045-2249
eISSN
1465-7279
D.O.I.
10.1093/beheco/ary066
Publisher site
See Article on Publisher Site

Abstract

Abstract Accurate and timely decisions are critical for foraging, predator avoidance, and reproductive success. However, there is often a trade-off between speed and accuracy in decision-making, where individuals that make decisions more quickly make more mistakes. An individual’s personality may influence its decision-making style (i.e., whether it errs more in the speed or accuracy of a decision) and this relationship may change depending on contexts. Despite growing research on invertebrate personality, how personality correlates with decision-making style is still largely unknown and little research has assessed these relationships across tasks of varying difficulty. Here, we test the relationship between aggressiveness and decision-making style in Portia labiata, a specialized spider-eating jumping spider, in both a simple and a difficult task. We found that aggressive spiders made fewer directional changes before completing the tasks, regardless of task difficulty. However, decision accuracy was jointly determined by both aggressiveness and task difficulty. Aggressive spiders made more accurate decisions in the simple task, while docile spiders made more accurate decisions in the difficult task. We conclude that the relationship between personality and decision-making style in P. labiata is context dependent. We also discuss how the association between aggressiveness and decision-making style may serve important functions in maintaining behavioral variation in a natural population. INTRODUCTION The speed and accuracy of decision-making can be critical in the context of foraging, predator avoidance, and reproductive success (Ings and Chittka 2008; Raine and Chittka 2008; Chittka et al. 2009; Cole et al. 2012; Morand‐Ferron et al. 2016). Making an accurate decision requires sufficient information, but processing this information takes time, and most decisions need to be made quickly for the decision-maker to benefit (Bogacz et al. 2010). Therefore, there is the potential for a trade-off between these attributes, where an individual may make either a fast, error-prone decision or a slow, accurate decision (Chittka et al. 2009; Bogacz et al. 2010). Individuals are generally consistent in decision-making styles (i.e., whether they more often err on the speed side of the trade-off or the accuracy side; Sih and Del Giudice 2012). Personality describes consistent interindividual differences in behavior in a population. An individual’s personality is known to influence its decision-making, but whether and how personality influences decision-making across contexts has rarely been studied explicitly, especially in invertebrates (Sih et al. 2004; Carere and Locurto 2011; Jandt et al. 2014; Ducatez et al. 2015; Wang et al. 2015; Nawroth et al. 2017). Personality can influence individual cognitive task performance (Sih and Del Giudice 2012; Griffin et al. 2015; Morand‐Ferron et al. 2016). However, the relationship between personality and performance may vary across tasks (Carere and Locurto 2011; Sih and Del Giudice 2012; Griffin et al. 2015). For example, the relationship may be influenced by the associated levels of risk and reward (Sih and Del Giudice 2012). Proactive (bold, aggressive, or exploratory) individuals tend to accept a higher level of risk to get a reward, while reactive (shy, docile, or timid) individuals are more likely to “play it safe” (Sih and Del Giudice 2012). As a result, proactive animals are generally faster in learning how to obtain a reward than reactive individuals, while reactive individuals are faster to learn to avoid risks (Sih and Del Giudice 2012). Apart from associated levels of risk and reward, when trained to associate a certain stimulus with a reward, reactive individuals generally perform less well than proactive individuals when they must discriminate the rewarding stimuli from another, but are faster than proactive individuals to learn the reverse rule when the stimulus associated with the reward is reversed (Carere and Locurto 2011). For instance, proactive black-capped chickadees (Poecile atricapillus) perform better at discrimination tasks than reactive chickadees, but the reactive chickadees perform better in reversal learning tasks (Guillette et al. 2009; Guillette et al. 2011). As with learning, decision-making speed, accuracy and the relationship between speed and accuracy can be influenced by task difficulty (Gold and Shadlen 2007; Chittka et al. 2009). For instance, bumblebees (Bombus terrestris) trained to associate a reward with a certain flower color decide quickly and accurately when presented with two dissimilar flower colors, but their speed decreases significantly when deciding between two similarly colored flowers (Dyer and Chittka 2004). This speed and accuracy trade-off increases with task difficulty probably because it takes longer to acquire and process all the necessary information for an accurate decision, and so individuals make quick decisions with less than sufficient information (Gold and Shadlen 2007; Chittka et al. 2009; Mamuneas et al. 2015). For a simple task, the trade-off is less severe as less information is required. Given this, the relationship between an individual’s personality type and decision-making style may also vary with task difficulty. Research has shown proactive three-spined stickleback fish (Gasterosteus aculeatus), for example, make fast and accurate decisions in a simple T-shaped maze (Mamuneas et al. 2015), while proactive guppies (Poecilia reticulata) sacrifice accuracy for quick decision making in a more complicated maze set-up (Burns and Rodd 2008). Despite a significant body of research on the correlation between personality and decision-making style across taxa, the relationship between personality and decision-making style across tasks of varying difficulty has not yet been studied (Chittka et al. 2009; Mamuneas et al. 2015). Jumping spiders (Araneae: Salticidae) of the genus Portia make many decisions