sign in sign up
<<
Home , Mosquito

© 2020. Published by The Company of Biologists Ltd | Journal of Experimental (2020) 223, jeb221218. doi:10.1242/jeb.221218

RESEARCH ARTICLE Distinct navigation behaviors in , and mosquito larvae Eleanor K. Lutz, Kim T. Ha and Jeffrey A. Riffell*

ABSTRACT differences appear to be flexible and dependent upon the Mosquitoes spread deadly that impact millions of people environment (Yee et al., 2004). Despite growing interest (Skiff every . Understanding mosquito and behavior is vital and Yee, 2014; Reiskind and Shawn Janairo, 2018; Zahouli et al., for and prevention. However, many important 2017), the strategies larvae use to locate sources of food, or respond questions remain unanswered in the field of mosquito neuroethology, to environmental stimuli, including sources of nutrients or toxic particularly in our understanding of the larval stage. In this study, we chemicals, remain poorly understood across many disease investigate the innate exploration behavior of six different of species. A better understanding of larval navigation and foraging disease vector mosquito larvae. We show that these species exhibit behavior across mosquito species may help inform strikingly different movement paths, corresponding to a wide range of techniques by suggesting where, when and how much ingestible exploration behaviors. We also investigated the response of each to apply to maximize while minimizing species to an appetitive food cue, aversive cue or neutral control. cost and environmental impact. In contrast to the large differences in exploration behavior, all species Comparative approaches can provide important insights into the appeared to gather near preferred cues through random aggregation sensory and evolutionary bases of behavior. Despite the limited data rather than directed navigation, and exhibited slower speeds once on chemosensory behaviors by mosquito larvae, previous work has encountering food patches. Our results identify key behavioral demonstrated differences across species in locomotion and responses differences among important disease vector species, and suggest to food-related stimuli (Skiff and Yee, 2014), with these differences that navigation and exploration among even closely related mosquito attributed to the adaptive use of nutrient resources and impacts of species may be much more distinct than previously thought. competition. For instance, larvae of multiple species of Culex or Aedes may inhabit spatially restricted containers where individuals KEY WORDS: Chemosensory, Disease vector, Chemotaxis, experience interspecific competition for limited resources (Bevins, Olfaction, Gustation, Foraging 2008; Yee et al., 2004; Juliano, 2009; Workman and Walton, 2003). Although these species may exhibit differences in locomotory and INTRODUCTION feeding responses, such as suspension feeding or browsing on Mosquitoes are global disease vectors that transmit diseases such as surfaces, they can exhibit similar behaviors when experiencing , and . To limit the spread of these similar environmental resources, such as browsing on the surface of a disease vector mosquitoes, researchers have identified larval leaf (Yee et al., 2004). Larval resource use and competition have mosquito control as a highly effective public health tool (Weeks important effects on adult body size, and in the case of Aedes et al., 2019). In particular, naturally occurring such as albopictus, can increase arboviral infection. Characterizing larval methionine and Bacillus sp. bacteria have recently increased in behaviors under different chemosensory conditions may allow us to popularity as an environmentally safe alternative to synthetic determine how different species respond to different microhabitats or such as DDT (Weeks et al., 2019; Regis et al., 2000). competitive environments. These larvicides must be ingested by larvae to be effective, and Mosquito larvae may also provide important insight into the many factors affect larval feeding rate, including foraging strategy, algorithms associated with search behaviors by aquatic . A chemosensory preference and competition with conspecifics or previous study has shown that larvae find food individuals from other species (Hartman, 2016; Ramoska and randomly, rather than demonstrating directed motion toward Hopkins, 1981; Merritt et al., 1992). In addition, larval behaviors preferred cues. Once a food-rich area is located, larvae decrease and development rate also play an important role in adult population their swimming speed to remain in the favorable environment (Lutz levels. For instance, direct competition for limited food resources at et al., 2019). Previous studies in (Aly and the larval stage is thought to be a major factor in the presence of Mulla, 1986) and (Aly, 1985) larvae showed that these certain disease vectors (Bevins, 2008; Alto et al., 2005). Qualitative species also discover food at random, although the methods used in studies of mosquito larvae have shown different patterns of feeding these studies prevented deeper analysis into the mechanism of and swimming behaviors (Ramoska and Hopkins, 1981; Merritt navigation. This is an unusually simple foraging strategy rarely et al., 1992; Yee et al., 2004), although these inter-specific found in other insects, or even in adult mosquitoes (Takken and Knols, 2010). Do other disease vector mosquito species also forage by randomly encountering food cues? Department of Biology, University of Washington, Box 351800, Seattle, WA 98195, USA. In the present study, we investigated the foraging and navigation behavior of six species of mosquito larvae, drawn from the three *Author for correspondence ( [email protected]) major disease vector genera Aedes, Anopheles and Culex (Ruzzante E.K.L., 0000-0002-2434-2254; J.A.R., 0000-0002-7645-5779 et al., 2019) (Fig. 1A). These species were selected for their importance to public health and for their diversity in ecological

Received 4 January 2020; Accepted 25 February 2020 specialization and habitat choice. In the Aedes , we Journal of Experimental Biology

1 RESEARCH ARTICLE Journal of Experimental Biology (2020) 223, jeb221218. doi:10.1242/jeb.221218

A Anophelinae Anopheles Cellia Anopheles arabiensis Fig. 1. Summary of experimental design and methods. (A) A simplified phylogeny of Anopheles coluzzii the six mosquito species investigated in this study, adapted from Ruzzante et al. (2019) Culicidae Culex Culex (branch lengths are not representative of phylogenetic distance). (B) Diagram of the Culex tarsalis experimental rig, adapted from Bui et al. Aedes aegypti (2019), including a Basler Scout Machine Vision GigE camera (1), infrared lighting (2) Aedes Stegomyia and a behavior arena filled with water (3). (C) Map of stimulus distribution within the Subfamily Tribe Genus Subgenus Species arena at 15 min post-stimulus addition, adapted from Bui et al. (2019). B C Stimulus concentration (%) 100 80 60 40 20 0

