Temporal Bet-Hedging in Sand Fly Oviposition: Pharate Phlebotomus papatasi Sand Fly Neonates Regulate Hatching Time in Response to Organic Matter and Proximity to Conspecific Eggs

Abstract

In phlebotomine sand fly, the topic of the factors stimulating or inhibiting egg hatching has been largely ignored. In this study, we evaluated the hypothesis that pharate neonate sand flies are able to regulate their hatching time adaptively in response to cues indicative of the presence of food or potential competitors. In this study, we evaluated the independent and combined effects of organic matter (OM) extract and proximity to conspecific eggs on the hatching proportion of Phlebotomus papatasi eggs. In one set of larval rearing jars, we introduced 16 eggs into a small hole in the center of a plaster base of the jar. In another set, we introduced a single egg into each small hole of a 4-by-4 symmetrical array. To one set, we added an aqueous OM extract, and to the other, we added deionized water (DI). OM stimulated egg hatching while egg clustering slightly inhibited egg hatching. Results of this experiment are biologically important because they show, for the first time, that pharate sand fly neonates are able to adaptively regulate their hatching time in response to external cues.

Risk spreading in time or space (known as “bet-hedging”) is a critical life-history strategy that enhances the lifetime reproductive success of organisms in stochastic environments (Hopper 1999). Hematophagous and other insect types may spread the risk in space by distributing eggs of a single clutch among several oviposition sites (“skip oviposition”) or in time using “environmentally cued delayed hatching” (Warkentin 2011, Day 2016). With mosquitoes, it was shown that water characteristics, such as temperature and flooding frequency, may affect the proportion of eggs hatched (Day 2016). In addition, bacteria and bacterially derived compounds, as indicators of larval food, were shown to stimulate egg hatching (Ponnusamy et al. 2011, Day 2016). The effect of conspecific eggs or larvae on mosquito egg hatching was shown to be quite complex with conspecific eggs reducing hatching proportion (Gillett 1959), but conspecific larvae having density-dependent or stage-specific effects (Livdahl et al. 1984, Edgerly and Marvier 1992).

Phlebotomine sand flies (Diptera: Psychodidae) are hematophagous insects that transmit protozoan parasites (Leishmania spp.), as well as bacterial (Bartonella bacilliformis) and also viral pathogens (Killick-Kendrick 1999, Ready 2013). Most significant are the human Leishmaniases. Phlebotomine sand flies can be found in warm arid to tropical regions around the world (Killick-Kendrick 1999). Female sand flies lay 40–70 eggs in sheltered organic matter (OM), rich and moist microhabitats, and under optimal conditions, eggs hatch within 6–17 days with hatching time affected by climatic and physiological factors such as temperature, moisture, and insemination of females (Killick-Kendrick 1999, Ready 2013, Lawyer et al. 2017).

In terms of oviposition site selection, it was shown that gravid sand flies are attracted to OM of various sources (Elnaiem and Ward 1992, Dougherty et al. 1995, Marayati et al. 2015). With respect to the effect of conspecifics, eggs were shown to be attractive to both new world and old world sand flies with dodecanoic acid being the main driver of this attraction (Elnaiem and Ward 1991, Dougherty et al. 1994, Srinivasan et al. 1995, Dougherty and Hamilton 1997, Kowacich et al. 2020). Recently, Kowacich et al. (2020) showed that sand flies are attracted to conspecific eggs at low/medium densities but are repelled at high densities. However, as of yet, it was not evaluated if pharate neonate sand flies are able to regulate their hatching time adaptively in response to cues indicative of the presence of food or potential competitors.

In this study, we evaluated the effect of OM and proximity to conspecific eggs on the hatching proportion of P. papatasi eggs. We hypothesized that, given the coprophagic diet of sand fly larvae, OM extract should stimulate egg hatching as an indicator of larval food. However, the effect of conspecific eggs could either be stimulating by indicating larval site suitability or inhibition by indicating a potential competitive risk.

