Aedes albopictus (Diptera: Culicidae) oviposition response to organic infusions from common flora of suburban Florida
We evaluated the oviposition response of gravid Aedes albopictus (Skuse) to six organic infusions. Laboratory and field-placed oviposition cups baited with water oak (Quercus nigra L.), longleaf pine (Pinus palustris P. Mill), or St. Augustine grass (Stenotaphrum secundatum (Walt.) Kuntze), as well as binary infusion mixtures of each, were used. In addition, a triple-cage, dual port olfactometer was used to measure upwind response of gravid individuals to these infusions. We found that Ae. albopictus deposited more eggs in infusion-baited cups compared with water alone. Moreover, significantly more eggs were laid in the water oak and a water oak-pine mixture as compared with the St. Augustine grass infusion in laboratory bioassays. However, a negative upwind response was observed with longleaf pine infusion in the olfactometer. In field cages, significantly more eggs were deposited in infusion-baited cups as compared with water alone and a greater percentage of eggs were deposited in cups containing a water oak and the water oak-longleaf pine mixture as compared with cups containing single infusions or their mixtures.
An array of chemicals and environmental factors are known to influence mosquito oviposition behavior, and consequently their ultimate site selection, by orienting them toward (attractant) or away (repellent) from an oviposition source, and/or eliciting an oviposition event (stimulant) (Bentley and Day 1989). Many of these chemical cues include pheromones associated with eggs (Starratt and Osgood 1973), carboxylic acids and methyl esters associated with bacterial digestion of organic materials (Ponnusamy et al. 2008), and various naturally occurring bacteria in larval water (Hasselschwert and Rockett 1988). Organic infusions, commonly developed from a range of fermented plant material to animal waste products, are frequently used to increase the attraction of gravid mosquitoes to ovitraps and gravid traps (Service 1993). Oviposition response is largely due to the detection of semiochemicals that may act as attractants or repellents emanating from the mixture (Trexler et al. 2003). Hazard et al. (1967) identified Aerobacter aerogenes from a hay infusion as being responsible for stimulating mosquito oviposition in Aedes aegypti L. and Culex pipiens quinquefasciatus (Say). Subsequently, a range of bacteria found in mosquito larval-rearing water have been identified as being responsible for producing stimulants that have included Enterobacter cloacae, Acinitobacter calcoaceticus, and Psychrobacter immobilis (Benzon and Apperson 1988, Trexler et al. 2003). Moreover, Trexler et al. (2003) found that Sphingobacterium mulivaroum found in soil-contaminated cotton towels also acted as an oviposition stimulant.
Laboratory and field studies have previously demonstrated that Aedes albopictus (Skuse) lays significantly more eggs in ovitraps containing white oak leaves (Quercus alba L.) (Trexler et al. 1998), maple leaves (Acer buergerianum) (Dieng et al. 2002), guinea grass (Panicum maximum) (Santana et al. 2006), and Bermuda grass (Cynodon dactylan) (Zhang and Lei 2008) than in water-only controls. Gravid trap studies using red oak leaf (Quercus rubra L.)-baited infusions have also reported greater captures of adult Ae. albopictus compared with standard hay infusions (Burkett et al. 2004). Although these studies have shown enhanced oviposition by Ae. albopictus using plant infusions, there is limited information on attractiveness of infusions utilizing coniferous tree substrates to mosquitoes in general. Schreiber et al. (1996) provides one such example, demonstrating that tires containing slash pine (Pinus elliotti Englem) needles supported a greater number of mosquito larvae than those containing live oak leaves (Quercus virginiana L.). However, Ae. albopictus attraction, repellency, and stimulant responses to infusion mixtures from several plant species remains poorly understood.
Many suburban backyards in north-central Florida contain a mixture of water oak (Quercus nigra L.), longleaf pine (Pinus palustris P. Mill), and St. Augustine grass (Stenotaphrum secundatum (Walt.) Kuntze). Subsequently, fallen leaves from these trees and grass cuttings from lawn mowers frequently collect in rain-filled containers, providing ideal larval habitats for Ae. albopictus and other container-inhabiting mosquitoes. We examined the ovipositional response of Ae. albopictus to single species infusions of water oak, longleaf pine, and St. Augustine grass as well as binary mixtures of each as potential oviposition attractants. In addition, a triple-cage, dual port olfactometer was used to measure upwind response of gravid Ae. albopictus to these infusions.