in the context of foraging, such as prey-type choice and route choice to reach prey, and they have multifaceted foraging strategies, regularly making complex or long-distance detours that include shifting away from the prey, breaking visual contact, and choosing between multiple routes (Jackson and Wilcox 1993; Tarsitano and Jackson 1994; Tarsitano and Jackson 1997; Tarsitano 2006). This is considered a cognitively challenging task and indicates that Portia forage in scenarios that likely vary significantly in difficulty. Therefore, Portia is an excellent animal model to study the relationship between personality and decision-making across various task difficulties. In our recent study on P. labiata, we looked solely at the relationship between consistent individual differences in aggressiveness and prey-choice discrimination decisions (Chang et al. 2017a). We found that more aggressive P. labiata made faster prey-choice discrimination decisions than less aggressive individuals without the cost of accuracy. However, this may have been due to the simplicity of the task (discriminating between very small and very large prey items). Here, we investigated the relationship between personality and decision-making in P. labiata, but across two tasks that differed in difficulty: a relatively simple task and a more difficult task. Both tasks involved an arena with two vertical poles—one pole was the correct route and led to the prey item, while the other pole (the dead-end pole) was the incorrect route (Figure 1). While in the simple task, the spider could move as it pleased, in the difficult task the spider started on a tower above the walkways leading to the poles, so that spider could not approach prey directly but had to climb down and move away from the prey before being able to approach it. In the simple task, P. labiata could see the prey and approach it with a fixed orientation. However, in the difficult task, P. labiata would break visual contact with prey and choose between two indirect paths that move away from the prey before proceeding toward it. This break in visual contact and choice between two routes (one correct and one incorrect) made the “difficult” task more cognitively demanding than the simple task. We predicted that more aggressive P. labiata would complete both the simple and difficult tasks faster than less aggressive P. labiata because aggressive spiders tend to make faster decisions, but that, in line with the speed-accuracy trade-off hypothesis, aggressive P. labiata would be more likely to make an incorrect decision than docile spiders in the difficult task, but not in the simple task. MATERIALS AND METHODS Study subjects and maintenance Adult P. labiata females (N = 38) were collected from Malaysia in December 2016. Spiders were housed individually in plastic containers (60 × 50 × 50 mm), in laboratory conditions at 25 ± 1 °C, 80 ± 5% RH, on a 12:12 h light:dark cycle and fed 3–5 fruit flies (Drosophila melanogaster) twice a week. Experiments were conducted between December 2016 and February 2017, during active daylight hours (0900 to 1700 h). Spiders were photographed using a digital SLR camera (NIKON D800) with a microlens (MICRO NIKKOR 105mm 1:2.8G ED) and close-up ring flash (SIGMA EM-140DG). The carapace width and body length of each P. labiata were then analyzed using ImageJ 1.51 software to the nearest 0.01 mm. Individual body mass was measured using an electronic balance to the nearest 0.01 mg. Aggressiveness assay The aggressiveness of each P. labiata individual was assessed using a mirror-image stimulation test following the procedure described in Chang et al. 2017b. After the lid of the housing container was removed, the housing container was placed horizontally in front of a mirror (140 × 65 mm). The test spider orientated toward the mirror image and performed an agonistic display. After a maximum of 20-min interaction, we measured the shortest distance reached between the test spider and the mirror. The shorter the distance, the more aggressive the spider. The actual distance from the mirror was subtracted from 50 mm (the maximum distance from the mirror in the container), so that higher values indicated greater levels of aggression. To ensure the consistency of the responses, each individual was tested 5 times with an interval of 1–3 days between trials, and the mean of the 5 trials was taken as that individual’s aggressiveness score. Decision-making assay The apparatus design of the simple task followed Tarsitano (2006), and the apparatus for the difficult task was modified from apparatus design of the simple task. The prey in each task was a dead salticid Cosmophasis umbratica (a natural prey item for P. labiata; Bulbert et al. 2015) mounted on the pole in lifelike posture (Cross and Jackson 2016). The C. umbratica individual (carapace width: 2.04 mm; body length: 6.92 mm) was killed with carbon dioxide and preserved in 70% ethanol. A high-speed video camera (CASIO Exilim EX-100) was placed approximately 700 mm away from the setup, elevated and to the side of the arena, for video-recoding. The apparatus for the simple task is shown in Figure 1a. The base was made of a 300 mm × 450 mm transparent plastic board stacked on a white foam board. The lure was mounted on the prey pole at the 50 mm mark. The prey pole was connected to a horizontal 125 mm rampway, and the rampway was elevated by an “access pole,” in one corner of the arena. Climbing the access pole and moving across the rampway was the only possible way to reach the prey pole and C. umbratica lure. A secondary, isolated “dead-end” pole was placed in the opposite corner of the same end of the platform. It gave no access to the prey. In attempting to reach the prey, the spider could not jump from one pole to the other pole, which means it had to choose between them. Figure 1 View largeDownload slide Diagrams of (a) the simple-task apparatus and (b) the difficult-task apparatus (not to scale). The difficult-task apparatus was surrounded by water to a depth of 30 mm in a large tray. The walkways were 50 mm