Post-biting behavioral reprogramming underlies reproductive efficiency in Aedes aegypti mosquitoes Authors: Linhan Dong, Elisabeth F Bradford, Jord M Barnett, Laura B Duvall Published: 2025 ABSTRACT SUMMARYThe global spread and increasing populations of disease vector mosquitoes expose hundreds of millions of people to mosquito-borne illnesses each year. Female Aedes aegypti mosquitoes, global vectors of dengue and Zika, bite humans to obtain blood nutrients that support egg development. We report that, after a blood meal, female mosquitoes transition from a period of inactivity into active searching for egg-laying sites. Females with mature eggs show increased locomotor activity and a shift in circadian behavioral timing, leading to nocturnal humidity seeking and egg laying in an otherwise diurnal species. We show that the circadian clock gene cycle is critical for regulating this transition; its absence disrupts the timing of oviposition-related behaviors, causing reduced reproductive capacity when active site search is required. These findings suggest that, during egg development, circadian-clock-dependent behavioral reprogramming triggers nocturnal hyperactivity and oviposition site search, an essential process for mosquito reproduction and population viability. INTRODUCTION To produce viable eggs, female Aedes aegypti mosquitoes must bite and obtain a blood meal from vertebrate hosts such as humans. This unique reproductive biology underlies their capacity to transmit disease-causing pathogens, including dengue, Zika, and chikungunya viruses.1–4 A single blood meal can double the female’s body weight, supplying the nutrients needed to develop dozens of eggs. This high reproductive output contributes to rapid mosquito population growth and global expansion.5–8 Non-blood-fed females exhibit a strong host-seeking drive and are highly responsive to human-associated cues, including carbon dioxide (CO2), human odor, and body heat, to locate and bite hosts.9,10 Aedes aegypti are primarily diurnal, with peak biting activity and CO2 response persistence at dawn and dusk and olfactory responses that vary throughout the day.11–15 The circadian clock underlies the timing of nearly all biological rhythms and is governed by conserved clock genes that generate autoregulatory transcription-translation feedback loops.16–20 Disruption of core clock genes, such as cycle, through mutation or RNAi, impairs circadian rhythmicity and alters mosquito behavioral patterns.21–24 After blood feeding to repletion, females undergo a dramatic behavioral shift: locomotor activity and responsiveness to host cues are suppressed for days during blood digestion and egg maturation.25–29 During this period of host-seeking suppression, sensitivity to host-associated cues is reduced, and as eggs finally reach maturity, the gravid female transitions from inactivity to oviposition site selection.30–33 Host-seeking suppression is reversible and typically ends after successful egg laying, occurring about 3 days post-blood meal, although it can be prolonged if suitable oviposition sites are unavailable.34–37 Previous studies in Aedes mosquitoes have reported variable peak oviposition times ranging from mid-afternoon to evening or night, but the role of the circadian clock in shaping these transitions remains untested.38–40 The search for an appropriate egg-laying site is crucial, as this location will serve as the habitat for the aquatic stages of a female’s offspring and will ultimately determine her reproductive success.41,42 Mosquito species exhibit niche-specific strategies when selecting egg-laying sites: some travel miles to reach natural freshwater reservoirs, while others specialize in locating artificial water containers near human dwellings or rain-filled objects in urban environments.35,43 Aedes aegypti mosquitoes specialize in seeking freshwater sources around human habitats for egg laying.44 Gravid females rely on a combination of visual, humidity, and odorant cues to locate egg-laying sites33,45–47 and use their legs and proboscis to evaluate salinity and oviposition suitability.36 Thus, following temporary post-blood-meal activity suppression, females must undergo a behavioral transition to initiate active searching and evaluation of oviposition sites. To investigate this shift, we used oviposition and activity-monitoring assays to characterize the behavioral states in anticipation of egg laying. We found that Aedes aegypti females transition from quiescence to hyperactivity within 3 days of blood feeding, coinciding with egg maturation, and initiate a distinct behavioral program that approximately doubles their daily activity. We further show that hyperactivity is closely linked to humidity seeking, which is essential for effective oviposition site exploration and reproductive success. Notably, this hyperactive state diverges from the species’ typical diurnal activity and instead reflects a nighttime pattern of egg laying. The assignment of a nocturnal phase to egg-laying behaviors is lost in cycle (circadian clock) mutants. This unique gonadotrophic behavioral state demonstrates significant plasticity, with sensitivity to both internal and external environmental cues, and complete reversion upon successful oviposition. RESULTS Temporal reprogramming of activity in gravid females After ingesting a blood meal, mated Aedes aegypti females suppress locomotor activity while they digest blood and develop eggs, which are ready to be deposited approximately 3 days later25–27,29 (Figure 1A). To characterize the behavioral state associated with egg-bearing females, we fed females a sheep blood meal and monitored locomotor activity over the following 4 days (Figure 1B; see also Figure S1A). Consistent with previously described locomotor activity suppression,26,29 females exhibited a pronounced reduction in overall activity during the first 2 days post-blood-meal, with a single peak occurring at dusk (Figures 1B and 1C). However, by the third day, coinciding with egg maturation, mosquitoes entered a hyperactive state with activity levels nearly double those of non-blood-fed controls, which persisted into the fourth day in the absence of an egg-laying site (Figure 1C). Despite this reactivation, the activity pattern remained unimodal with a prominent evening activity peak, differing from the bimodal rhythm seen in non-fed controls (Figure 1B). Additionally, the anticipatory activity typically observed before lights off in non-blood-fed animals was absent on the first day after the blood meal but resumed and became prominent by 3 days post-blood-meal, indicating an increase in overall activity. Notably, we also detected nocturnal hyperactivity in gravid females, which contrasts sharply with non-blood-fed females that show minimal activity at night (Figure 1D). The locomotor activity monitor does not provide an egg-laying substrate, confirmed by the absence