MATERIALS AND METHODS
Leaves of water oak (WO) and needles of longleaf pine (LP) trees were collected from the grounds at the University of Florida, while St. Augustine grass (SAG) (‘Bitterblue’ cultivar) was cut with a lawnmower at the primary author's residence. These collections were used to manufacture all infusions. All sites were in the vicinity of Gainesville, FL. Special attention was taken to ensure all leaves and needles were free of foreign organic matter. Fresh-cut SAG was placed on a sheet and dried for four days under natural sunlight. Four batches of infusions were prepared by fermenting 25 g of dried leaves, 2.5 g brewer's yeast (MP Biomedicals, LLC, Solon, OH), and 2.5 g lactalbumin (Sigma-Aldrich, St. Louis, MO) in 2.5 liters of well water, approximating methods of Allan and Kline (1995) and held at (25−27° C) for ten days in a sealed plastic bucket. Infusions were first passed through sterile gauze dressing, transferred to 150 ml polypropylene cups, and sealed and frozen at −20° C. Infusion mixtures consisted of a 50:50 ratio of water oak-longleaf pine (WO-LP), water oak-St. Augustine grass (WO-SAG), and longleaf pine-St. Augustine grass (LP-SAG). On the day of testing, frozen aliquots were placed in a warm bath for 30 min or until they melted.
All mosquitoes used in the study were from a laboratory colony established in 2007 from field-collected eggs from sites in northern Florida. Similar methods from Gerberg et al. (1994) were used in maintaining the colony. Larvae were reared on finely ground TetraFin™ goldfish flakes (Tetra®, Blacksburg, VA), while female adults were fed defibrinated bovine blood as a protein source. All mosquitoes were 13 days of age (post-eclosion).
Laboratory cage bioassays
Bioassays were conducted in aluminum cages (30 × 30 × 30 cm) constructed of four clear Plexiglas® sides, with a gauze access sleeve at one end and window screening (80 × 80 mesh) at the opposite side. Each cage contained two black plastic oviposition cups (156 ml), one that contained an infusion and one with well-water only as a control. Infusions were diluted to a 10% concentration prior to the study, as previous laboratory experiments using standard hay infusions demonstrated that concentrations above 20% were repellent4.
Both cups contained a 6 × 4 cm strip of 76 seed germination paper (Anchor Paper, St. Paul, MN) for mosquito oviposition. Cups were diagonally set 14 cm apart from one another in each cage and the position of treatment and control (right or left) was noted and alternated to eliminate bias. These infusion choice-tests used gravid Ae. albopictus that were produced as follows: four days prior to testing, females from the F10 to F12 generations were exposed to sausage casings containing defibrinated bovine blood that had been placed in a 34° C water bath for 5 min. Casings were patted dry and suspended inside of a cage to facilitate feeding. This procedure was repeated every 20 min until all feeding activity ceased.
On the day of testing, ten gravid mosquitoes were aspirated from a chill table and placed inside each cage. Cages were stacked up to four high and completely randomized. Mosquitoes were permitted to oviposit in the plastic cups for 24 h, after which time the seed germination paper was collected. After drying, papers were placed into sealable plastic bags. Eggs were counted using a dissecting microscope and recorded into a spreadsheet. Each trial consisted of 35 cages per trial (six treatments and one control) and the experiment was repeated five times. Experiments were conducted from January to April 2008.
To determine if the infusion treatments elicited an upwind response, gravid Ae. albopictus were tested in a clear acrylic triple-cage dual port olfactometer (Posey et al. 1998) from November 2007 through March 2008 using procedures similar to those described by Allan and Kline (1995). Conditions inside the olfactometer were 28° C and 85% relative humidity. A micropipette was used to extract 500 µl of 100% infusion concentrate from the top of each aliquot and placed in the bottom of a 50 × 2.5 mm watch glass that was set inside separate arms of the olfactometer. A total of 50 gravid females (4-days post-blood fed) from the F5 through F11 generations were aspirated from a chill table, transferred to the olfactometer, and allowed a 1–h pretreatment acclimatization period. The number of dead mosquitoes was noted and not included in the analysis. Following acclimation, both arms of the olfactometer were opened, allowing access from the mosquito-containing chamber. A one liter/s airflow was passed over each treatment infusion and well water control. Olfactometer runs were conducted for 10 min, after which doors were closed and the number of females in each chamber counted. At the conclusion of three consecutive runs (one run per chamber), mosquitoes were aspirated out of the arms and placed back into the chamber and treatments were randomized using a new chamber that ensured every treatment was exposed in each chamber. Due to time constraints, a trial was conducted over two consecutive days. On the first day, three of the six infusion treatments were randomly selected and used for three runs on that day. The following day the remaining three treatments were evaluated in the olfactometer. Trials were replicated eight times for a total of n=24 observations for each treatment.