above the water surface. The wood poles were 9 mm × 9 mm in cross section. The tower was 60 mm away from the wall of the enclosure. Figure 1 View largeDownload slide Diagrams of (a) the simple-task apparatus and (b) the difficult-task apparatus (not to scale). The difficult-task apparatus was surrounded by water to a depth of 30 mm in a large tray. The walkways were 50 mm above the water surface. The wood poles were 9 mm × 9 mm in cross section. The tower was 60 mm away from the wall of the enclosure. At the beginning of each test, the spider was placed in a plastic transparent petri dish (30 mm diameter, 10 mm deep), 150 mm from the prey pole. To prevent the spider from being distracted by the external surrounding, white plastic board (430 mm × 590 mm) was used to surround the 4 sides of the apparatus. The wooden poles (9 mm × 9 mm in cross section) were layered with a transparent plastic that was cleaned with 70% ethanol before each trial to minimize any chemical cues from previous test spiders influencing directional decisions. The difficult-task apparatus matched the design of the simple-task apparatus with the access pole, rampway, prey pole, and dead-end pole (Figure 1b). Again, only the access pole allowed the spider to reach the prey. There are 3 important differences from the simple-task apparatus: 1) elevated starting point, 2) the apparatus was surrounded by water which limited the spider’s movement to along the 2 walkways, and 3) the walkways extended in different directions providing indirect routes to their different destinations. The elevated starting point is the most important modification to increase the difficulty of this task, which requires the spider to turn away from the prey, break visual contact with the prey item, and climb down to walk on the walkways. In addition, the apparatus was placed in a tray (internal size: 620 mm × 330 mm, external size: 660 mm × 370 mm) filled with water to a depth of 30 mm and had only 2 walkways on which to walk. The walkways were 50 mm above the water surface, and spiders could walk along the sides of the walkway. The starting point was 230 mm away from the prey pole and 60 mm away from the wall of the enclosure nearest to the tower. Portia labiata are known to avoid contacting water (Cross and Jackson 2016), and therefore the set-up ensured the spider had to travel along a fixed walkway to reach the access pole or the dead-end pole. In the simple-task set-up, the test spiders could move without any restriction so long as they stayed on the platform, and thus they could maintain their directions the whole way to either the access pole or dead-end pole. However, in the difficult-task apparatus, the test spiders could not move freely or jump from the starting tower to the prey directly, but they had to turn away from the prey and then follow the walkway to approach the prey. The walkways did not provide a straight path to the prey, so the spiders could pause and reorientate to keep the prey in sight before continuing on the path. If spiders made a wrong decision once leaving from the starting point, they had to double back along the same path and they could climb up or climb around the tower. Spiders were not exposed to the apparatus prior to their trials and only one successful trial was conducted for each P. labiata individual for both tasks. Although the correct pole (having access pole, rampway, and prey pole) is structurally different from incorrect pole (dead-end pole), spiders were likely to be motivated to capture prey, and so their goals were likely to reach the prey item rather than reach the access pole. Besides C. umbratica is a natural prey for P. labiata and spiders are their preferred prey, each spider was starved for 4–7 days before the trial ensuring that they were likely to be highly motivated to capture prey. Spiders were first given 15 min to acclimatize at the starting point under the transparent petri dish. All spiders oriented and gazed at the prey item when they were under the petri dish, prior to the commencement of the trial. The trial commenced when the petri dish was removed, allowing the spider to move as it wished, and ended when P. labiata either: 1) reached an ending point (walked on access pole, dead-end pole, arena space under lure, or under rampway in the simple task; either access pole or dead-end pole in the difficult task); 2) walked or jumped out of the apparatus; 3) stayed motionless for 30 min; or 4) 2.5 h elapsed without scenarios 1–3 occurring. Spiders that failed to reach either the access pole or the dead-end pole were tested again on the following day, but spiders were only allowed to be retested once. For successful trials, scenario 1, once spiders reached either the access pole or the dead-end pole, they were removed from the apparatus and were not allowed to touch the prey lure. In the difficult task, only 3 spiders jumped on the wall of the enclosure next to the starting tower (60 mm distance) and left from the apparatus, and these 3 spiders were retested on the next day, while most spiders followed the walkway and oriented toward the prey during the course of trials. This suggests that in the difficult task spiders were still motivated for prey capture and not trying to escape from the environments surrounded by water. To control for directional bias, test spiders were assigned at random to either a setup with the access pole and rampway on the left side or on the right side of the apparatus. In between the trials the apparatus and petri-dish were wiped down with 70% ethanol to minimize chemical cues that could provide directional cues to the next spider. Decision accuracy was recorded as either correct (reaching the access pole) or incorrect (reaching the dead-end pole). We also recorded the time taken by the spiders to reach either pole and the number of directional changes en route. In the simple task, the directional change was when the spider paused, reoriented itself, and changed direction of movement. In the difficult task, the directional change was when a spider initially chose one side