of eggs during the 4-day recording period. To determine whether this hyperactive state is reversed upon oviposition, we allowed females to lay eggs in “ovi-vials” with a wet paper substrate (Figure S1B). After oviposition, females were transferred to locomotor activity monitors alongside age-matched, non-blood-fed controls. Activity levels and bimodal patterns were restored post-oviposition, confirming that the behavioral changes associated with egg laying are reversible and state dependent (Figures 1E and 1F; see also Figure S1C). Further, gravid females displayed no apparent phase shift compared to non-blood-fed controls under constant darkness (DD) conditions (Figure 1G; see also Figure S2A), suggesting that this temporal reprogramming is likely not a direct result of a phase delay of internal circadian timing. Instead, we suspect that these behavioral changes are driven by downstream modulators of activity that alter the suppression of activity in darkness. Indeed, gravid females displayed similar amplitude of hyperactivity under both light-dark (LD) and DD conditions (Figure 1H). Constant light (LL) conditions, however, induced decreased phase synchrony and hyperactivity in both blood-fed and non-blood-fed females (Figures S2B and S2C). Overall, our data suggest that the gravid state follows hyperactive programming with the unconventional appearance of nighttime activity, even under dark conditions that normally suppress locomotion.24,48 Post-biting behavioral reprogramming in gravid Aedes aegypti mosquitoes (A) Schematic of the reproductive timeline of wild-type female Aedes aegypti mosquitoes. After feeding on a host, females begin blood meal digestion and egg development. Once eggs mature, females seek freshwater sites to lay eggs. (B) Locomotor activity profile of non-blood-fed (n = 58) and blood-fed (n = 56) wild-type Aedes aegypti mosquitoes. (B–F) Locomotor activity was recorded under 12 h light:12 h dark LD cycles with 30 min dawn/dusk light transition periods. (C) Total daily activity of non-blood-fed (n = 58) and blood-fed (n = 56) wild-type Aedes aegypti mosquitoes. Mann-Whitney test. (D) Average nocturnal activity (ZT13–ZT24) of non-blood-fed (n = 58) and blood-fed (n = 56) wild-type Aedes aegypti mosquitoes. Mann-Whitney test. (E) Average daily activity of non-blood-fed (n = 29) and oviposited (n = 18) wild-type Aedes aegypti mosquitoes. Mann-Whitney test. See also Figures S1B and S1C. (F) Average nocturnal activity (ZT13–ZT24) of non-blood-fed (n = 29) and oviposited (n = 18) wild-type Aedes aegypti mosquitoes. Mann-Whitney test. See also Figures S1B and S1C. (G) Locomotor activity profile of non-blood-fed (n = 34) and blood-fed (n = 40) wild-type Aedes aegypti mosquitoes under 1 day of LD entrainment, followed by 3 days of constant darkness (DD). (H) Total activity days 3–4 post-blood feeding under LD (B) and DD (G) conditions. Unpaired t test. ZT, zeitgeber time, with lights on at ZT0 and off at ZT12. The white and gray shading in (B) represents the light and dark phases, respectively. The light gray shading in (G) represents the respective light phase from prior LD entrainment. The lines and shaded lines in (B) and (G) show the mean and SEM. The lines and error bars in (C)–(F) and (H) show the mean and SEM, and the dots show individual animals. ***p < 0.001, ****p < 0.0001, and p > 0.05, not significant (ns). The exact p values are provided in Table S1. See also Figures S1 and S2. The circadian rhythmicity and synchrony of gravid females under DD conditions (Figure 1G) suggest that the clock temporally regulates this hyperactive state.49 To determine whether a functional circadian clock is required for behavioral reprogramming in gravid females, we examined females with a loss-of-function mutation in the core clock gene cycle.23 In these mutants, the disrupted CYCLE protein is unable to dimerize with its partner CLOCK, resulting in a disabled circadian clock and loss of bimodal activity under LD conditions.23 In wild-type non-blood-fed females housed under 12:12 LD conditions, morning and evening activity peaks are separated by low midday activity (“siesta”), and this bimodal activity modulation is lost in cycle mutants (Figure S3A). Locomotor activity recording in DD conditions reveals decreased rhythmicity (Figures S3B and S3C), and cycle mutants exhibited drastically reduced locomotor activity in DD, suggesting that the suppression of activity in darkness is more dominant in the absence of a functional clock (Figures S3B and S3D). Following a blood meal, cycle mutants failed to exhibit the elevated activity levels and nocturnal hyperactivity observed in gravid wild-type females (Figures 2A–2C). These mutants also lacked anticipatory activity before dusk and did not show extended nocturnal activity into the night (Figure 2D). These results demonstrate that the circadian clock, specifically the clock gene cycle, is essential for the emergence of the hyperactive, nocturnally shifted behavioral state in gravid Aedes aegypti females. Alternative activity programming in gravid Aedes aegypti females is dependent on the clock gene cycle (A) Locomotor activity profile of non-blood-fed (n = 21) and blood-fed (n = 19) cycle mutant females. (B) Total daily activity of non-blood-fed (n = 21) and blood-fed (n = 19) cycle mutants. Mann-Whitney test. (C) Average nocturnal activity (ZT13–ZT24) of non-blood-fed (n = 21) and blood-fed (n = 19) cycle mutants. Mann-Whitney test. (D) Average daily activity profile of blood-fed wild-type females (n = 56) and cycle mutants (n = 19), days 3 and 4 post-blood-meal. The white and gray shading in (A) and (D) represents the light and dark phases, respectively. The lines and shaded lines in (A) and (D) show the mean and SEM. The lines and error bars in (B) and (C) show the mean and SEM, and the dots show individual animals. ****p < 30.0001 and p > 0.05, not significant (ns). The exact p values are provided in Table S1. See also Figure S3. Nighttime activity is temporally linked to humidity seeking While host seeking remains suppressed during egg development, we reasoned that the hyperactive state in gravid females may reflect a shift in motivational state from seeking hosts to seeking suitable oviposition sites. Aedes aegypti are highly adapted to human environments, seeking out artificial water containers for egg laying.44 Additionally, they can exhibit “skip oviposition,” distributing eggs across multiple sites to maximize the chances of offspring survival.50–53 We hypothesize that hyperactivity facilitates oviposition site search specifically at night. To test whether the observed activity shift is linked to sensory-driven oviposition behaviors, we examined humidity seeking, as an increased humidity level is an ecologically relevant cue that guides gravid females to freshwater egg-laying sites.46,47,54 We developed a custom humidity preference assay in which mosquitoes