1

2

3

investigated Ae. aegypti and Ae. albopictus – the most important guidelines (16 h:8 h light:dark for C. tarsalis; 12 h:12 h for all other vectors of dengue fever and . Although these two species). One day before the experiment, L3-stage larvae were species are closely related, Ae. aegypti preferentially breeds in isolated in Falcon™ 50 ml conical centrifuge tubes (Thermo Fisher manmade containers (Christophers, 1960), while Ae. albopictus is a Scientific, Waltham, MA, USA) containing ∼15 ml Milli-Q water. generalist that may also inhabit rural and forested areas (Mousson Larvae were denied food for at least 24 h before the experiment. et al., 2005). In the Anopheles genus, we examined the malaria that died before eclosion or pupated during the experiment vectors An. arabiensis and An. coluzzii. Interestingly, although An. were omitted, and all animals were tested during the light phase of arabiensis and An. coluzzii inhabitat similar human-associated their circadian . larval habitats (Minakawa et al., 1999; Etang et al., 2016), An. arabiensis drastically outcompetes An. coluzzii in mixed-species Preparation of odor stimuli larval competition (Hartman, 2016). This suggests that these closely Stimuli were used that elicited robust behavioral responses across related species may rely on different foraging strategies, and that species. Two stimuli were used: an attractive food solution and larval habitat specialization may not solely predict foraging quinine, a compound that elicits aversion in Ae. aegypti larvae (Lutz behavior. Finally, we investigated Culex quinquefasciatus,a et al., 2019). The food extract solution was made fresh daily by container-breeding mosquito, and Culex tarsalis, which breed in dissolving 0.5% food (Petco; Hikari Tropic First Bites) in Milli-Q large vegetative areas such as rice fields. These two Culex species water for 1 h, then passing the mixture through a 0.2 μm filter (VWR exhibit oviposition behavior that corresponds to risk in International, 28145-477) to remove solid particulates. Quinine their natural larval habitat (high risk and high predator avoidance for hydrochloride was prepared at 10 mmol l−1 in Milli-Q water C. tarsalis, low risk and low predator avoidance for container- (Sigma-Aldrich, Q1125). For all species, we saw no difference in dwelling C. quinquefasciatus) (Van Dam and Walton, 2008). Our mortality between the three treatments (P=1 for all species; Fig. S2), exploratory study reveals striking differences in exploration suggesting that exposure to quinine or food extract did not behavior between all six species. significantly harm larvae physiologically.

MATERIALS AND METHODS Behavior arena and imaging Mosquitoes We computed the trajectories of individual larvae in a custom Six species of wild-type mosquitoes were obtained from BEI behavior arena as previously described (Lutz et al., 2019; Bui et al., Resources (National Institute of and Infectious Diseases, 2019). Briefly, individual larvae were introduced to an 8×3 cm National Institutes of Health): Ae. aegypti (strain COSTA RICA, rectangular behaviorarenacontaining 20 ml of distilled Milli-Q water. MRA-726, contributed by William G. Brogdon), Ae. albopictus Larvae were allowed to acclimate within the dark arena for 15 min, (strain ATM-NJ95, Centers for Disease Control and Prevention while being recorded by a Basler Scout Machine Vision GigE camera for distribution by BEI, NR-48979), An. arabiensis (strain under infrared light. Subsequently, 100 µl of one stimulus was DONGOLA, MRA-856, contributed by Mark Q. Benedict), pipetted gently into the upper left corner of the arena (Fig. 1C), and An. coluzzii (strain Ngousso, MRA-1279, contributed by Frédéric larval behavior was recorded for an additional 15 min. In a separate Simard), C. quinquefasciatus (strain JHB, NR-43025) and experiment without larvae, we pipetted 100 µl of fluorescein dye into C. tarsalis (strain YOLO, NR-43026). All species were reared in an identically shaped arena, in order to map stimulus concentration Milli-Q water in a shallow tray (26×35×4 cm) and fed with fish food within the arena throughout the 15 min experiment (Lutz et al., 2019) (Petco; Hikari Tropic First Bites). Larvae were reared using the (Fig. 1C). Trajectory paths were extracted from each video using circadian cycle recommended by species-specific BEI rearing Multitracker software by van Breugel et al. (2018) and additional code Journal of Experimental Biology