of the elevated walkway, but then turned back on itself and moved to the other side before reaching either pole. A total of 38 adult female P. labiata were tested in the simple task, and 36 adult females in the difficult task. Statistical analysis All analyses were performed in R 3.3.2 (R Core Team 2015). A linear mixed effect model, in the lme4 package (Bates et al. 2015), was used to quantify the consistency of the aggressiveness. Body size (carapace width) and trial number were included as fixed effects, and spider identity (ID) was coded as a random effect. Repeatability was calculated as the between-individual variance divided by the sum of the between-individual variance and the residual variance (Nakagawa and Schielzeth 2010). To determine whether there was significantly more between-individual difference than within-individual difference, we used a likelihood ratio test in the lmtest package (Zeileis and Hothorn 2002) to compare the full model to a model excluding spider ID. To test the effect of aggressiveness and task difficulty on decision accuracy (either access pole or dead-end pole) and the number of directional changes taken en route, we used generalized linear mixed effect models (lme4 package) with a binomial error structure for decision accuracy and a Poisson error structure for the number of directional changes. Six clear outliers (more than 2 standard deviations above the mean) were removed from the data set prior to analysis. The results of analysis with all data points are provided in the electronic supplementary materials (Supplementary Table S1 and Supplementary Figure S1). Furthermore, in the simple task, 8 individuals were excluded from the analysis as they did not make a clear decision between the access pole and the dead-end pole (either ending the trial underneath the lure, underneath the rampway, or jumping straight from the base to the prey pole). All individuals made clear decisions in the difficult task. We also used a linear mixed effect model (nlme, Pinheiro et al. 2017) to evaluate the effect of aggressiveness level and task difficulty on the time taken to reach either the access pole or dead-end pole. Carapace width, position of the access pole (right or left), task difficulty (simple or difficult), aggressiveness level, and the interaction between task difficulty and the aggressiveness were included as fixed effects for all models. Spider ID was also included as a random effect for all models. The fixed effect variables were centered on their means if the initial model was unable to converge. The significance of the explanatory variable was then tested using Wald Z-test, where the least significant variables were removed in a stepwise simplification. Ethics note All experiments on Portia labiata in this study comply with the current legal requirements of Singapore in which the research was conducted, and with all National University of Singapore guidelines (OSHM/PI/13/FOS-289). RESULTS Adult P. labiata females showed consistent interindividual differences in aggressiveness (ICC = 0.37, 95% CI = 0.25–0.45, P < 0.001). Neither carapace width (χ2 = 0.03, P = 0.87) nor trial number (χ2 = 3.07, P = 0.08) had a significant effect on the aggressiveness. In the simple task, 23 spiders reached access pole, and 7 spiders reached dead-end pole. Eight spiders did not make clear choice between access pole and dead-end pole, and they were excluded from the statistical analysis. Spiders spent 4.8 ± 1.19 min to reach either pole and changed direction 3.13 ± 0.72 times. In the difficult task, 3 spiders jumped on the enclosure wall from the starting tower, and they were retested on the following day. In total, 17 spiders reached access pole, and 19 spiders reached dead-end pole. Spiders spent 21.26 ± 3.07 min to reach either pole and changed direction 4.33 ± 1.00 times. Decision accuracy was determined by the interaction between aggressiveness and task difficulty (Table 1, Figure 2a). In the simple task, aggressive spiders more often made correct decisions than docile spiders, but in the difficult task docile spiders more often than aggressive spiders to make accurate decisions. Table 1 Results of generalized linear mixed effect model of the effect of aggressiveness, the position of the access pole (positive values indicate it was positioned on the right), carapace width, task difficulty (positive values indicate the difficult task), and the interaction between aggressiveness and task difficulty on the accuracy of the decision-making Predictor of accuracy Estimate SE Z P Intercept 1.22 0.46 2.68 0.007 Position of the access pole −0.06 0.58 −0.11 0.92 Carapace width −10.92 13.37 −0.82 0.41 Task difficulty 1.51 1.45 1.04 0.30 Aggressiveness 0.28 0.26 1.10 0.27 Task × Aggressiveness −0.43 0.17 −2.47 0.01 Predictor of accuracy Estimate SE Z P Intercept 1.22 0.46 2.68 0.007 Position of the access pole −0.06 0.58 −0.11 0.92 Carapace width −10.92 13.37 −0.82 0.41 Task difficulty 1.51 1.45 1.04 0.30 Aggressiveness 0.28 0.26 1.10 0.27 Task × Aggressiveness −0.43 0.17 −2.47 0.01 Significant effects are shown in bold. View Large Table 1 Results of generalized linear mixed effect model of the effect of aggressiveness, the position of the access pole (positive values indicate it was positioned on the right), carapace width, task difficulty (positive values indicate the difficult task), and the interaction between aggressiveness and task difficulty on the accuracy of the decision-making Predictor of accuracy Estimate SE Z P Intercept 1.22 0.46 2.68 0.007 Position of the access pole −0.06 0.58 −0.11 0.92 Carapace width −10.92 13.37 −0.82 0.41 Task difficulty 1.51 1.45 1.04 0.30 Aggressiveness 0.28 0.26 1.10 0.27 Task × Aggressiveness −0.43 0.17 −2.47 0.01 Predictor of accuracy Estimate SE Z P Intercept 1.22 0.46 2.68 0.007 Position of the access pole −0.06 0.58 −0.11 0.92 Carapace width −10.92 13.37 −0.82 0.41 Task difficulty 1.51 1.45 1.04 0.30 Aggressiveness 0.28 0.26 1.10 0.27 Task × Aggressiveness −0.43 0.17 −2.47 0.01 Significant effects are shown in bold. View Large Figure 2 View largeDownload