Field cage bioassays
Outdoor cage trials were conducted from June to September 2007 using four circular screened cages (2.13 m high × 2.74 m diam) constructed of a PVC pipe frame (2.54 cm diam) and screening (18 × 14 mesh). Cages were linearly set and spaced 1.82 m apart in a semi-shaded environment. One flowering Gardenia jasminoides J. Ellis, approximately 1 m in height, was placed in the center of the cage to provide resting sites. Two plastic cups containing cotton balls soaked in a 5% sucrose solution were placed in the cages to provide a carbohydrate resource.
Oviposition was monitored using 11 × 9 cm black plastic cups with a 1 cm diameter drainage hole positioned 5.5 cm from the bottom. Seed germination paper was cut into 10 × 10 cm squares and pressed against the inside surface of each cup. Cups were filled with 200 ml of a 20% infusion (40 ml) water or well water only. This concentration was based on previous field studies demonstrating gravid Ae. albopictus were most attracted to hay infusion concentrations between 20 and 30%, as compared with 10% used in laboratory studies4.
Two trials were conducted using four cages, each with three single-source infusion treatments (WO, LP, and SAG) and a well water control (total 4 cups per cage). An additional two trials were conducted similar to the previous trials with the three 50:50 infusion mixtures (WO-LP, WO-SAG, LP-SAG, and well water control).
Oviposition cups were placed on top of concrete blocks (19.05 cm in height), positioned approximately 1 m from the center of the cage and 1 m apart at 90° angles to one another. A 14×14 cm nylon curtain fabric was placed over the top of each oviposition cup to prevent mosquitoes from prematurely depositing eggs within the cup before fully acclimatizing to the cage. Previously blood-fed (3 days prior) Ae. albopictus from F2 through F4 generations were placed on a chill table to sort for gravid females. One hundred gravid females were released in the center of each cage, and were permitted to acclimatize for 1 h, at which point oviposition cup covers were removed by pulling an attached string. After 48 h, oviposition papers were collected and eggs counted using a dissecting microscope.
A completely randomized design was used in the laboratory bioassay cage experiments. Treatment, trial, and infusion batch were fixed effects and a treatment by trial interaction term was included in the model. A paired t-test was first conducted on raw means to determine well water and infusion treatment differences within each cage. The total amount of eggs deposited in treatments was transformed with log10 (n+1) and analyzed by analysis of variance (ANOVA) to detect differences between fixed effects.
A randomized block design was used to analyze the olfactometer bioassay. Oviposition response was measured as the percentage of mosquitoes that responded positively to the treatment. Abbott's correction was used to adjust data for those mosquitoes that flew into water-only control arms of the olfactometer (Abbott 1925). Means were square-root transformed and analyzed by ANOVA to detect differences between fixed effects. A blocked design was used to mitigate potential differences caused by aspirating mosquitoes back and forth between runs. Treatment, infusion batch, olfactometer chamber, block, and mosquito generation were fixed effects in the model.
A randomized block design was used for field cage trials. Treatments were randomized in the cage to eliminate position bias. Statistical analysis was conducted using methods similar to Allan et al. (2005). To determine infusion attractancy, or repellency, the total number of eggs laid on each seed-germination paper was divided by the total number of eggs laid in the respective cage (treatment + control). An arcsine transformation was performed on the percentage of eggs deposited in each cup and analyzed by ANOVA. Treatment, cage, trial, and infusion batch were fixed effects in the model. All statistical analyses were conducted using the PROC GLM procedure of SAS (SAS 2006). Multiple mean comparisons were made with the Student-Newman-Keuls multiple range test. For all analyses, differences were considered significant when α=0.05.