2 RESEARCH ARTICLE Journal of Experimental Biology (2020) 223, jeb221218. doi:10.1242/jeb.221218 developed previously (Lutz et al., 2019). We visually inspected each RESULTS trajectory path and manually corrected errors and omissions Larval exploration behavior in clean water introduced by the tracking software. To study the navigation behavior of each mosquito species, we used a semi-automated video analysis method previously reported in Trajectory quantification Lutz et al. (2019) (Fig. 1B). We first investigated behavior in clean During foraging and swimming behaviors, mosquito can water during the 15 min acclimation period (Ae. aegypti n=67; Ae. exhibit species-specific differences in their swimming kinematics albopictus n=70; An. arabiensis n=93; An. coluzzii n=108; and behaviors, including changing the duration of activity (Sih, C. quinquefasciatus n=110; C. tarsalis n=53). The arena size 1986), increasing or decreasing their swim speeds, or exhibiting used in these experiments (3×8 cm) was chosen based on previous complex changes in locomotion (Merritt et al., 1992). We thus field research showing that 95% of Ae. aegypti and Ae. albopictus quantified 10 aspects of larval navigation in clean water to represent oviposition sites in the field were man-made containers, and that many of these ecologically relevant behaviors. Time spent moving more than a quarter of these observed oviposition sites were under was quantified as a percentage (0–100%), with movement defined 5 cm in radius (Chan et al., 1971), comparable to our experimental as >1 mm s−1. Total distance traveled was measured in meters. To arena. Similarly, An. gambiae larvae were most commonly found in normalize for any size-specific differences across individuals or very small pools such as hoofprints in a different field study species (Fig. S2), we converted larval speed measurements into (Minakawa et al., 2004). body lengths per second (BL s−1). Experimenters were blind to We observed striking behavioral differences across mosquito larval species or sex when measuring body lengths. Thus, the units species in many aspects of exploration behavior (Fig. 2A). For BL s−1 were used for quantifying maximum speed, mean speed example, C. tarsalis explored the environment slowly using when moving, mean speed in the first minute, and the difference in distinctive sweeping circles (mean number of spirals=13.1), while mean speed between first and last minutes. Spirals were defined as a the two Anopheles species interspersed long rests with fast, straight distinct time period in which larvae engaged in >4 s of continuous sprints (An. arabiensis mean number of spirals=0.2; An. coluzzii spiraling movement. Sharp turns were defined as turns of >45 deg 0.4). The two Aedes species, as well as C. tarsalis, spent the conducted at a speed of >4 mm s−1. Continuous paths were defined majority of the time moving (Ae. aegypti 56%; Ae. albopictus 52%; as sustained movement at the same Δ angle, not including spirals. C. tarsalis 46%), albeit at a much lower mean speed (Ae. aegypti Rests were defined as periods of time >10 s of no movement. 0.8 BL s−1; Ae. albopictus 0.7 BL s−1; C. tarsalis 0.7 BL s−1) compared with other species (mean speed 1.3–1.9 BL s−1; time Statistical analyses spent moving 9–35%). To further investigate these observations, we Statistical analyses were performed in R (https://www.r-project.org/) quantified 10 different aspects of each larval trajectory, based on and in Python (https://www.python.org/). We used a non-parametric metrics we believed to be relevant to foraging and exploration Kruskal–Wallis test with Bonferroni correction to compare behavior (Fig. 2B–K). We found significant differences across navigation characteristics across species for each of the 10 aspects species in all quantified measures (Fig. 2B–K; P<0.001, Kruskal– of larval navigation (Fig. 2), because we found that not all variables Wallis test with Holm–Bonferroni correction). followed a normal distribution (Shapiro–Wilk test, P>0.05). To create For example, some metrics measure the frequency of exploration the Euclidean distance matrix for larval similarity analysis (Fig. S1), behavior in starved larvae, such as time spent moving and total we first standardized all variables to zero mean and unit variance. distance traveled. Other metrics quantify known search behavior To compare larval trajectories across species, both as a group of patterns in insects, such as the looping spirals observed in local six and in species–species pairs, we used a perMANOVA and test search behavior (Bell, 1990), the frequency of sharp turns and the of multivariate dispersion ANOVA with a Bonferroni correction number of continuous straight-line paths. Some metrics were added (Table 1). We used a Monte Carlo permutation test to select to assess larval response to disturbance. For example, introducing significant eigenvectors for visualization in our PCA ordination animals to the arena during the acclimation phase is likely to elicit (Fig. 3). We used a pairwise t-test to compare larval preference for disturbance response to mechanical movement. Thus, we measured different stimuli for each species (Fig. 4A). Preference was defined as the mean speed of animals during the initial minute following the median concentration preferred by the larvae during the 15 min introduction to the arena, as well as throughout the entire 15 min experiment, normalized to the areas chosen by the same larva during acclimation period. We also calculated a metric subtracting the the preceding 15 min acclimation phase. This normalization was speed during the initial minute from the last (15th) minute, to necessary to control for innate larval preference for corners or walls quantify the change in larval behavior post-disturbance. Because we reported in some species (Lutz et al., 2019 preprint). Discovery time found significant differences in larval size across species (Fig. S2), across different stimuli were compared for each species using a non- we normalized all speed measurements to each individual’s body parametric Kruskal–Wallis test (Fig. 4B). A non-parametric test was length (BL s−1). Finally, some metrics were intended to measure the used because we found that discovery time data did not follow a physiological capacity of the starved larvae, such as the longest rest normal distribution (Shapiro–Wilk test, P>0.05). Discovery time was period and maximum observed speed. defined as the time taken (in seconds) to first encounter a section of the behavioral arena ≥50% concentration, normalized to the time Exploration differences among species taken to first encounter the same area during the clean water We next investigated whether these observed differences were acclimation period. We used a Fisher’s exact test with Bonferroni consistent with known phylogenetic relationships. In particular, do correction to assess mortality differences among larval species, as different species exhibit different trajectory patterns when all well as among experimental treatments in larvae of the same species navigation variables are considered? To answer this question, we (Fig. S2). We used a non-parametric Kruskal–Wallis test to compare created a Euclidean distance matrix of larval trajectories by body length between different species (Fig. S2), because we found incorporating all 10 navigation variables (Fig. S1). We found that body length data did not follow a normal distribution for all significant differences among the six mosquito species species (Shapiro–Wilk test, P>0.05). (perMANOVA P<0.001, pseudo-F=70.7; Fig. 3). To visualize Journal of Experimental Biology