slide The relationship between aggressiveness and (a) accuracy of decision-making, (b) the time taken to complete each task, and (c) the number of directional changes adult female Portia labiata made before completing each task (excluding outliers). Open circles and the dashed line represent the data and fitted line for the simple task. Solid circles and the solid line represent the data and fitted line for the difficult task. Figure 2 View largeDownload slide The relationship between aggressiveness and (a) accuracy of decision-making, (b) the time taken to complete each task, and (c) the number of directional changes adult female Portia labiata made before completing each task (excluding outliers). Open circles and the dashed line represent the data and fitted line for the simple task. Solid circles and the solid line represent the data and fitted line for the difficult task. Task difficulty, but not aggressiveness, predicted the time taken by the spider to complete the task (Table 2, Figure 2b), with all spiders taking longer to complete the difficult task than the simple task. On the contrary, aggressiveness predicted the number of directional changes, regardless of the difficulty of the task (Figure 2c). Aggressive P. labiata made fewer directional changes than docile P. labiata. The position of the access pole also influenced the number of directional changes. When the access pole was in right side of the apparatus, P. labiata tended to make fewer directional changes (Table 2). Table 2 Results of linear mixed effect model of the effect of aggressiveness, position of the access pole (positive values indicate it was positioned on the right), carapace width, task difficulty (positive values indicate the difficult task), and the interaction between aggressiveness and task difficulty on the time taken to complete the task and generalized linear mixed effect model for the number of directional changes made by the spiders (excluding outliers) Response variable Predictor Estimate SE t/Z P Time taken to complete task Intercept 287.63 156.77 1.83 0.08 Position of the access pole −336.29 212.72 −1.58 0.13 Carapace width 4500.94 4956.26 0.91 0.37 Task difficulty 987.95 212.27 4.65 0.0001 Aggressiveness −61.55 82.34 −0.75 0.46 Task × Aggressiveness 13.16 165.26 0.08 0.94 Number of directional changes Intercept 1.55 0.21 7.30 <0.0001 Position of the access pole −0.38 0.18 −2.10 0.04 Carapace width −3.78 3.94 −0.96 0.34 Task difficulty 0.04 0.18 0.25 0.80 Aggressiveness −0.20 0.07 −3.04 0.002 Task × Aggressiveness −0.10 0.13 −0.77 0.44 Response variable Predictor Estimate SE t/Z P Time taken to complete task Intercept 287.63 156.77 1.83 0.08 Position of the access pole −336.29 212.72 −1.58 0.13 Carapace width 4500.94 4956.26 0.91 0.37 Task difficulty 987.95 212.27 4.65 0.0001 Aggressiveness −61.55 82.34 −0.75 0.46 Task × Aggressiveness 13.16 165.26 0.08 0.94 Number of directional changes Intercept 1.55 0.21 7.30 <0.0001 Position of the access pole −0.38 0.18 −2.10 0.04 Carapace width −3.78 3.94 −0.96 0.34 Task difficulty 0.04 0.18 0.25 0.80 Aggressiveness −0.20 0.07 −3.04 0.002 Task × Aggressiveness −0.10 0.13 −0.77 0.44 Significant effects are shown in bold. View Large Table 2 Results of linear mixed effect model of the effect of aggressiveness, position of the access pole (positive values indicate it was positioned on the right), carapace width, task difficulty (positive values indicate the difficult task), and the interaction between aggressiveness and task difficulty on the time taken to complete the task and generalized linear mixed effect model for the number of directional changes made by the spiders (excluding outliers) Response variable Predictor Estimate SE t/Z P Time taken to complete task Intercept 287.63 156.77 1.83 0.08 Position of the access pole −336.29 212.72 −1.58 0.13 Carapace width 4500.94 4956.26 0.91 0.37 Task difficulty 987.95 212.27 4.65 0.0001 Aggressiveness −61.55 82.34 −0.75 0.46 Task × Aggressiveness 13.16 165.26 0.08 0.94 Number of directional changes Intercept 1.55 0.21 7.30 <0.0001 Position of the access pole −0.38 0.18 −2.10 0.04 Carapace width −3.78 3.94 −0.96 0.34 Task difficulty 0.04 0.18 0.25 0.80 Aggressiveness −0.20 0.07 −3.04 0.002 Task × Aggressiveness −0.10 0.13 −0.77 0.44 Response variable Predictor Estimate SE t/Z P Time taken to complete task Intercept 287.63 156.77 1.83 0.08 Position of the access pole −336.29 212.72 −1.58 0.13 Carapace width 4500.94 4956.26 0.91 0.37 Task difficulty 987.95 212.27 4.65 0.0001 Aggressiveness −61.55 82.34 −0.75 0.46 Task × Aggressiveness 13.16 165.26 0.08 0.94 Number of directional changes Intercept 1.55 0.21 7.30 <0.0001 Position of the access pole −0.38 0.18 −2.10 0.04 Carapace width −3.78 3.94 −0.96 0.34 Task difficulty 0.04 0.18 0.25 0.80 Aggressiveness −0.20 0.07 −3.04 0.002 Task × Aggressiveness −0.10 0.13 −0.77 0.44 Significant effects are shown in bold. View Large DISCUSSION Both task difficulty and aggressiveness jointly determine the accuracy of decision-making in P. labiata. More aggressive P. labiata perform simple tasks more accurately, whereas less aggressive spiders complete difficult tasks more accurately. This indicates that the relationship between personality and accuracy in the decision making is dependent on the difficulty of the task. Task difficulty was also the best predictor for the time taken to complete the task. In the simple task, spiders could move freely and were allowed to approach prey with a fixed direction. However, in the difficult task, spiders took longer to reach the prey, emphasizing how breaking visual contact and restricting spiders from direct approach increases the difficulty of the task. We did not find a relationship between aggressiveness and the time spiders took to complete either the simple or difficult task, despite the results of our previous study (Chang et al. 2017a), where aggressive spiders chose a more rewarding prey type in a shorter time than docile spiders. This is likely because the tasks in this study involved deciding on the best route (choosing which path leads toward