were housed in a chamber with mesh windows above dry and water-containing Petri dishes while preventing direct contact with the water (Figures 3A and 3B). Therefore, females retain their eggs in this arena due to the lack of an egg-laying substrate. Using this system, infrared (IR) images of the arena were acquired every minute, and hourly time-minimum projections were generated (Figure 3C). These time-minimum projections accumulate dark pixels as mosquitoes land and perform local search behaviors. Therefore, the number of dark pixels can be used as an activity index. When the activity index was measured from the entire arena (Figure S4A), we observed similar temporal activity patterns and hyperactivity in gravid females as well as a lack of hyperactivity in cycle mutants, even when females were group housed in this larger arena (Figure S4B). To investigate humidity-seeking activity, we quantified activity index specifically on the “wet” versus “dry” windows (Figure 3C). In our behavioral setup, the wet and dry mesh windows show an ~7% difference in relative humidity that quickly declines as the distance from the mesh increases (Figure 3D). Wild-type non-blood-fed females show consistently low humidity-seeking behavior (Figure 3E). In contrast, gravid females show a rhythmic pattern with pronounced humidity seeking in an approximately 5-h window after dusk (Figure 3E). This rhythm emerged 3 days post-blood-meal, coinciding with the onset of hyperactivity. In cycle mutants, although overall humidity preference increased after a blood meal, the rhythmic timing observed in wild-type females was lost (Figures 3E and 3G). This suggests that proper timing of humidity seeking requires a functional circadian clock, even in the presence of LD cycles. These results show strong temporal coupling between nocturnal activity and humidity seeking in gravid females, indicating that the circadian system plays a critical role in coordinating the timing of oviposition-related behaviors to ensure that egg laying occurs at an optimal time of day. Rhythmic, nocturnal humidity seeking is regulated by the clock gene cycle (A) Experimental timeline of the humidity-seeking assay. Twenty female mosquitoes were used in each trial. (B) Schematic of the humidity-seeking arena. (C) Representative 1 h time-minimum projections used to calculate activity index. (D) Difference in relative humidity above wet and dry zones in the humidity-seeking arena (B and C). Three humidity measurements were taken. (E) Humidity-seeking preference index (PI) of wild-type and cycle mutant females (blood-fed and non-blood-fed). (E–G) n = 6 biological replicates with 20 females per replicate. Mann-Whitney test. (F) Cumulative humidity-seeking activity of wild-type and cycle mutant females (blood-fed and non-blood-fed, days 3–4 post-blood-meal). Mann-Whitney test. (G) Average humidity-seeking profile of blood-fed wild-type and cycle mutant females, days 3–4 post-blood-meal. BioDare2 eJTK rhythmicity test. The white and gray shading in (E) and (G) represents the light and dark phases, respectively. The lines in (D)–(G) show the mean, and the error bars in (F) show the SEM. The dots in (D), (F), and (G) and the thin lines in (E) show independent replicates. **p < 0.01. The exact p values are provided in Table S1. See also Figures S4 and S5. Nocturnal egg laying in a diurnal mosquito In non-blood-fed Aedes aegypti mosquitoes, daily behavioral timing follows a diurnal, bimodal pattern with peaks at dawn and dusk in laboratory and field studies.11,13–15,55 Our data suggest that gravid females exhibit an alternative circadian program characterized by nocturnal hyperactivity and humidity seeking that may reflect the preferred time of day for egg laying. Aedes aegypti mosquitoes require contact with a wet surface to lay eggs (Figure 4A).43,45 While previous studies have explored oviposition rhythms by manually replacing egg-laying substrates every few hours, this approach is labor intensive and susceptible to unwanted experimenter influence.39,40,53,56,57 Across insect species, monitoring the internal timing of oviposition remains a technical challenge, and efforts have focused on developing automated systems to generate oviposition rhythms with minimal experimenter bias.58 Here, we introduce “EggTimer,” a behavioral monitoring platform that enables continuous automated tracking of egg laying (Figure 4B). In this setup, mosquitoes are individually housed above a wet paper substrate connected to a water reservoir and illuminated with IR light, to allow real-time visualization of egg deposition and locomotor activity. Aedes aegypti eggs are initially white due to their transparent endochorion but melanize and darken within 3 h after being laid.41,59–61 We developed a custom image-processing algorithm to detect pixel-level changes in darkness, enabling accurate determination of oviposition timing (Figure 4C). This algorithm is designed to reduce “noise” from mosquito excretions (Figure S5A) and shows a temporal oviposition profile comparable to manual counting (Figure S5B). Using this system, we observed that, despite their diurnal locomotor activity prior to blood feeding, Aedes aegypti females displayed a unimodal oviposition rhythm when given continuous access to an egg-laying substrate, with oviposition occurring most frequently immediately after dusk (Figures 4D and 4E). In cycle mutants, rhythmicity and nocturnal preference are disrupted (Figures 4D and 4E), and eggs are laid throughout the day, demonstrating that circadian control is essential for proper oviposition timing. To investigate if nighttime egg laying is a direct response to lights-off, we measured egg-laying timing in DD conditions. Interestingly, egg-laying timing is phase advanced without external zeitgeber cues and occurs primarily on the second day post-blood-meal (Figure S6). This phase advance is similar to the shift in locomotor activity in DD that reflects a free-running period of ~22.6 h in Aedes aegypti. Egg laying in DD in wild-type females is mostly observed within a 5-h window before CT12 but is less temporally restricted in cycle mutants (Figure S6). Overall, this demonstrates that lights-off is not required for mosquito egg laying and that synchronous egg laying is under circadian control. Hyperactivity is contingent upon egg retention (A) A gravid female Aedes aegypti mosquito laying eggs on a wet paper substrate. Newly laid eggs are white and darken over time. (B) Schematic of the EggTimer oviposition monitor. (C) Representative images of an egg-laying female acquired with the EggTimer oviposition monitor. The boxes indicate the shadow of the female mosquito. (D) Egg-laying rhythm in wild-type (n = 41) and cycle mutant (n = 30) females. (E) Percentage of eggs laid in scotophase (ZT12–ZT24). Mann-Whitney test. (F) Representative egg-laying and locomotor activity of wild-type females in the EggTimer