3 RESEARCH ARTICLE Journal of Experimental Biology (2020) 223, jeb221218. doi:10.1242/jeb.221218

A Aedes aegypti Anopheles arabiensis Culex quinquefasciatus

Aedes albopictus Anopheles coluzzii Culex tarsalis

B C *** 10% moving *** 0.4 m 100 90% still 6

4 50 2

0 0 Time moving (%) Time Distance traveled (m) D E ) 0.1 BL s–1 ) 1.7 BL s–1 –1 *** *** 3 –1 24

2 16

1 8

0 0 Max. speed (BL s Max. speed (BL Mean speed (BL s Mean speed (BL F G ) )

0.1 BL s–1 –1 –0.1 BL s–1 –1 *** *** 3 3 Initial Last

2 0 1

0 -3 Initial speed (BL s Initial speed (BL Speed change (BL s Speed change (BL H I *** 0.7%*** 388 s 20 900

600 10 300

0 Longest rest (s) 0 Sharp turns (% time) K J *** 3 spirals *** 31 paths 50 3 300 1 200 25

Spirals 100

0 2 0 Continuous paths

Fig. 2. See next page for legend. these results, we reduced the dimensionality of this data using 10, P>0.05). In brief, we found that larvae appeared to cluster in principal component analysis (PCA) based on the same Euclidean ordination space into four distinct categories: (1) very fast animals distance matrix (Fig. 3A). PC1 and PC2 explained a significant (upper left); (2) animals that traveled a long distance and conducted proportion of variation in the data, but not subsequent PC axes many sharp turns (upper right); (3) animals that rested for long

(Monte Carlo permutation test, PC1, P<0.001; PC2, P<0.001; PC3– periods of time (lower left); and (4) animals that traveled in spiral Journal of Experimental Biology

4 RESEARCH ARTICLE Journal of Experimental Biology (2020) 223, jeb221218. doi:10.1242/jeb.221218

Fig. 2. Larvae exhibit species-specific behavioral differences in the corresponding larval behavior during the acclimation phase. Aedes absence of chemosensory stimuli. (A) Example trajectories visualizing aegypti preferred a median food concentration of 20% more than – one individual from each species, navigating in clean water. (B K) Species- would be expected from their pre-experiment behavior (P=0.0002, specific distributions for each of 10 navigation variables quantified from larval behavior in clean water. For each variable, a violin plot visualizes the pairwise t-test). We observed similar attraction for all other species distribution for each species, with scatter points showing values for individual (Fig. 4A): Ae. albopictus (+32%, P<0.001), An. arabiensis (+13%, animals. A black bar marks the mean value for each species. Asterisks above P=0.005), An. coluzzii (+7%, P=0.04), C. quinquefasciatus (+14%, each plot represent the significance of differences across species (Kruskal– P=0.006) and C. tarsalis (+16%, P=0.001). Further, we investigated Wallis test; ***P<0.001). For all plots, a sample trajectory from a C. changes in larval behavior after the introduction of 10 mmol l−1 quinquefasciatus individual is shown on the right in gray, with red and black quinine, an aversive tastant. Similar to our previous study (Lutz et al., lines highlighting sections of the trajectory included in that variable. From left to right in all graphs: Ae. aegypti (navy); Ae. albopictus (purple); An. arabiensis 2019), Ae. aegypti significantly avoided quinine, preferring a median (red); An. coluzzii (yellow); C. quinquefasciatus (green); C. tarsalis (aqua). concentration of 9% less than would be expected from their pre- (B) Time spent moving (%); (C) total distance traveled (m); (D) mean speed experiment behavior (P<0.001). We observed similar aversion in Ae. when moving [body lengths (BL) s−1]; (E) maximum speed (BL s−1); (F) initial albopictus (−11%, P<0.001), An. arabiensis (−7%, P=0.002) and speed, or mean speed in first minute (BL s−1); (G) speed modification, or An. coluzzii (−7%, P=0.001) (Fig. 4A). Interestingly, neither −1 difference in mean speed between first and last minutes (BL s ); (H) sharp C. quinquefasciatus nor C. tarsalis exhibited aversion to quinine turns (% total time spent turning >45 deg); (I) longest continuous rest period; (C. quinquefasciatus P=0.42; C. tarsalis P=0.60). In response to the (J) number of spirals; and (K) number of continuous paths that are not spirals. addition of distilled water – a negative control for mechanical disturbance – all species exhibited no change in preference (P>0.05). patterns and changed their behavior drastically during the 15 min We next explored the mechanism of larval navigation to food period (lower right). sources. In our previous work, we found that Ae. aegypti explore We next asked whether sister mosquito species display exploratory their environment using a non-directional search strategy that results behavior that is more similar to each other than to other species. in random discovery of preferred cues (Lutz et al., 2019). In all A post hoc pairwise perMANOVA for each species–species pair species in this