the prey), rather than just the close-range direct attack required in the previous study. While aggressive spiders took less time than docile spiders in the attack aspects of prey capture tasks, in these new experiments, where prey capture involved route choice, spiders may adopt multiple secondary goals along the way (at the beginning of the walkway, after descending from the tower, for example), and after arriving at the secondary goal spiders may reorient toward the prey again before making the next decision (Hill 1979; Tarsitano 2006). That may be the reason why both aggressive and docile spiders took similar amount of time. As the correct pole leading to the prey differed structurally from the incorrect pole, a secondary question may be whether spiders were motivated by prey or by the 3-part structure of the correct pole; however, as test spiders were starved before the experiments and spiders are their preferred prey item, it is more likely that they were motivated by the prey. As the potential routes became more complex with breaking visual contact and restricting spiders from direct approach, the docile spiders showed greater accuracy than the aggressive spiders. Taken together, these studies suggest that there may be a trade-off between speed and accuracy in a multifaceted and difficult task and that the role personality plays in this trade-off is context dependent (Chittka et al. 2009). Though aggressive P. labiata did not take less time to reach the access or dead-end pole, they did make fewer directional changes in both tasks compared to docile P. labiata. This could be because docile individuals pause and change direction more often as they process the information they have gathered more slowly, while aggressive spiders change direction less as they may have processed that same information faster. However, our results do not appear to support this hypothesis as we would therefore expect to see increasing numbers of directional changes with increasing task difficulty, which we did not. Instead, it may be that aggressive individuals rely on less spatial information across a larger area while exploring, whereas docile individuals tend to gather more information in the environments with multiple locational cues per unit area (Sih and Del Giudice 2012). When spiders make directional changes, they may be exploring their environments and various potential routes to reach the prey. Docile spiders may explore the environments more thoroughly before making decisions, or they may tend to question the decisions that they already made and then collect more information to confirm the decisions. In the field, the viewing space of the prey may be hindered by many obstructions (Jackson and Wilcox 1993; Tarsitano and Jackson 1997), and scanning the environment from a stationary position is likely a less efficient method of gathering information than exploring the environment to decide on the best route. However, this exploratory behavior may not be essential because P. labiata is known to be able to execute planned routes correctly to reach the prey and without continuously orienting toward the prey along the way (Cross and Jackson 2016). Proactive animals tend to have better spatial cognition than reactive individuals (Sih and Del Giudice 2012), especially for active hunting animals that may be required to actively gather information from the environments. This may explain why the aggressive spiders showed higher accuracy than docile spiders in the simple task. However, in the difficult task, a better understanding of the potential routes is important to make an accurate decision. That is perhaps why the docile spiders perform more accurately (though at a slower speed with more detailed exploration) than aggressive spiders. Context-dependent fitness consequences may be one possible mechanism maintaining the between-individual variation in aggressiveness and decision-making speed in a population. For example, when fast-moving, nondangerous prey is most prevalent in the spider’s environment, aggressive individuals may benefit more from the speed of their decision-making in foraging decisions, as prey flee as soon as they detect the presence of predators. Fast decision-making predators may have a higher chance to capture the prey than slow decision makers. In a cognitive task with a low cost of making errors, the best strategy may be to “guess” the solution quickly to avoid taking too long to decide (Burns 2005; Kay et al. 2006; Burns and Dyer 2008). For example, it is known that fast-but-inaccurate decision-making in bumblebees (Bombus terrestris) results in the collection of more nectar than slow-but-accurate decision-making (Burns 2005). Conversely, when a harmful prey type is most prevalent, the cost of making an error may be high, such as death or injury (Chittka and Osorio 2007; Chittka et al. 2009). In this instance, foraging accuracy will matter more to success than speed, and so docile, slow decision-making spiders may have higher foraging success. Research on how personality is related to decision-making when the cost of making an error is high would shed new light on the fitness consequences of these phenotypes. Portia labiata potentially live and forage in different environments with varying topological complexities (Tarsitano and Jackson 1994). Some habitats may frequently require animals to break visual contact with the prey item while proceeding toward it. In these environments, the accuracy of decision-making would again play a large role in determining foraging success, and we might predict docile spiders will perform better than aggressive spiders. On the other hand, aggressive spiders with their fast decision-making may perform better at foraging in simpler environments than the docile spiders. It has been shown that environmental complexity plays an important role in predator-prey interactions for a salticid (Keiser et al. 2018), which supports the hypothesis that habitat-dependent foraging success may be key to maintain within-population variation in aggressiveness and decision-making