assay. The individuals shown are from one cohort. (G) Fold change in daily activity comparing blood-fed and non-blood-fed wild-type females in the locomotor activity monitor (n = 56) and the EggTimer (n = 37). The white and gray shading in (D) and (F) represents the light and dark phases, respectively. The dots and error bars/shading in (D), (E), and (G) show the mean and SEM. **p < 0.01. The exact p values are provided in Table S1. See also Figures S6 and S7. Hyperactivity is contingent upon egg retention In this assay, gravid females typically laid eggs between the second and third days post-blood-meal, suggesting that when housed directly on an egg-laying substrate, hyperactivity onset may occur earlier than when no suitable oviposition site is available. To test this, we conducted extended EggTimer recordings on blood-fed females using an experimental timeline that matched our locomotor activity recordings (Figure 1B). We tracked locomotor activity using the mosquito’s shadow on the wet paper substrate to infer the female’s activity level and temporal relationship to oviposition (Figure 4C; see also Figure S7A). By the second day post-blood-meal, females showed a reactivation of activity that was tightly aligned with the onset of egg laying (Figure 4F). Notably, both overall and nocturnal hyperactivity peaked on the day of oviposition but returned to baseline the following day (Figure 4F; see also Figure S7B). To assess whether the presence of an oviposition substrate influenced the magnitude or timing of hyperactivity, we compared locomotor activity over 4 days in two conditions: females housed with an oviposition substrate (EggTimer assay) and females without access to an oviposition substrate (standard locomotor activity assay; Figure 1). Females without a substrate showed sustained hyperactivity beginning on day 3 post-blood-meal, while those with a substrate exhibited earlier hyperactivity on day 2 that resolved by day 3 (Figure 4G). This suggests that access to a substrate reduces the need for prolonged movement (see also Figure S7C). Nonetheless, a nocturnal shift in activity persists and remains tightly coupled to nighttime egg laying. These findings indicate that the hyperactive state is internally driven but modulated by environmental context. Hyperactivity supports long-range oviposition site seeking Gravid females must navigate and evaluate a complex and variable sensory landscape when selecting oviposition sites in human environments.62 Suboptimal or hazardous sites, such as those with high salinity, can severely compromise reproductive success. If suitable sites are not found within a few days of egg maturation, females risk laying eggs in lethal conditions or may fail to lay them altogether.34–37 Our data suggest that gravid females enter a hyperactive behavioral state with altered activity timing to maximize the likelihood of successful oviposition—the final and critical step of the reproductive cycle. Without a functional clock, cycle mutants do not exhibit hyperactivity and nocturnal humidity seeking, and we hypothesize that this disrupts their ability to search for and locate suitable oviposition sites. To model a scenario requiring extended search behavior, we designed an assay consisting of five sequentially connected cages: the first three cages next to the starting cage each contained an ovi-cup filled with 0.24 M NaCl (lethal to larval growth36), while the final cage contained an ovi-cup with freshwater (Figure 5A). We released a group of 20 gravid females in the starting cage at zeitgeber time (ZT) 11.5 (30 min before dusk) of the third day post-blood-meal and allowed them to explore and lay eggs for the next 5 h, coinciding with the peak humidity seeking and egg laying observed in our earlier assays. To ensure that differences in egg laying were due to exploratory and not physiological deficits, we confirmed that cycle mutants lay eggs as efficiently as wild-type females when housed directly on a freshwater oviposition substrate and tested each genotype separately to ensure that community-based cues did not influence site selection63 (Figures S8A and S8B). Hyperactivity underlies reproductive efficiency (A) Schematic of the long-range freshwater-seeking behavioral assay and ovi-cups used for egg collection. Animals 3 days post-blood-meal (p.b.m.) were released in groups of 20 at ZT11.5 and allowed 5 h in the behavioral assay. (B) Heatmap of egg-laying results by wild-type (n = 6) and cycle mutant (n = 11) Aedes aegypti mosquitoes in the long-range freshwater-seeking assay. (C) Eggs laid per female across all cups by wild-type (n = 6) and cycle mutant (n = 11) females in cages 1–4 of the long-range freshwater-seeking assay. Mann-Whitney test. (D) Number of surviving offspring per female predicted with oviposition-site lethality by wild type (n = 5) and cycle mutants (n = 11). Mann-Whitney test. (E) Schematic of the short-range freshwater-seeking behavioral assay. Animals 3 days p.b.m. in groups of 20 were released at ZT11.5 and allowed 5 h in the behavioral assay. (F) Heatmap of egg-laying results by wild type (n = 6) and cycle mutants (n = 6) in the short-range fresh-water-seeking assay. (G) Freshwater preference of wild-type and cycle mutants in the long- versus short-range freshwater-seeking assay (n = 6 for all conditions and genotypes).Trials with fewer than eight eggs per female were excluded to ensure sufficient participation. Mann-Whitney test. The line and error bars in (C), (D), and (G) show the mean and SEM, and the dots show the individual trials. The rows in (B) and (F) are sorted by the total number of eggs collected. *p < 0.05, **p < 0.01, ***p < 0.001, and p > 0.05, not significant (ns). The exact p values are provided in Table S1. See also Figure S8. In the long-range assay, wild-type females laid eggs almost exclusively in the farthest freshwater cup, showing a strong avoidance of the lethal substrates (Figure 5B). Cycle mutants, however, deposited more eggs in the lethal cups and showed reduced egg deposition in the freshwater option (Figure 5B). Despite similar egg-laying efficiency when placed directly on an oviposition substrate, cycle mutants laid significantly fewer eggs overall in the long-range freshwater-seeking assay, likely due to a lack of hyperactivity required to access oviposition sites (Figure 5C). This suggests that reduced egg output is not due to general impairment in egg production but may be attributable to defects in egg-laying timing (Figure 4D) and site searching in the long-range assay. The combination of limited exploratory behavior and inappropriate oviposition in lethal sites results in significantly lower reproductive success in cycle mutants, measured by predicted surviving offspring36 (Figure 5D). To investigate whether cycle mutants show impairments in detecting salinity or making oviposition choices on a small spatial scale, we established