study, we found that larval preference also appeared showed that all species differed significantly in navigation from each to be consistent with random cue discovery. Discovery time did not other, including sister species (Table 1). We observed that both Aedes significantly differ between water, food and quinine for any species species were more similar to each other than to any other species (P>0.05, Kruskal–Wallis test; Fig. 4B), suggesting that larvae (comparison of pseudo-F statistics across species–species pairs), and encounter environmental cues by random chance. In addition, we both Anopheles species were also closest to each other than to non- did not observe strong differences in orientation (Fig. 4C) or turn sister species. However, C. quinquefasciatus was most similar to frequency (Fig. 4D) between different stimuli for any of the six Ae. aegypti,whileC. tarsalis was most similar to Ae. albopictus.Itis species. Although surprising, these results are consistent with earlier interesting to note that C. quinquefasciatus and Ae. aegypti both literature using both video-tracking methods (Lutz et al., 2019) and inhabit man-made containers, while C. tarsalis and Ae. albopictus researcher observations (Aly and Mulla, 1986; Aly, 1985). can inhabit large vegetated areas such as rice fields and lakes. In our previous study, we were able to conduct deep analyses and Although our study only compares six species and is not intended to simulations into the mechanism of Ae. aegypti navigation, using a draw phylogenetic conclusions, our limited panel of results suggest dataset of over 500 individual animals observed independently, with that both evolutionary history and ecological specialization may approximately 2 million total data points. In the present study, we correlate with similar navigation behaviors in different species. did not have the necessary data to conduct the simulations or 3000 experiments necessary for similar analyses across species. Larval response to attractive and aversive cues Nevertheless, we visualized some of the same behavioral changes Next, we examined the change in larval behavior after the as a reference for future experiments (Figs S3, S4 and S5). We found introduction of 0.5% food extract, 10 mmol l−1 quinine or distilled several interesting patterns in these datasets. For example, in the water. These stimuli were chosen to investigate larval cue-finding vast majority of cases, animals did not appear to change their behavior, because previous studies have shown that these cues elicit behavior – such as the number of looped searches or sharp turns – robust preferences in Ae. aegypti mosquito larvae (Lutz et al., 2019; after addition of the stimulus. In cases where animals did change Bui et al., 2019). Corroborating a previous study (Lutz et al., 2019), their behavior – such as An. coluzzii, which decreased initial speed we found that Ae. aegypti significantly preferred 0.5% food extract. in the post-stimulus period – animals seemed to exhibit the same To quantify preference, we normalized the median concentration behavioral changes for all experiments, independently of the preferred by each larvae during the experiment phase, to the stimulus added. Corroborating our previous results, we found that Table 1. Comparisons of swimming patterns among species Aedes aegypti Aedes albopictus Anopheles arabiensis Anopheles coluzzii Culex quinquefasciatus Culex tarsalis Aedes aegypti – 11.12*** 89.92*** 115.66*** 25.66*** 20.13*** Aedes albopictus 5.54 – 99.66*** 118.58*** 38.73*** 16.96*** Anopheles arabiensis 0.76 0.03 – 33.10*** 71.41*** 48.50*** Anopheles coluzzii 1.44 6.79 18.19*** – 69.04*** 63.56*** Culex quinquefasciatus 4.90 18.77** 21.33*** 0.06 – 29.69*** Culex tarsalis 0.02 4.84 0.22 3.15 3.21 – Values in the upper right half of the matrix represent pseudo-F-statistics from pairwise perMANOVA tests. Asterisks after each value indicate the significance of the corresponding P-value after Bonferroni correction: **P<0.01; ***P<0.001. In the upper right, significant values with a high pseudo-F-statistic represent species–species pairs that exhibit statistically significant differences in overall navigation behavior. Values in the lower left half of the matrix represent F-statistics for a pairwise test of multivariate dispersion (ANOVA with Bonferroni correction). Significant values with a high F-statistic in the bottom half of the matrix represent species–species pairs with statistically significant differences in the intra-species variability of navigation behavior. These results suggest that some, but not all, of the observed differences between species–species pairs may be due to variance among individuals, rather than to differences in raw behavioral metrics. Journal of Experimental Biology