style (Réale et al. 2007; Guillette et al. 2011). Habitat environmental condition may also influence the development of both personality type and decision-making style. For example, in the salticid Marpissa muscosa, individuals reared in more complex environments at the early stages tend to become more exploratory than those reared in simpler environments (Liedtke et al. 2015), suggesting that in complex environments it is necessary to acquire and process more information for all tasks, than in simpler environments. Therefore, spiders from different local habitats may differ in decision-making style because spiders from complex environmental habitats may tend to invest more in information gathering than spiders from simpler habitats, and as a result, make decisions more slowly. In summary, this study is the first to show for an invertebrate that the accuracy with which decisions are made is influenced by both personality and task difficulty. More aggressive spiders made fewer directional changes in both the simple and difficult tasks, but its potential cost (a reduction in accuracy) only became clear in the difficult task. In nature, animals face various tasks involving various costs of making incorrect or slow decisions, and whether it is better to focus on accuracy or speed depends on the decision-making context. Further research on how aggressive and docile individuals use and process information from their environments across contexts, and the fitness outcomes of these differences, will provide a clearer understanding of the processes which maintain behavioral diversity within populations. SUPPLEMENTARY MATERIAL Supplementary data are available at Behavioral Ecology online. FUNDING This study was supported by the Singapore Ministry of Education AcRF grants to D.L. (R-154-000-621-112 and R-154-000-638-112) and University of Malaya Research grant to Y.N-R (UMRG-RG379-17AFR). The authors thank Erick Yusuf Kencana for assisting in the spider collection and Poh Moi Goh for providing fruit flies for spider feeding. The authors also thank Robert R. Jackson and 1 anonymous referee for their insightful comments on the manuscript. Data accessibility: Analyses reported in this article can be reproduced using the data provided by Chang et al. (2018). REFERENCES Bates D , Machler M , Bolker B , Walker S . 2015 . Fitting linear mixed-effects models using lme4 . J Stat Softw . 67 : 1 – 48 . Google Scholar CrossRef Search ADS Bogacz R , Wagenmakers E-J , Forstmann BU , Nieuwenhuis S . 2010 . The neural basis of the speed–accuracy tradeoff . Trends Neurosci . 33 : 10 – 16 . Google Scholar CrossRef Search ADS PubMed Bulbert MW , O’Hanlon JC , Zappettini S , Zhang S , Li D . 2015 . Sexually selected UV signals in the tropical ornate jumping spider, Cosmophasis umbratica may incur costs from predation . Ecol Evol . 5 : 914 – 920 . Google Scholar CrossRef Search ADS PubMed Burns JG . 2005 . Impulsive bees forage better: the advantage of quick, sometimes inaccurate foraging decisions . Anim Behav . 70 : e1 – e5 . Google Scholar CrossRef Search ADS Burns JG , Dyer AG . 2008 . Diversity of speed-accuracy strategies benefits social insects . Curr Biol . 18 : R953 – R954 . Google Scholar CrossRef Search ADS PubMed Burns JG , Rodd FH . 2008 . Hastiness, brain size and predation regime affect the performance of wild guppies in a spatial memory task . Anim Behav . 76 : 911 – 922 . Google Scholar CrossRef Search ADS Carere C , Locurto C . 2011 . Interaction between animal personality and animal cognition . Curr Zool . 57 : 491 – 498 . Google Scholar CrossRef Search ADS Chang C-c , Lim ZY , Klomp DA , Norma-Rashid Y , Li D . 2018 . Data from: aggressive spiders make the wrong decision in a difficult task . Dryad Digital Repository . http://dx.doi.org/10.5061/dryad.0t20pq0 Chang C-c , Ng PJ , Li D . 2017a . Aggressive jumping spiders make quicker decisions for preferred prey but not at the cost of accuracy . Behav Ecol . 28 : 479 – 484 . Chang C-c , Teo HY , Norma-Rashid Y , Li D . 2017b . Predator personality and prey behavioural predictability jointly determine foraging performance . Sci Rep . 7 : 40734 . Google Scholar CrossRef Search ADS Chittka L , Osorio D . 2007 . Cognitive dimensions of predator responses to imperfect mimicry . PLoS Biol . 5 : e339 . Google Scholar CrossRef Search ADS PubMed Chittka L , Skorupski P , Raine NE . 2009 . Speed-accuracy tradeoffs in animal decision making . Trends Ecol Evol . 24 : 400 – 407 . Google Scholar CrossRef Search ADS PubMed Cole EF , Morand-Ferron J , Hinks AE , Quinn JL . 2012 . Cognitive ability influences reproductive life history variation in the wild . Curr Biol . 22 : 1808 – 1812 . Google Scholar CrossRef Search ADS PubMed Cross FR , Jackson RR . 2016 . The execution of planned detours by spider-eating predators . J Exp Anal Behav . 105 : 194 – 210 . Google Scholar CrossRef Search ADS PubMed Ducatez S , Audet JN , Lefebvre L . 2015 . Problem-solving and learning in Carib grackles: individuals show a consistent speed-accuracy trade-off . Anim Cogn . 18 : 485 – 496 . Google Scholar CrossRef Search ADS PubMed Dyer AG , Chittka L . 2004 . Bumblebees (Bombus terrestris) sacrifice foraging speed to solve difficult colour discrimination tasks . J Comp Physiol A Neuroethol Sens Neural Behav Physiol . 190 : 759 – 763 . Google Scholar PubMed Gold JI , Shadlen MN . 2007 . The neural basis of decision making . Annu Rev Neurosci . 30 : 535 – 574 . Google Scholar CrossRef Search ADS PubMed Griffin AS , Guillette LM , Healy SD . 2015 . Cognition and personality: an analysis of an emerging field . Trends Ecol Evol . 30 : 207 – 214 . Google Scholar CrossRef Search ADS PubMed Guillette LM , Reddon AR , Hoeschele M , Sturdy CB . 2011 . Sometimes slower is better: slow-exploring birds are more sensitive to changes in a vocal discrimination task . Proc Biol Sci . 278 : 767 – 773 . Google Scholar CrossRef Search ADS PubMed Guillette LM , Reddon AR , Hurd PL , Sturdy CB . 2009 . Exploration of a novel space is associated with individual differences in learning speed in black-capped chickadees, Poecile atricapillus . Behav Processes . 