Laboratory cage bioassays
Significantly more eggs were deposited in all infusion treatments when compared with the well water control (Table 1). Oviposition cups containing LP and WO-LP infusion had the greatest number of eggs. While no significant differences were detected among infusion batches, significant differences were present between treatments (F = 2.91; df = 5, 115; P= 0.0163) (Table 1). Significantly more eggs were deposited in cups containing LP and WO-LP than those containing SAG. Also, there were significant differences within trial (F = 13.17; df = 4, 115; P <0.0001) and treatment trial interactions (F = 1.88; df = 20, 115; P= 0.0203).
Table 1. Oviposition response of 4-day-old gravid Ae. albopictus to six infusions1 and a well water control in laboratory cage bioassays.
|Pine||25||276.6 (17.3) a||84.72 (7.06)||24||6.57||0.0001|
|Oak-Pine||25||259.6 (17.3) a||110.60 (8.80)||24||8.89||0.0001|
|Oak||24||238.3 (17.8) ab||95.16 (10.7)||23||8.89||0.0001|
|Pine-Grass||25||230.8 (22.0) ab||86.20 (11.0)||24||8.47||0.0001|
|Grass-Oak||24||224.6 (15.0) ab||81.80 (8.80)||23||11.10||0.0001|
|Grass||25||212.8 (20.3) b||57.00 (11.4)||24||6.57||0.0001|
|Well water||25||102.8 (12.0) NI*||114.80 (11.4)||24||–1.53||0.1390|
Gravid Ae. albopictus exhibited a significantly stronger upwind response to all infusion treatments, with the exception of LP, when pair-wise comparisons with their well water controls were considered (Table 2). Although not statistically different, more mosquitoes responded positively to the LP infusion compared with well water only. While no significant differences were present between infusion batches or blocks, significant differences were detected between mosquito generations (F = 5.53; df = 5, 126; P= 0.0001), with earlier generation mosquitoes most responsive. All infusions elicited a greater attraction response by Ae. albopictus than the LP infusion (F = 4.54; df = 5, 126; P= 0.0006) (Table 2).
Table 2. Upwind response of 4-day-old post-blood fed Ae. albopictus to 500 µl of 100% infusion concentrate compared with well water control inside a dual-port olfactometer for 10 min.
|Oak-pine||24||4.85 ± 0.76 a||1.41 ± 0.25||23||4.815||<0.0001|
|Grass||24||3.92 ± 0.52 a||1.37 ± 0.26||23||4.155||<0.0004|
|Oak-grass||24||3.81 ± 0.68 a||0.83 ± 0.20||23||3.974||<0.0006|
|Oak||24||3.28 ± 0.56 a||1.08 ± 0.26||23||3.795||<0.0009|
|Pine-grass||24||3.35 ± 0.57 a||0.83 ± 0.23||23||4.047||<0.0005|
|Pine||24||1.15 ± 0.78 b||2.00 ± 0.49||23||–0.730||0.4727|
Field cage bioassays
Significantly more eggs were observed in oviposition cups containing an infusion than in cups with well water only (F = 8.68; df = 6, 39; P< 0.0001). While no significant differences were observed between infusions, a greater percentage of eggs were deposited in cups containing WO and the WO-LP mixture compared with cups containing single infusions or their mixtures (Table 3).
Table 3. Mean 48-h oviposition response of 100 (3-day-old post-blood fed) gravid Aedes albopictus released into field cages with oviposition cups containing infusion water1 of varying composition and a well water control, Gainesville, FL, June September 2007.