5 RESEARCH ARTICLE Journal of Experimental Biology (2020) 223, jeb221218. doi:10.1242/jeb.221218

A Aedes aegypti B Aedes albopictus 1 4: Mean speed 5: Initial speed Anopheles arabiensis Anopheles coluzzii Culex quinquefasciatus 8: Max. speed Culex tarsalis

4 0.5 2: Distance traveled

6: Sharp turns

PC1: 41.5% PC1: 41.5% 3: Continuous runs 0 0 9: Longest rest

1: Time moving

10: Spirals

–0.5 –4

7: Speed modification

–1

–6 –3 PC2: 33.7% 0 3 6 –1 –0.5PC2: 33.7% 0 0.5 1

Fig. 3. Larvae exhibit species-specific behavioral differences in the absence of chemosensory stimuli. (A) PCA biplot of 10 individual variables describing larval trajectories based on the Euclidean distance matrix visualized in Fig. S1. PC1 and PC2 explain a significant proportion of variance in the original data (Monte Carlo test, PC1: 41.5%, P<0.001; PC2: 33.7%, P<0.001), while all other PC axes did not (P>0.05). Scatter points depict individual larvae in the ordination space, colored by species: Ae. aegypti (navy circles, n=67), Ae. albopictus (purple × markers, n=70), An. arabiensis (red squares, n=93), An. coluzzii (yellow diamonds, n=108), C. quinquefasciatus (green triangles, n=110) and C. tarsalis (aqua + markers, n=53). (B) Vector arrows (1–10) indicate the direction of variable gradients in the ordination space. Ordered from highest to lowest contribution to PC1 and PC2, these variables include: (1) time spent moving (variable contribution to PC1: 21.4%; PC2: 0.8%); (2) distance traveled (PC1: 17.9%; PC2: 4.5%); (3) continuous paths (PC1: 21.5%; PC2: 0%); (4) mean speed (PC1: 0.4%; PC2: 24.5%); (5) initial speed in the first minute (PC1: 0.1%; PC2: 24.6%); (6) sharp turns (PC1: 15.8%; PC2: 3.2%); (7) speed modification, or the difference in mean speed between the first and last minutes (PC1: 2.2%; PC2: 19.4%); (8) maximum speed (PC1: 0.2%; PC2: 20.1%); (9) longest rest period (PC1: 16%; PC2: 0.1%); and (10) spirals (PC1: 4.6%; PC2: 2.8%). in our current work, Ae. aegypti also appear to aggregate near may exhibit different navigation strategies in larger environments preferred cues by decreasing their movement speed near preferred where the chemical gradients may be shallower or influenced by areas (Fig. S4). Interestingly, we observed the same pattern for Ae. turbulent kinetic motion. Additionally, are there physiological albopictus but not for any other species (Fig. S4). Similarities limitations to larval chemosensation, such as the sensitivity of between closely related species were reflected in the PCA analyses, receptors or complexity of neural processing circuits, that prevent whereas the exploration responses were often distinct between larvae from utilizing more complex navigation processes? species (Fig. 3A, Fig. S4K). However, once food was added, all the Second, is there an evolutionary benefit to different navigation species significantly slowed their swimming speeds, causing them behaviors exhibited by each species in clean water? Interspecific to cluster in the PCA space (Fig. S4N). Although further larval competition significantly affects distribution of mosquito experiments are necessary to understand these results, it is likely species in the wild and in laboratory experiments (Hartman, 2016; that the Anopheles and Culex species in our study use different Juliano, 1998; Braks et al., 2004). It is possible that this competitive kinematic changes to navigate with respect to chemosensory cues, environment drives larvae to exploit different foraging niches such as adjusting their turning frequency. through different navigation strategies. Alternatively, it is possible that the different environmental conditions preferred by each species, DISCUSSION such as lakes, streams and containers, result in different navigation Our results raise several interesting questions for future research. In strategies consistent across habitats. Although our study did not our experiments investigating larval cue-finding, we predicted that examine enough species to quantitatively answer this question, it is larvae may exhibit navigation strategies adapted to their interesting to note that Aedes and Anopheles mosquito larvae environment, with species living in small containers displaying exhibited greater similarity to sister species even when the sister different strategies than those that breed in larger lakes or streams. species inhabited vastly different natural larval habitats. By contrast, However, our results showed that the six different species of the two Culex species exhibited the greatest similarity toward mosquito larvae were strikingly homogeneous in their chemosensory non-sister species that inhabited similar ecological environments. responses to food: none of the species were able to change their It is important to note that our investigative study may not address behavior to find food cues faster in our experimental paradigm. Are important characteristics of mosquitoes found in thewild. Forexample, there intrinsic physical properties to chemical diffusion in small, although all larvae analyzed in this study successfully completed stagnant aquatic environments that makes more directed navigation development under the same laboratory rearing conditions, it is likely particularly difficult? Although many of the species examined are that environmental variables including temperature, concentration of naturally adapted to habitats of similar size to the experimental arena dissolved organic matter, and water depth were more optimal for some

(Skiff and Yee, 2014; Christophers, 1960), it is possible that larvae species than for others. Indeed, species exhibited significantly different Journal of Experimental Biology

6 RESEARCH ARTICLE Journal of Experimental Biology (2020) 223, jeb221218. doi:10.1242/jeb.221218

A Aedes aegypti Aedes albopictus Anopheles arabiensis Anopheles coluzzii Culex quinquefasciatus Culex tarsalis n.s. n.s. n.s. n.s. n.s.n.s. n.s. n.s. 80 *** *** *** *** ** ** ** * ** **

40

0

−40

−80 Water Quinine Food Water Quinine Food Water Quinine Food Water Quinine Food Water Quinine Food Water Quinine Food Preferred concentration (normalized)

B n.s. n.s. n.s. n.s. n.s. n.s. 15

10

5

0

–5

–10 Time to discover cue (min) Time

–15 Water Quinine Food Water Quinine Food Water Quinine Food Water Quinine Food Water Quinine Food Water Quinine Food C

S SSSSS

D 0.03

0.02

0.01 Frequency (%) 0 −180 0 180 −180 0 180 −180 0 180 −180 0 180 −180 0 180 −180 0 180 Water Food Quinine Change in angle (deg)

Fig. 4. Larval gathering near preferred cues is more consistent with passive exploration than with active navigation. (A) All larval species were significantly attracted to food extract. With the exception of the two Culex species, larvae were also significantly repelled by the aversive tastant quinine. (B) However, none of the six larval species were able to change their navigation behavior to find food faster, or delay occupying high-concentration areas of quinine: Ae. aegypti (P=0.60, Kruskal–Wallis test); Ae. albopictus (P=0.71); An. arabiensis (P=0.52); An. coluzzii (P=0.22); C. quinquefasciatus (P=0.64); C. tarsalis (P=0.93). (C) For all species, animals did not appear to change their behavior to move towards or away from any stimuli. Polar plots showing orientation for all animals throughout the 15 min experiment for water (gray), food (pink) or quinine (light blue). The label ‘S’ marks the approximate direction of each stimulus. Each bar in the polar plot represents the proportion of time that animals moved in that specific direction. For visualization purposes, all polar plots are normalized to the maximum frequency for each stimulus. (D) Turning frequency for each species did not appear to change depending on the added stimulus. Frequency histograms for exploratory larval turning frequency in response to the addition of water (gray), food (pink) or quinine (light blue). Y-axis ranges are identical for all six plots. To reduce noise from passively drifting animals, plots in C and D only include data where larvae were moving at ≥1mms−1. Although our experiments did not have sufficient power to statistically inspect these differences, animals do not appear to significantly change orientation across stimuli. mortality rates post-experiment, suggesting that the 24 h starvation to quantitatively compare exploration behavior among mosquito period may have been more stressful for some species (Fig. S2). larvae using machine vision rather than researcher observations. Even In addition, our experimental trials only observed larvae for a total among the small subset of species examined in this study, we saw of 15 min after stimulus addition, and it is possible that larvae may immediate and clear differences in exploration, stimulus preference exhibit different behaviors over longer time scales. and chemosensory navigation. Future studies incorporating additional Nevertheless, we believe that this study reveals an important mosquito species – especially outgroups that are not disease vectors – area of future research. To our knowledge, this is the first study would add fascinating comparisons that may help clarify the Journal of Experimental Biology