a short-range freshwater-seeking assay (Figure 5E). In this short-range setup (40 times smaller than the long-range assay), cycle mutants consistently laid more eggs in the freshwater cup (Figure 5F). We then calculated freshwater preference in all trials with more than 8 eggs per female. The defect in freshwater preference of cycle mutants in the long-range assay was recovered in a short-range setup, indicating that they can recognize suitable oviposition sites at close range (Figure 5G). We hypothesize that the impaired long-range oviposition behavior in cycle mutants results from a combination of increased internal pressure to lay eggs and reduced exploratory behavior, due to the absence of nocturnal hyperactivity, ultimately compromising their ability to locate optimal egg-laying sites. Indeed, when no suitable oviposition substrate is available, wild-type females will eventually deposit eggs in a salt-water cup, although with reduced efficiency and avoidance of the paper substrate (Figures S8C and S8D). DISCUSSION A transient and reversible shift in temporal niche While circadian regulation is typically defined by robust, self-sustained rhythms under constant conditions, examples of plasticity in daily behavioral timing that result in temporal niche switches are found across the tree of life.64 These shifts can be triggered by environmental and physiological factors, including energy balance65,66 and seasonal changes.67,68 In this study, we characterize a striking example of temporal niche flexibility in the diurnal mosquito Aedes aegypti, which exhibits a transient, dusk-to-nighttime preference for egg laying. This oviposition timing mirrors patterns observed in nocturnal mosquito species such as Anopheles gambiae,69 despite Aedes aegypti being otherwise day active. The convergence of oviposition timing across both diurnal and nocturnal species points to the existence of evolutionarily conserved circadian mechanisms that coordinate reproductive timing across diverse mosquito lineages and suggests that nighttime egg laying may confer general adaptive advantages. Newly laid eggs lack a melanized endochorion, making them more susceptible to UV radiation and desiccation.59 Laying eggs at sunset may offer optimal environmental conditions, including lower temperatures, higher humidity, and diminished UV levels, that promote egg viability. Our findings show that Aedes aegypti females, although typically diurnal, are capable of reprogramming their activity to adopt a nocturnal egg-laying schedule. This behavioral reprogramming requires the molecular circadian clock. In cycle mutants, the rhythmicity and timing of both humidity seeking and oviposition are disrupted, even under entraining LD conditions, indicating that an intact molecular clock is essential for coordinating these behaviors. Supporting this, gravid wild-type females maintain circadian rhythmicity under DD conditions, suggesting that the behavioral changes arise from clock-driven output pathways. Intriguingly, “restless” egg-laying nights in gravid females draw parallels with seasonal behavioral changes in migratory birds known as “zugunruhe” or nocturnal restlessness observed in otherwise diurnal birds.70 This circannual migratory drive is conserved in nonmigratory birds,71 suggesting a potential utility of this alternative program in migratory flexibility in variable environments. However, unlike seasonal niche switches, the shift to nocturnal activity in mosquitoes is transient, tightly coupled to reproductive physiology, and reversed immediately following egg deposition. The features we observe in female mosquitoes define a unique form of temporal niche plasticity that is both specific to egg-bearing females and under strong circadian control. Such plasticity highlights the adaptability of behavioral timing in the service of reproductive success. Hyperactivity as a mechanism for egg-laying efficiency Accompanying the shift in behavioral timing, gravid females enter a hyperactive state that overrides post-blood-meal activity suppression. Interestingly, although locomotor hyperactivity occurs before and after lights off, humidity-seeking and egg-laying activities are mostly nocturnal. This pattern suggests that egg-laying-specific sensory behaviors may be further regulated by environmental cues. Given that this heightened activity is sustained during egg retention and rapidly ceases following oviposition, we propose that it represents a short-term, energetically costly strategy that enhances the likelihood of locating suitable egg-laying sites, which ultimately maximizes reproductive success. Mosquito blood feeding occurs across a range of environments varying in urbanization, vegetation, and microclimate.35,43 While some mosquito species undertake long-range migrations in search of oviposition sites, Aedes aegypti is more commonly associated with human-modified habitats, actively seeking out small artificial containers such as tires, pots, and gutters.32,44,54,72 In such unpredictable landscapes, hyperactivity may enable gravid females to detect and integrate a wide array of sensory cues—physical, chemical, and visual—that vary in signal strength and range. Indeed, previous descriptions of mosquito oviposition behavior refer to a phase of seemingly “random activity,” thought to aid in the discovery of suitable sites.45,54 Our findings support this notion, highlighting hyperactivity as an adaptive behavioral state that facilitates site evaluation under uncertain environmental conditions. The rapid cessation of hyperactivity post-oviposition suggests that signals from the gravid ovary may sustain this elevated activity state, and earlier studies have shown that transplanting a gravid ovary into a non-blood-fed female can elicit pre-oviposition behaviors, implicating ovarian signals in modulating behavior.73 Future work aimed at identifying the molecular nature of these signals could uncover novel targets to reduce reproductive efficiency. The mosquito gonotrophic cycle as a target for behavioral control The unique reproductive biology of the female mosquito provides a compelling model to investigate the innate regulation of behavior. Our study describes distinct internal states that define host-seeking, blood-digestion, and egg-laying phases that are accompanied by shifts in sensory tuning and circadian timing. Together, this sequence of transitions reveals an extraordinary degree of behavioral plasticity. Identifying the mechanisms that control behavior across these key reproductive stages will advance our understanding of mosquito biology and open new approaches for vector control. These results demonstrate that the circadian clock is essential for the behavioral reprogramming that supports efficient egg laying. Targeting this tightly regulated gonadotrophic cycle by disrupting the circadian system could provide novel approaches to prevent mosquito