Materials and Methods

Insects and colony maintenance

Phlebotomus papatasi sand flies that originated from Abkük, Turkey (April, 2004) were maintained following the mass-rearing technique described by Lawyer et al. (2017). The adults were blood fed on live anesthetized ICR mice (Envigo, Indianapolis, IN) (SoBran protocol no. UNC-002-2019). Immature and adult sand flies were maintained in environmental chambers (Caron^®, Marietta, OH) at 26°C, 85% relative humidity (RH) under a 14-h light/10-h dark reverse photoperiod.

Study design

We implemented a factorial design to test the independent and combined effects of OM and proximity to conspecific eggs. Experiments were conducted in 500-mL Nalgene jars (Model 81063, diameter = 11 cm; Nalgene™) with a 2.2 cm layer of Whip-Mix^® Orthodontic Plaster (Model: 5577352; Henry Schein, Inc., Melville, New York) on the bottom. In one set of jars, we drilled a shallow (2 mm deep, 5 mm diameter) hole in the center of the plaster base, and in the other set, we drilled a centralized 4-by-4 grid of small holes (1 mm diameter, 1 mm deep, 2 cm apart). We introduced 16 eggs to each jar. In the first set, all eggs were clustered in the central hole (hereafter, “Cluster” [C] dispersion), whereas in the other, a single egg was inserted gently into each hole in the 4-by-4 grid (hereafter, “Even” [E] dispersion). For the OM treatment, we used 0.1 g of used larval rearing medium mixed thoroughly with 1 mL of deionized water (DI) in a 7-mL scintillating vial. The OM was then allowed 5 min to settle, and the supernatant was used as our aqueous OM extract. OM extract or DI water control was then dispensed to each experimental jar at a dose of 50 μL of OM extract or DI water. In total, we had four treatment combinations: even egg dispersion with OM extract (E+), even egg dispersion with DI water (E-), clustered egg dispersion with OM extract (C+), and clustered egg dispersion with DI water (C-). Hatching status of the eggs was evaluated visually under a microscope (6.7 × to 45 × ; Olympus SZ61, Center Valley, PA). Eggs used were estimated (based on Marayati et al. 2015) to be 5 days old or younger at the beginning of the experiment. Four to six replicate jars per treatment were used in each replicate session. Replicate sessions took place between January 2019 and February 2020 for a total of six replicate sessions (n = 29 per treatment).

Data analyses

Our response variable for all analyses in this study was the proportion of eggs hatched. With “daily hatching proportion,” we measured the number of unhatched eggs by the end of the previous day (as the denominator) and counted the number of hatched eggs on the following day. With “cumulative hatching proportion,” we summed the number of hatched eggs to a particular day and divided it by the original number of eggs, which was 16. We used a weighted logistic regression, with proportion of hatched eggs as the response variable and egg dispersion (“E” or “C”), OM treatment (“-” or “+”), and “Day” as predictor variables. Number of unhatched eggs in a jar was used as the weighting factor (W). With “daily hatching proportion” W was the number of unhatched eggs by the end of the previous day and with “cumulative hatching proportion” W = 16.

Results

Daily hatching proportion

OM extract had a highly significant positive effect (odds ratio [OR] = 1.28) (Fig. 1A) on daily hatching proportion (Table 1A). The effect of egg clustering was negative (OR = 0.88) (Fig. 1A) but only marginally significant (Table 1A). These effects were approximately consistent across all 4 days (Fig. 1B) of the experiment with no significant statistical interactions for any factor with respect to time. Specifically, the effects of OM and clustering were not statistically significant on day 1, but they were highly significant on day 2 (p ≤ 0.001). On days 3 and 4, the effects of OM and clustering were not statistically significant and on day 4, the difference in hatching proportions among treatments appeared to shrink (Fig. 1B). On average, in reference to day 1, hatching proportion slightly, but significantly (Table 1A), increased at day 2 (OR = 1.55) but then substantially increased on day 3 (OR = 10.38) and day 4 (OR = 19.1) (Fig. 1B and Table 1A).

FIG. 1.