7 RESEARCH ARTICLE Journal of Experimental Biology (2020) 223, jeb221218. doi:10.1242/jeb.221218 evolutionary basis of exploration behavior in mosquitoes. systems for life-history-specific foraging strategies. BMC Neurosci. 20, 27. doi:10. Furthermore, with the advent of new tools to examine the sensory 1186/s12868-019-0511-y Chan, K. L., Ho, B. C. and Chan, Y. C. (1971). Aedes aegypti (L.) and Aedes bases of behaviors, including binary expression systems for cell-type albopictus (Skuse) in Singapore City. 2. Larval habitats. Bull. World Health Organ specific labeling and manipulation (Riabinina et al., 2016; Matthews 44, 629-633. and Vosshall, 2020), it may soon be possible to interrogate and Christophers, S. (1960). Aedes aegypti (L.) the Yellow Fever Mosquito: its Life History, Bionomics and Structure. Cambridge University Press. functionally link different navigational behaviors with the sensory Etang, J., Mbida, A. M., Akono, P. N., Binyang, J., Moukoko, C. E. E., Lehman, channels mediating those responses. Because laboratory arena sizes L. G., Awono-Ambene, P., Talipouo, A., Eyisab, W. E., Tagne, D. et al. (2016). and rearing conditions are already within the natural range of many Anopheles coluzzii larval habitat and resistance in the island area of Manoka, Cameroon. BMC Infect. Dis. 16, 217. doi:10.1186/s12879-016-1542-y mosquito species, mosquito larvae may be a promising field for Hartman, M. F. (2016). Malaria mosquito larvae in competition for limited resources. phylogenetic behavior questions from a practical perspective. Finally, https://oaktrust.library.tamu.edu/handle/1969.1/164484 our results underscore the importance of understanding disease vector Juliano, S. A. (1998). Species introduction and replacement among mosquitoes: behavior at all life history stages. We suggest that species-specific interspecific resource competition or apparent competition? Ecology 79, 255-268. doi:10.1890/0012-9658(1998)079[0255:SIARAM]2.0.CO;2 vector control research may be particularly important to improving Juliano, S. A. (2009). Species interactions among larval mosquitoes: context disease prevention methods. dependence across habitat gradients. Annu. Rev. Entomol. 54, 37-56. doi:10. 1146/annurev.ento.54.110807.090611 Acknowledgements Lutz, E. K., Grewal, T. S. and Riffell, J. A. (2019). Computational and experimental We thank Floris van Breugel for assistance in developing methods for video data insights into the chemosensory navigation of Aedes aegypti mosquito larvae. Proc. R. Soc. B 286, 20191495. doi:10.1098/rspb.2019.1495 analysis, and Julian Olden for advice on statistical methods. We also thank Binh Matthews, B. J. and Vosshall, L. B. (2020). How to turn an organism into a model Nguyen and Kara Kiyokawa for maintaining the Riffell lab mosquito colony, and organism in 10 ‘easy’ steps. J. Exp. Biol. 223, jeb218198. doi:10.1242/jeb.218198 Dustin Miller for advice on rearing mosquitoes procured from the Centers for Disease Merritt, R. W., Dadd, R. H. and Walker, E. D. (1992). Feeding behavior, natural Control and Prevention. Finally, we thank two anonymous peer reviewers for their food, and nutritional relationships of larval mosquitoes. Annu. Rev. Entomol. 37, time and effort in suggesting helpful revisions to the manuscript. 349-374. doi:10.1146/annurev.en.37.010192.002025 Minakawa, N., Beier, J. C., Githure, J. I., Mutero, C. M. and Yan, G. (1999). Spatial Competing interests distribution and habitat characterization of Anopheline mosquito larvae in Western The authors declare no competing or financial interests. Kenya. Am. J. Trop. Med. Hyg. 61, 1010-1016. doi:10.4269/ajtmh.1999.61.1010 Minakawa, N., Sonye, G., Mogi, M. and Yan, G. (2004). Habitat characteristics of Author contributions s.s. larvae in a Kenyan highland. Med. Vet. Entomol. 18, Conceptualization: E.K.L., J.A.R.; Methodology: E.K.L., K.T.H.; Software: E.K.L.; 301-305. doi:10.1111/j.0269-283X.2004.00503.x Mousson, L., Dauga, C., Garrigues, T., Schaffner, F., Vazeille, M. and Failloux, Validation: E.K.L., J.A.R.; Formal analysis: E.K.L.; Investigation: E.K.L., K.T.H., A.-B. (2005). Phylogeography of Aedes (Stegomyia) aegypti (L.) and Aedes J.A.R.; Resources: E.K.L., J.A.R.; Data curation: E.K.L.; Writing - original draft: (Stegomyia) albopictus (Skuse) (Diptera: Culicidae) based on mitochondrial DNA E.K.L., J.A.R.; Writing - review & editing: E.K.L., J.A.R.; Visualization: E.K.L.; variations. Genet. Res 86, 1-11. doi:10.1017/S0016672305007627 Supervision: J.A.R.; Project administration: J.A.R.; Funding acquisition: J.A.R. Ramoska, W. A. and Hopkins, T. L. (1981). Effects of mosquito larval feeding behavior on Bacillus sphaericus efficacy. J. Invertebr. Pathol. 