reproduction and to mitigate disease transmission. Limitations of the study While our study leverages laboratory assays to investigate behavioral signatures of egg-laying mosquitoes, several factors may limit the generalizability of these findings. In this study, we reported temporal features of mosquito daily activity profiles using a laboratory-based locomotor activity assay, which does not fully reflect the behavioral transitions of blood-fed mosquitoes in the wild. While these compact assays are valuable, larger and more naturalistic behavioral arenas may offer ecologically relevant insights, particularly into the intensity of the hyperactivity phase and its role in the search for an optimal oviposition site. Our assays were conducted under conventional LD cycles, yet environmental cues altered by human activity, such as artificial light at night,48 may significantly influence behavioral timing. This is especially relevant for human-preferring, pathogen-vectoring species such as Aedes aegypti. In our work, we investigated the behavioral difference in a cycle-knockout mutant. It is important to note, however, that this genetic tool does not allow us to distinguish cycle-specific from broader circadian clock effects. Addressing this limitation will require the development and testing of additional clock gene mutants. Spatially restricted genetic tools could help further dissect central versus peripheral clock regulation of reproductive behaviors. Lastly, future work to investigate the oscillation of molecular clock components with high temporal resolution could provide insight into the molecular underpinnings of this behavioral reprogramming. RESOURCE AVAILABILITY Lead contact Requests for resources and reagents should be addressed to the lead contact, Laura B. Duvall (lbd2126@columbia.edu). Materials availability This study did not generate new unique reagents. Data and code availability All data used in the main and supplemental figures are available in the supplemental information. All data used in the main and supplemental figures are available in the supplemental information. All original code used in this study can be found on GitHub and Zenodo (https://doi.org/10.5281/zenodo.17507875). All original code used in this study can be found on GitHub and Zenodo (https://doi.org/10.5281/zenodo.17507875). Any additional information required to re-analyze the data reported in the paper is available from the lead contact upon request. Any additional information required to re-analyze the data reported in the paper is available from the lead contact upon request. STAR★METHODS EXPERIMENTAL MODEL AND STUDY PARTICIPANT DETAILS Aedes aegypti Liverpool mosquitoes and cycle23 mutant strains were maintained and reared at 28°C, 70–80% relative humidity with a 12 h light: 12 h dark schedule. Eggs were hatched using 1 L of TetraMin fish food suspension in deionized deoxygenated water (one tablet/L). Larvae were fed with TetraMin fish food tablets each day as needed. Adult female mosquitoes used for experiments were 3–21 days old and co-housed with males, with ad libitum access to 10% sucrose solution until further experimental procedures. For passaging, adult female mosquitoes were blood-fed on defibrinated sheep blood (Quad5) with 1 mM ATP at 37°C using Hemotek artificial membrane feeders. METHOD DETAILS Blood feeding for behavioral assays All animals tested for post-blood-meal behavior were entrained for at least 3 days under a 12 h light: 12 h dark schedule. Access to sucrose solution was removed and a cotton wick soaked with DI water was provided approximately 24 h before blood-feeding. Animals were allowed to blood-feed on defibrinated sheep blood (Quad5) with 1 mM ATP at 37°C using Hemotek artificial membrane feeders. The feeder was rubbed on experimenter’s arms to provide human odor. The experimenter’s breath was used to improve feeding rate. After 30 min of access to a feeder, cages were transferred to 4°C to cold anesthetize animals. Females were scored by eye for complete abdominal distension indicating that they fed to repletion, non-fed animals were removed, and fully blood-fed animals were returned to 28°C, 70–80% relative humidity. All feeding was conducted at ZT8–9. Locomotor activity assay Locomotor activity was assessed using a modified version of the MozzieBox locomotor activity assay.11 This modified design is constructed with four 23 cm miniature t-slotted framing (McMaster-Carr) and custom cut acrylic sheets (3 mm thickness, black or clear) that has overall dimensions of 29 × 18 × 31 cm to fit inside a compact incubator (Heratherm, Thermo Fisher) to allow for temperature control. All recordings were performed with incubator temperature set at 28°C. Infrared LED lights (850nm, Waveform Lighting) were installed at the bottom panel, and four 4 × 4 cm cooling fans (Noctua NF-A4×10 FLX) were installed to improve heat dissipation and air circulation. RaspberryPi Camera Module 3 NoIR Wide angle cameras covered with IR longpass filters (Edmund Optics 43–948) were mounted on top of the acrylic frames to take images every minute. White LED strips were installed and controlled together with the camera via custom Python script. All platforms and components were assembled with corner brackets and surface brackets for miniature t-slotted framing (McMaster-Carr). Python scripts can be found on Github (https://github.com/DongLinhan/EggTimer2025). Each well of a 12-well tissue culture plate was supplied with 1.5 mL of 10% sucrose in 1% agarose, with 1% (v/v) 15% p-hydroxybenzoic acid methyl ester in 95% ethanol (antifungal). Female mosquitoes (non-blood-fed, blood-fed, oviposited) were individually loaded into the wells under cold anesthesia. All recordings of blood-fed or oviposited females were performed with age-matched controls from the same rearing cage. Loading of blood-fed animals was conducted immediately after blood-feeding at ZT8. For post-oviposition activity recording, females were loaded into individual ovi-vials on day 2 post-blood-meal, constructed using Drosophila vials (VWR 75813–164), folded filter paper (55 mm, Whatman), sitting in 5 mL of DI water. Animals were only loaded for post-oviposition activity recording if a full clutch of over 50 eggs was laid in the ovi-vial by day 4 post-blood-meal. Up to four plates were loaded into the monitor before ZT10, and recording started the day after at ZT0. For all LD recording, four LD cycles with 30 min dawn and dusk periods were applied and IR images were taken every minute. For recordings in Figures 1G and S2B, one LD cycle was applied before 3 days of DD/LL conditions. For recordings in Figure S3, three LD cycles were first applied, before switching to DD for 7 days. Downstream analysis of images uses the same pipeline as in Dong et al.11 Briefly, images were converted into a 30 frames-per-second video using a custom MATLAB code, and the video was further analyzed