Daily egg hatching. The effects of treatment of eggs with OM aqueous extract (OM+) or water (OM-) and egg dispersion (C = clustered, E = evenly dispersed) on the mean daily hatching proportion of Phlebotomus papatasi eggs (A). This figure (B) also depicts the temporal dynamics of daily hatch with respect to the effects of organic water extract and egg spatial dispersion (C− = clustered eggs treated with DI water; C+ = clustered eggs treated with OM aqueous extract; E− = evenly dispersed eggs treated with DI water, E+ = evenly dispersed eggs treated with OM aqueous extract). Daily hatching proportion was calculated based on the number of eggs hatched in an experimental jar at a particular day divided by the number of unhatched eggs in the previous day. The * and ф markers indicate significant effects of OM extract or egg clustering, respectively, at each of the study days. *(or ф) = p < 0.01, **(or фф) = p < 0.001, ***(or ффф) = p < 0.0001. DI, deionized water; OM, organic matter.

Table 1.

Logistic Regression of the Effects of Organic Matter, Egg Clustering, and Day on Daily (A) and Cumulative (B) Egg Hatching Proportion

Parameter	Coefficient	SE	z	p
A
Intercept	−2.666	0.104	−25.689	<0.0001
Organic matter	0.245	0.067	3.609	0.0003
Egg clustering	−0.118	0.068	−1.745	0.081
Day 2	0.444	0.121	3.687	0.0002
Day 3	2.342	0.104	22.315	<0.0001
Day 4	2.955	0.114	25.795	<0.0001

B
Intercept	−2.682	0.101	−26.484	<0.0001
Organic matter	0.314	0.060	5.254	<0.0001
Egg clustering	−0.165	0.059	−2.769	0.0056
Day 2	0.986	0.111	8.915	<0.0001
Day 3	2.721	0.102	26.490	<0.0001
Day 4	4.032	0.108	37.035	<0.0001

Cumulative hatching proportion

Consistent with the above, both OM (OR = 1.36) and clustering (OR = 0.84) had statistically significant effects (Table 1B) with positive and negative effect of OM and clustering, respectively (Fig. 2A). Despite no statistically significant interaction between “day” with any of the other factors, the effects of OM and clustering varied slightly among the four days of the experiment (Fig. 2B). The effects of OM and egg clustering were not significant on day 1, but they were highly significant on day 2 (p ≤ 0.001) (Fig. 2B). On days 3 and 4, only the effect of OM was significant (Fig. 2A, B). Final average cumulative proportion of eggs hatched by day 4 of the experiment was equal to 80% (±1.28%).

FIG. 2.

Cumulative egg hatching. The effects of treatment of eggs with OM aqueous extract (OM+) or water (OM-) and egg dispersion (C = clustered, E = evenly dispersed) on the mean cumulative hatching proportion of Phlebotomus papatasi eggs (A). This figure also depicts the temporal dynamics of mean cumulative egg hatch with respect to the effects of organic water extract and egg spatial dispersion (C− = clustered eggs treated with DI water; C+ = clustered eggs treated with OM aqueous extract; E− = evenly dispersed eggs treated with DI water, E+ = evenly dispersed eggs treated with OM aqueous extract). Cumulative egg hatching proportion was calculated as the cumulative number of eggs hatched per experimental jar until a certain day divided by the initial number of eggs, which was 16. The * and ф markers indicate significant effects of OM or egg clustering, respectively, at each of the study days: *(or ф) = p < 0.01, **(or фф) = p < 0.001, ***(or ффф) = p < 0.0001.

Discussion

Results of this experiment are biologically important because they show, for the first time, that pharate sand fly neonates are able to regulate their hatching time in response to external cues. As expected, hatching of eggs was stimulated by OM aqueous extract, which could indicate the presence of food for the coprophagic larvae. Recently, we showed that oviposition site selection of P. papatasi is mediated by several bacterial species isolated from the same larval rearing substrate used to produce the extract used in this study (Kakumanu et al. 2020). However, the role of bacteria and the compounds they produce in stimulating egg hatching remains to be studied. As demonstrated by Ponnusamy et al. (2011) with mosquitoes, bacteria that constitute an important food source for larvae, stimulate egg hatching directly by producing compounds that stimulate hatching and indirectly by reducing dissolved oxygen level. It would be interesting to evaluate if changes in oxygen level also affect sand fly egg hatching. This is not unlikely given that eggs are often laid in poorly oxygenated underground environments.