37, 269-272. doi:10. Funding 1016/0022-2011(81)90086-0 This work was supported in part by National Institutes of Health grants Regis, L., da Silva, S. B. and Melo-Santos, M. A. (2000). The use of bacterial RO1DCO13693-04 and R21AI137947 to J.A.R.; National Science Foundation grant larvicides in mosquito and black control programmes in Brazil. Mem. Inst. DGE-1256082 to E.K.L.; Air Force Office of Scientific Research grant FA9550-16-1- Oswaldo Cruz. 95 Suppl. 1, 207-210. doi:10.1590/S0074-02762000000700035 0167 to J.A.R.; a CoMotion Innovations Scholarship to K.T.H.; a Robin Mariko Harris Reiskind, M. H. and Shawn Janairo, M. (2018). Tracking Aedes aegypti (Diptera: Award to E.K.L.; and the and Tom Wyckoff Award to E.K.L. Deposited in PMC Culicidae) larval behavior across development: effects of temperature and ’ for release after 12 months. nutrients on individuals foraging behavior. J. Med. Entomol. 7, e2239. doi:10. 1093/jme/tjy073 Riabinina, O., Task, D., Marr, E., Lin, C.-C., Alford, R., O’Brochta, D. A. and Data availability Potter, C. J. (2016). Organization of olfactory centres in the malaria mosquito All code used to analyze data in this paper is available from GitHub at: Anopheles gambiae. Nat. Commun. 7, 13010. doi:10.1038/ncomms13010 https://github.com/eleanorlutz/aedes-aegypti-2020 Ruzzante, L., Reijnders, M. J. M. F. and Waterhouse, R. M. (2019). Of genes and genomes: mosquito evolution and diversity. Trends Parasitol. 35, 32-51. doi:10. Supplementary information 1016/j.pt.2018.10.003 Supplementary information available online at Sih, A. (1986). Antipredator responses and the perception of danger by mosquito http://jeb.biologists.org/lookup/doi/10.1242/jeb.221218.supplemental larvae. Ecology 67, 434-441. doi:10.2307/1938587 Skiff, J. J. and Yee, D. A. (2014). Behavioral differences among four co-occurring References species of container mosquito larvae: effects of depth and resource Alto, B. W., Lounibos, L. P., Higgs, S. and Juliano, S. A. (2005). Larval environments. J. Med. Entomol. 51, 375-381. doi:10.1603/ME13159 Takken, W. and Knols, B. G. J. (2010). Olfaction in Vector– Interactions. competition differentially affects infection in Aedes mosquitoes. Ecology Wageningen Academic Pub. 86, 3279-3288. doi:10.1890/05-0209 van Breugel, F., Huda, A. and Dickinson, M. H. (2018). Distinct activity-gated Aly, C. (1985). Feeding rate of larval Aedes vexans stimulated by food substances. pathways mediate attraction and aversion to CO in . Nature 564, J. Am. Mosq. Control Assoc. 1, 506-510. 2 420-424. doi:10.1038/s41586-018-0732-8 Aly, C. and Mulla, S. (1986). Orientation and ingestion rates of larval Anopheles Van Dam, A. R. and Walton, W. E. (2008). The effect of predatory fish exudates on albimanus in response to floating particles. Entomol. Exp. Appl. 42, 83-90. doi:10. the ovipostional behaviour of three mosquito species: Culex quinquefasciatus, 1111/j.1570-7458.1986.tb02191.x Aedes aegypti and Culex tarsalis. Med. Vet. Entomol. 22, 399-404. doi:10.1111/j. Bell, W. J. (1990). Searching behavior patterns in insects. Annu. Rev. Entomol. 35, 1365-2915.2008.00764.x 447-467. doi:10.1146/annurev.en.35.010190.002311 Weeks, E. N., Baniszewski, J., Gezan, S. A., Allan, S. A., Cuda, J. P. and Bevins, S. N. (2008). Invasive mosquitoes, larval competition, and indirect effects Stevens, B. R. (2019). Methionine as a safe and effective novel biorational on the vector competence of native mosquito species (Diptera: Culicidae). Biol. mosquito larvicide. Manag. Sci. 75, 346-355. doi:10.1002/ps.5118 Invasions 10, 1109-1117. doi:10.1007/s10530-007-9188-8 Workman, P. D. and Walton, W. E. (2003). Larval behavior of four Culex (Diptera: ́ Braks, M. A. H., Honorio, N. A., Lounibos, L. P., Lourenço-De-Oliveira, R. and Culicidae) associated with treatment wetlands in the southwestern United States. Juliano, S. A. (2004). Interspecific competition between two of J. Vector Ecol. 28, 213-228. container mosquitoes, Aedes aegypti and Aedes albopictus (Diptera: Culicidae), Yee, D. A., Kesavaraju, B. and Juliano, S. A. (2004). Larval feeding behavior of in Brazil. Ann. Entomol. Soc. Am. 97, 130-139. doi:10.1603/0013-8746(2004)097 three co-occurring species of container mosquitoes. J. Vector Ecol. 29, 315. [0130:ICBTIS]2.0.CO;2 Zahouli, J. B. Z., Koudou, B. G., Müller, P., Malone, D., Tano, Y. and Utzinger, J. Bui, M., Shyong, J., Lutz, E. K., Yang, T., Li, M., Truong, K., Arvidson, R., (2017). Urbanization is a main driver for the larval ecology of Aedes mosquitoes in Buchman, A., Riffell, J. A. and Akbari, O. S. (2019). Live calcium imaging of arbovirus-endemic settings in south-eastern Côte d’Ivoire. PLoS Negl. Trop. Dis. Aedes aegypti neuronal tissues reveals differential importance of chemosensory 11, e0005751. doi:10.1371/journal.pntd.0005751 Journal of Experimental Biology

8