|Oak||6||37.1 (3.13) a||F6,39= 8.68, P <0.0001|
|Oak-Pine||6||35.4 (4.01) a|| |
|Oak-grass||6||30.5 (4.60) a|| |
|Pine-Grass||6||28.8 (2.50) a|| |
|Grass||6||28.7 (2.73) a|| |
|Pine||6||26.5 (2.44) a|| |
|Well water||12||13.4 (2.30) b|| |
Our results demonstrated that gravid Ae. albopictus exhibited an oviposition preference for longleaf pine and water oak-longleaf pine infusions compared with St. Augustine grass in laboratory bioassays. Indeed Burkett et al. (2004) found similar preferences with red oak leaf infusions compared with hay or grass. However, when we compared the LP infusion with upwind observations in the olfactometer and field cage oviposition trials, a different set of results were encountered. We believe that the contradictions in attractancy between our three methods of evaluation may be due to differences in concentration. Laboratory experiments contained 10% dilutions of the LP infusion, whereas field cage and olfactometer experiments utilized 20% and 100% concentrations, respectively. Moreover, longleaf pine infusions may contain volatile oils that may repel Ae. albopictus oviposition response at higher concentrations. This may be due to nerol, a compound found in pine oil and reported as a repellent by Butler (2007). However, the fact that the WO-LP and LP-SAG infusions elicited a greater upwind response in the olfactometer as compared with LP suggests that either pine oil was masked by the other organic ingredients in the olfactometer, or that in laboratory and outdoor cage trials the pine oil concentration had been reduced from a repellent concentration (100%) used in the olfactometer to a level below repellency in the non-olfactometer studies (10–20%). Either of these would alter or negate the repellent properties observed in the olfactometer. Significant differences between trials may have been the result of using different mosquito generations, as earlier generation mosquitoes were more responsive to infusions compared to later ones. Slight differences in infusion concentration or bacteria load may have altered mosquito response in several treatments, causing differences within treatments and trials.
There are a number of variables that may alter the degree of infusion attractiveness. Protein concentration and bacteria levels are known to transform an infusion from an attractant to a repellent (Gubler 1971) and their subsequent levels are dependent on several factors. For example, the presence of bacteria isolated from oak leaves has been shown to have a positive influence on Ae. albopictus oviposition behavior (Trexler et al. 2003), while those of alder leaf extracts have deleterious effects on larvae (David et al. 2000). A second factor that may change the attractiveness of an infusion is the duration of fermentation. Kramer and Mulla (1979) reported that while Cx. p. quinquefasciatus was not attracted to 5-day-old, 1% chicken manure, a positive response was observed on day 6 and the response gradually increased until day 11. Santana et al. (2006) later demonstrated that Ae. albopictus were most attracted to guinea grass infusions fermented from 15 to 20 days compared with those fermented for 30 days. Finally, the stage at which the leaves are used may produce different levels of chemical cues. For example, Santana et al. (2006) also demonstrated that Ae. albopictus deposited more eggs in infusions made from fresh guinea grass leaves than from dried leaves.
Organic infusions have successfully been used in ovitraps for surveying populations of Ae. aegypti (Reiter et al. 1991), but the application of these oviposition attractants may also serve as potential control measures. Aedes albopictus employs an oviposition behavior called “skip oviposition” whereby eggs are distributed throughout several sites (Burkett-Cadena and Mullen 2007). This strategy to distribute their eggs over an area may act as a survival mechanism (Rozeboom et al. 1973), unlike Culex spp. which deposit all of their eggs in one location. Therefore, infusion-baited ovitraps that attract and stimulate Ae. albopictus oviposition may mitigate this behavior by discouraging them from utilizing less attractive breeding sites, thereby concentrating their eggs to just a few containers (Ponnusamy et al. 2008). Based on our results, we recommend that ovitraps used for surveying Ae. albopictus be baited with oak or oak-pine infusions rather than using water alone. However, the fact that Ae. albopictus demonstrated little oviposition preference for one organic infusion over another is not surprising as Burkett-Cadena and Mullen (2007) and Zhang and Lei (2008) reported similar findings. Further research is clearly warranted in selecting an optimum infusion concentration at which maximum oviposition attraction can be attained. Moreover, future studies should identify the responsible bacteria or volatile compounds from organic infusions that attract mosquitoes, especially invasive species, as these organic attractants could be mass-produced and used as surveillance tools.
Obenauer, P. 2009. Surveillance of Aedes albopictus (Skuse) (Diptera: Culicidae) suburban and sylvatic populations using traps and attractants in north central Florida. Ph.D. dissertation, University of Florida, Gainesville.
We thank Erin Vrzal and Catherine Zettel-Nalen for assistance with laboratory and field experiments. We are grateful to Dr. John Sivinski for providing field cages and Charlie Stuhl with their set-up. This study was supported by the University of Florida Agricultural Experiment Station federal formula funds, Project No. FLA-04598, received from the Cooperative State Research, Education and Extension Service, U.S. Department of Agriculture. The views expressed in this article are those of the author and do not necessarily reflect the official policy or position of the Department of the Navy, Department of Defense, nor the U.S. government.