using Ethovision (version 17) to extract positional features for movement analysis. Captured movements between frames are calibrated to centimeters and reported as “activity score”. Analysis of circadian rhythmicity and phase synchrony used the same pipeline and custom coding as in Dong et al.11 Briefly, raw distances were transformed into activity scores that represents activity in 1-min bins. Calculation of rhythmic power was performed using the phase package (1.2.9) in R (version 4.2.3).74 An animal was considered rhythmic if rhythmic power is greater than 50. For phase quantifications of DD activity, a Savitzky-Golay filter was applied to animal activity scores in the third and fourth day of recording. Phase synchrony was determined using the Rayleigh R test. Humidity-seeking assay Humidity-seeking activity was assessed using the MozzieMonitor platform75 with custom-built humidity-seeking arenas (Figure 3B). The humidity-seeking arena is a 10 × 10 × 18 cm acrylic box with two 5 × 5 cm open mesh windows constructed with 3 mm clear acrylic sheets. Facing the mesh windows, a round Petri dish (100 mm × 15 mm, Celltreat) filled with DI water and an empty Petri dish were placed directly under the arena. Sucrose solution was provided to the arenas via a cotton wick partially soaked in 10% sucrose. For each experiment, 20 blood-fed females and 20 non-blood-fed control females were aspirated into two separate behavioral arenas 1 day after blood feeding. Animals were allowed to acclimate in the behavioral arena and image acquisition began the following day at ZT0 for 3 days under LD conditions with 30 min dawn and dusk periods. IR images were taken every minute, and a time-maximum projection was created using custom Python code as background. Every 60 images were converted into a time-minimum projection, and pixels were counted if below the darkness threshold (less than 100). Pixel counts were subtracted from the background image and grouped by user-defined ROI using custom Python coding and reported as “activity index”. For estimates of total activity in Figure S4, the ROI is defined to the entire bottom side of the humidity-seeking arena. The analysis code used can be found on Github (https://github.com/DongLinhan/EggTimer2025). Rhythmicity test of humidity seeking activity was carried out using BioDare276 (https://biodare2.ed.ac.uk). Average PI was grouped by 3-h intervals and this average value was analyzed via the BD2 eJTK method with no detrending applied and the BD2_CLASSIC preset. Test parameters and results are included in Table S1. EggTimer oviposition monitor The EggTimer assay was modified from the MozzieMonitor platform75 with adaptations to platform heights and orientation. A RaspberryPi Camera Module 3 NoIR Wide camera was mounted on the bottom platform covered with an IR longpass filter (Edmund Optics 43–948). Miniature t-slotted framing (30 cm, McMaster-Carr) was installed 11 cm above the camera to support a removable platform made with 3 mm acrylic sheet. An IR lighting platform was installed 9 cm above this removable platform, and two 14 × 14 cm USB fans (Newegg 9SIB7W5HBZ4935) were added to the IR lighting platform to provide ventilation and prevent overheating. An additional acrylic sheet covered in Kimwipe was installed 1.5 cm below the IR lights to provide diffusion of the IR light. Two open plastic containers (6 × 8 × 13.5 cm) filled with DI water were placed on the camera platform to provide a water source for the wet paper substrate. Up to two 12-well tissue culture plates can be imaged on the imaging platform. The imaging platform is covered in two layers of wet Kimwipes, leaving overhangs on each side of the platform to be soaked in the water source. For oviposition recordings, blood-fed females were loaded into individual wells of a 12-well tissue culture plate 1 day after blood feeding under cold anesthesia. Animals were allowed to acclimate in the behavioral arena and image acquisition began the following day at ZT0 for 3 days under LD conditions with 30-min dawn and dusk periods. For long-term oviposition/locomotion recordings, loading of blood-fed animals was conducted immediately after blood feeding at ZT8 to match the timeline of the locomotor activity recordings. For recordings of oviposition activity in DD, loading of blood-fed animals was conducted immediately after blood feeding at ZT8 and subject to 1 LD cycle in the following day before switching to DD on the second day post-blood-feeding. Images were taken every minute and female movement analysis followed the same pipeline as the locomotor activity assay. For egg-laying activity, we established a pipeline to detect pixel value change overtime that is characteristic of a darkening egg. While white, un-melanized eggs are visible and induce pixel value change under IR imaging, we used the pixel darkening profile to verify this initial change to reduce noise signals from animal movement. Pixel values were extracted from images acquired during recordings and time series of pixel values were processed with a Savitzky-Golay filter (window size = 15, polynomial order = 3). For each individual time series of pixel values, two rolling windows (A: 30 min, B: 240 min) were used to detect a significant and smooth drop in pixel value that persists until the end of the recording. The end of window A was then recorded as the egg-laying onset if window A and B matched the numerical criteria established in our algorithm. Detailed analysis code, along with data processing procedures and corresponding threshold values are deposited on Github (https://github.com/DongLinhan/EggTimer2025). On-substrate assay Egg-laying behavior was assessed by housing females 3 days post-blood-feeding directly in ovi-cups, eliminating the need for long-range oviposition site-seeking. A 4-oz cup with cap (VWR 8900–662) was filled with 50 mL DI water or 0.24 M NaCl and four pieces of filter paper (55mm, Whatman). A 35 × 10 mm plastic Petri dish (Greiner) was fastened to the inner side of the ovi-cup cap with screws and was used to house females until the start of the experiment. Before each experiment, 10 females were cold anesthetized and placed in the Petri dish inside the ovi-cup, and returned to 28°C, 70–80% relative humidity to acclimate for 90 min. At ZT11.5, the Petri dish cap was lowered by 1 cm, releasing females into the arena. At ZT13, ovi-cups were moved to a cold room to anesthetize and remove females, concluding the experiment. All handling was done with gloves to avoid human odor contamination. For salt-water trials, eggs were manually counted. For freshwater trials, images of the eggs were taken using a webcam (Logitech) and dark pixels were converted into egg counts. Analysis code can be found on Github (https://github.com/DongLinhan/EggTimer2025). Long-range freshwater seeking assay Long-range freshwater-seeking behavior was assessed in an arena consisting