82 : 265 – 270 . Google Scholar CrossRef Search ADS PubMed Hill DE . 1979 . Orientation by jumping spiders of the genus Phidippus (Araneae: Salticidae) during the pursuit of prey . Behav Ecol Sociobiol . 5 : 301 – 322 . Google Scholar CrossRef Search ADS Ings TC , Chittka L . 2008 . Speed-accuracy tradeoffs and false alarms in bee responses to cryptic predators . Curr Biol . 18 : 1520 – 1524 . Google Scholar CrossRef Search ADS PubMed Jackson RR , Wilcox RS . 1993 . Observations in nature of detouring behaviour by Portia fimbriata, a web‐invading aggressive mimic jumping spider from Queensland . J Zool . 230 : 135 – 139 . Google Scholar CrossRef Search ADS Jandt JM , Bengston S , Pinter-Wollman N , Pruitt JN , Raine NE , Dornhaus A , Sih A . 2014 . Behavioural syndromes and social insects: personality at multiple levels . Biol Rev Camb Philos Soc . 89 : 48 – 67 . Google Scholar CrossRef Search ADS PubMed Kay LM , Beshel J , Martin C . 2006 . When good enough is best . Neuron . 51 : 277 – 278 . Google Scholar CrossRef Search ADS PubMed Keiser CN , Ingley SJ , Toscano BJ , Scharf I , Pruitt JN . 2018 . Habitat complexity dampens selection on prey activity level . Ethology . 124:25–32. Liedtke J , Redekop D , Schneider JM , Schuett W . 2015 . Early environmental conditions shape personality types in a jumping spider . Front Ecol Evol . 3 : 134 . Google Scholar CrossRef Search ADS Mamuneas D , Spence AJ , Manica A , King AJ . 2015 . Bolder stickleback fish make faster decisions, but they are not less accurate . Behav Ecol . 26 : 91 – 96 . Google Scholar CrossRef Search ADS Morand‐Ferron J , Cole EF , Quinn JL . 2016 . Studying the evolutionary ecology of cognition in the wild: a review of practical and conceptual challenges . Biol Rev . 91 : 367 – 389 . Google Scholar CrossRef Search ADS PubMed Nakagawa S , Schielzeth H . 2010 . Repeatability for Gaussian and non-Gaussian data: a practical guide for biologists . Biol Rev Camb Philos Soc . 85 : 935 – 956 . Google Scholar PubMed Nawroth C , Prentice PM , McElligott AG . 2017 . Individual personality differences in goats predict their performance in visual learning and non-associative cognitive tasks . Behav Processes . 134 : 43 – 53 . Google Scholar CrossRef Search ADS PubMed Pinheiro J , Bates D , DebRoy S , Sarkar D , R Core Team . 2017 . nlme: linear and nonlinear mixed effects models. R package version 3.1-131. Available from: https://CRAN.R-project.org/package=nlme . R Core Team . 2015 . R: a language and environment for statistical computing . Vienna (Austria) : R Foundation for Statistical Computing . Raine NE , Chittka L . 2008 . The correlation of learning speed and natural foraging success in bumble-bees . Proc Biol Sci . 275 : 803 – 808 . Google Scholar CrossRef Search ADS PubMed Réale D , Reader SM , Sol D , McDougall PT , Dingemanse NJ . 2007 . Integrating animal temperament within ecology and evolution . Biol Rev . 82 : 291 – 318 . Google Scholar CrossRef Search ADS PubMed Sih A , Bell A , Johnson JC . 2004 . Behavioral syndromes: an ecological and evolutionary overview . Trends Ecol Evol . 19 : 372 – 378 . Google Scholar CrossRef Search ADS PubMed Sih A , Del Giudice M . 2012 . Linking behavioural syndromes and cognition: a behavioural ecology perspective . Philos Trans R Soc Lond B Biol Sci . 367 : 2762 – 2772 . Google Scholar CrossRef Search ADS PubMed Tarsitano M . 2006 . Route selection by a jumping spider (Portia labiata) during the locomotory phase of a detour . Anim Behav . 72 : 1437 – 1442 . Google Scholar CrossRef Search ADS Tarsitano M , Jackson RR . 1994 . Jumping spiders make predatory detours requiring movement away from prey . Behaviour . 131 : 65 – 73 . Google Scholar CrossRef Search ADS Tarsitano MS , Jackson RR . 1997 . Araneophagic jumping spiders discriminate between detour routes that do and do not lead to prey . Anim Behav . 53 : 257 – 266 . Google Scholar CrossRef Search ADS Wang M-Y , Brennan CH , Lachlan RF , Chittka L . 2015 . Speed–accuracy trade-offs and individually consistent decision making by individuals and dyads of zebrafish in a colour discrimination task . Anim Behav . 103 : 277 – 283 . Google Scholar CrossRef Search ADS Zeileis A , Hothorn T . 2002 . Diagnostic checking in regression relationships . R News . 2 : 7 – 10 . © The Author(s) 2018. Published by Oxford University Press on behalf of the International Society for Behavioral Ecology. All rights reserved. For permissions, please e-mail: journals.permissions@oup.com This article is published and distributed under the terms of the Oxford University Press, Standard Journals Publication Model (https://academic.oup.com/journals/pages/about_us/legal/notices)

Journal

Behavioral EcologyOxford University Press

Published: May 21, 2018

There are no references for this article.

You’re reading a free preview. Subscribe to read the entire article.


DeepDyve is your
personal research library

It’s your single place to instantly
discover and read the research
that matters to you.

Enjoy affordable access to
over 18 million articles from more than
15,000 peer-reviewed journals.

All for just $49/month

Explore the DeepDyve Library

Search

Query the DeepDyve database, plus search all of PubMed and Google Scholar seamlessly

Organize

Save any article or search result from DeepDyve, PubMed, and Google Scholar... all in one place.

Access

Get unlimited, online access to over 18 million full-text articles from more than 15,000 scientific journals.

Your journals are on DeepDyve

Read from thousands of the leading scholarly journals from SpringerNature, Elsevier, Wiley-Blackwell, Oxford University Press and more.

All the latest content is available, no embargo periods.

See the journals in your area

DeepDyve

Freelancer

DeepDyve

Pro

Price

FREE

$49/month
$360/year

Save searches from
Google Scholar,
PubMed

Create lists to
organize your research

Export lists, citations

Read DeepDyve articles

Abstract access only

Unlimited access to over
18 million full-text articles

Print

20 pages / month

PDF Discount

20% off