Egg hatching tended to be inhibited when eggs were aggregated compared with when eggs were evenly dispersed. This trend was exhibited for both daily and cumulative egg hatch proportion but was statistically significant for the latter. This observation is consistent with the possibility that egg aggregation may signal the potential for larval competition. Hence, it would be adaptive for a pharate neonate to delay its hatching to times when this risk is reduced. This effect, however, might be density dependent being possibly weaker at lower egg density and stronger at a higher egg density. With oviposition site attraction, we recently showed that gravid P. papatasi females exhibit a hump-shaped dose response with respect to conspecific egg density (Kowacich et al. 2020). The egg density used in this study (16 eggs/jar) corresponds to a region of neutral attraction, which would suggest that gravid females and pharate neonates have a differential perception of the risks and benefits associated with the presence of conspecifics as it related to oviposition site suitability. Competitive effects on neonates are expected to be particularly adverse in the presence of older larval stages that are known to cannibalize on younger larvae (Lawyer et al. 2017). Indeed, we recently showed that gravid females are attracted to oviposition sites with eggs and first instar larvae but are repelled/deterred from sites containing older stages (Kowacich et al. 2020). Nonetheless, this negative effect of conspecific eggs observed in this study is not trivial given that larval aggregation is often thought to be beneficial. This is because single neonate larvae tend to get entangled in fungal hyphae, whereas a group of them can chew their way out (Lawyer et al. 2017). As in mosquitoes (Livdahl et al. 1984), these intraspecific interactions are probably complex, and their effects are likely to be density and stage specific. Therefore, more research into this novel topic is needed.

Daily hatching proportion differed substantially among the 4 days of the experiment. Eggs at day 1 of the experiment were estimated to be 5 days old or younger. Daily hatch proportion was low on days 1 and 2 of the experiment (7–11%), but then increased sharply on day 3 (46%) and peaked on day 4 (58%). This is consistent with previously published results, indicating eggs tend to hatch 7–10 days postoviposition (Killick-Kendrick 1999). By the final day of the experiment (∼9 days postoviposition) about 80% of all eggs have hatched.

The effects of OM extract and conspecific aggregation, as observed in this study, were statistically (or marginally) significant but were not of a large magnitude. This might be because, overall, eggs in this experiment were maintained under fairly optimal abiotic (warm and moist) and biotic (relatively low egg density [16 eggs per jar]) conditions. It is possible that under more stressful conditions (e.g., drier environment, higher conspecific density), these factors would induce stronger effects. It is also possible that the fact that we used mature sclerotized eggs could partially explain their relatively low sensitivity. It would be worth repeating these experiments using newly laid unsclerotized eggs to evaluate if, indeed, pharate neonates are more sensitive to the presence of OM or conspecific eggs. Nonetheless, the consistency of our results indicates that pharate sand fly neonates are able to regulate their hatching time adaptively in response to external cues. Phlebotomus papatasi and other sand fly species tend to lay their eggs, especially in arid areas, inside rodent burrows that provide immature stages with a moist and cool microclimate as well as with food in the form of decomposing fecal matter (Wasserberg et al. 2003, Mascari et al. 2013). However, these habitats tend to be ephemeral as burrows are often deserted due to rodent host's death or dispersal (Wasserberg et al. 2003). Hence, it makes adaptive sense for pharate neonates to maintain a certain degree of phenotypic plasticity, enabling them to regulate hatching to times when conditions are more suitable. Finally, understanding the mechanisms that regulate sand fly egg hatching could have significant epidemiological consequences given that such factors (e.g., soil eutrophication through application of organic fertilizers into agricultural fields) may possibly synchronize egg hatching across the landscape leading to sand fly population outbreaks as often described following anthropogenic disturbances (Wasserberg et al. 2003).

Author Disclosure Statement

No conflicting financial interests exist.

Funding Information

This study was partially supported by grant number R01AI123327 to Gideon Wasserberg from the National Institute of Health: National Institute of Allergy and Infectious Diseases. Teaching assistantship to Dannielle Kowacich was funded by UNC-Greensboro’s Biology department. Hieu Nguyen was partially funded by UNC-Greensboro’s Undergraduate Research and Creativity Award.