of five 30 × 30 × 30 cm cages (BugDorm DP1000) connected via custom 20 cm mesh tunnels with openings 15 cm in diameter (BugDorm DP1000-N) (Figure 5A). All assays were performed in an environmentally-controlled room at 28°C, 70–80% relative humidity. Animals were tested 3 days post-blood-feeding and 20 females were moved to an 8-oz plastic cup with a cap (VWR 8900–664) under cold anesthesia. Females were placed in the starting cage to acclimate for 90 min before experimentation. Ovi-cups were made with 4-oz cups (VWR 8900–662) with four pieces of filter paper (55mm, Whatman) lining the sides and filled with 50 mL 0.24 M NaCl solution, which is lethal to mosquito larvae,36 or DI water. Salt ovi-cups were placed in cages 1–3 (closest to starting cage), and the freshwater ovi-cup was placed in cage 4 (farthest from starting cage). Handling of arena components and mosquitoes was done with gloves to avoid contamination with human odor. At ZT11.5, females housed in the 8-oz cup were released into the arena and allowed 5 h in the behavioral setup. Ovi-cups were removed at ZT16.5 and eggs laid in ovi-cups were allowed to melanize overnight and manually counted the following day. To report preference between salt and freshwater, we excluded trials with fewer than 150 eggs total collected to prevent skewed preference due to low participation. Predicted survival was calculated using published data,36 where freshwater substrate led to 92.22% survival; and 0.24 M NaCl used in this study is above the threshold concentration that leads to immediate larval lethality. Short-range freshwater seeking assay Short-range freshwater-seeking behavior was assessed 3 days post-blood-feeding, and 20 females were aspirated into a 14 × 14 × 14 cm small mesh cage (BugDorm 4M1515) and acclimated for 90 min. Two 4-oz ovi-cups (VWR 8900–662) filled with either 50 mL DI water or 50 mL 0.24 M NaCl with four pieces of filter paper (55mm, Whatman) lining the sides were placed in the cage at ZT11.5. Experiments were concluded at ZT16.5 by removing the ovi-cups, and eggs laid in ovi-cups were allowed to melanize overnight and manually counted the following day. QUANTIFICATION AND STATISTICAL ANALYSIS All statistical analyses were performed with GraphPad Prism or custom Python code. Statistical methods and sample sizes are included in figure legends. Nonparametric tests were used for data that does not follow normal distribution. Source data and statistical test details, including exact n and p values, can be found in Table S1. Supplementary Material SUPPLEMENTAL INFORMATION Supplemental information can be found online at https://doi.org/10.1016/j.celrep.2025.116663. KEY RESOURCES TABLEREAGENT or RESOURCESOURCEIDENTIFIER Chemicals, peptides, and recombinant proteins AgaroseSigma-AldrichA9539Sodium chlorideSigma-AldrichS9888Methyl 4-hydroxybenzoate (p-Hydroxy-benzoic acid methyl ester)Sigma-AldrichH5501 Deposited data Source dataThis paper Table S1 CodeThis paper https://github.com/DongLinhan/EggTimer2025 Experimental models: Organisms/strains Aedes aegypti: Liverpool wild typeDuvall LabN/AAedes aegypti: cycle mutantsShetty et al.23N/A Software and algorithms MozzieBox locomotor activity assay and analysisThis paper https://github.com/DongLinhan/EggTimer2025 MATLAB (2020b)The MathWorks, Inc. https://www.mathworks.com/ Ethovision (version 17)Noldus https://noldus.com/ethovision-xt R (version 4.2.3)The R Foundation https://www.r-project.org R “phase” package (version 1.2.9)The R Foundation https://cran.r-project.org/web/packages/phase/index.html Humidity-seeking assay and analysisThis paper https://github.com/DongLinhan/EggTimer2025 BioDare2BioDare2 https://biodare2.ed.ac.uk EggTimer oviposition monitor and analysisThis paper https://github.com/DongLinhan/EggTimer2025 GraphPad Prism (version 10.3.0)GraphPad Software https://www.graphpad.com/ Other TetraMin tropical granulesTetra16122Defibrinated sheep bloodQuad Five6105006-feeder Hemotek system with 3mL reservoirsHemotekAS6W1-3Miniature T-Slotted FramingMcMaster-Carr1959N1Corner Bracket for Single Rail Miniature T-Slotted FramingMcMaster-Carr1959N41L-Shaped Surface Bracket for Single Rail Miniature T-Slotted FramingMcMaster-Carr1959N43Heratherm™ Compact Microbiological IncubatorThermo Scientific50125590Infrared 850 nm IR LED Strip LightWaveform Lighting7041.850.5MNoctua NF-A6x25 FLX, Premium Quiet Fan, 3-Pin (60mm, Brown)AmazonB009NQMESSRaspberry Pi Camera Module 3 Wide NoIR - 12MP 120 Degree - Wide Angle Infrared LensAdafruit56601″ Diameter, Optical Cast Plastic IR Longpass FilterEdmund Optics43–498Whatman filter paper, 55 mmFisher Scientific09-805B12-Well Polystyrene Tissue Culture PlatesCellTreat229112Round Petri dish, 100 mm × 15 mmCellTreat229692Round Petri dish, 35 mm × 10 mmN/AKimwipesKimberly-Clark34120Plastic cup, 4 oz.Avantor89009–662Plastic cup, 8 oz.Avantor89009–664BugDorm-1 Insect Rearing CageBugDormDP1000BugDorm-1 Stockinet SleeveBugDormDP1000-N Highlights Aedes aegypti mosquitoes override post-biting activity suppression with hyperactivity Aedes aegypti mosquitoes override post-biting activity suppression with hyperactivity Humidity seeking and egg laying occur at nighttime in daytime-biting mosquitoes Humidity seeking and egg laying occur at nighttime in daytime-biting mosquitoes Hyperactivity supports long-range search of egg-laying site Hyperactivity supports long-range search of egg-laying site Cycle mutants show defects in behavioral timing and egg-laying efficiency Cycle mutants show defects in behavioral timing and egg-laying efficiency ACKNOWLEDGMENTS We thank Thomas Gabel for assistance with animal husbandry and Vinaya Shetty and Zach Adelman for providing the cycle mutant line. We thank Nils Reinhard, Ben Matthews, Paul Taghert, and members of the Duvall lab for useful discussions and feedback on the manuscript. This work was supported by the following grants: NIGMS (R35 GM137888) (L.B.D.), the Beckman Young Investigator Award (L.B.D.), the Pew Scholar in Biomedical Sciences Award (L.B.D.), and the Klingenstein-Simons Fellowship Award in Neuroscience (L.B.D.). Footnotes DECLARATION OF INTERESTS The authors declare no competing interests. REFERENCES Associated Data Supplementary Materials Data Availability Statement All data used in the main and supplemental figures are available in the supplemental information. All data used in the main and supplemental figures are available in the supplemental information. All original code used in this study can be found on GitHub and Zenodo (https://doi.org/10.5281/zenodo.17507875). All original code used in this study can be found on GitHub and Zenodo (https://doi.org/10.5281/zenodo.17507875). Any additional information required to re-analyze the data reported in the paper is available from the lead contact upon request. Any additional information required to re-analyze the data reported in the paper is available from the lead contact upon request.