xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xlink http:www.w3.org1999xlink xmlns:xsi http:www.w3.org2001XMLSchema-instance
leader 00000nas 2200000Ka 4500
controlfield tag 008 000000c19749999pautr p s 0 0eng d
datafield ind1 8 ind2 024
subfield code a M39-00285
Muldoon, James, H.
Decisiones de los gusanos len (Diptera Vermileonidae) con los efectos de la humedad del sustrato y la densidad de las larvas en la construccin de trampas
Pit-trap construction decisions by wormlions (Diptera Vermileonidae) effects of substrate moisture and larval density
During their sessile larval stage, wormlions (Diptera vermileonidae) provide a unique insight into the importance of habitat selection because their local conditions greatly influence their growth and
survivorship. Like their ecological relatives, antlions (Neuroptera myrmeleontidae), wormlion larvae are a
sit-and-wait predators that construct conical pits to capture their prey. In this study, I demonstrate effects of moisture on larval pit-trap construction to explain the presence of and potential problems with wormlion aggregations. I found that larvae prefer drier soil conditions, for pit presence and size was negatively affected by large amounts of water. Results suggest larvae lack the strength to manipulate the compact, more water-saturated soil and must wait until soil conditions improved to construct their pits. Because dry spaces are limited in rainforests, wormlion larvae are likely to live in dense aggregations. To determine the effects of varying densities on larval aggregations, I compared the number of pits constructed within high
concentrations to that observed in the field. As wormlion density increased, pit establishment decreased. Lower pit success in higher densities may be attributed to both direct and indirect competitive interference.
Durante el estado larval ssil, los gusanos len (Diptera: Vermileonidae) proveen una visin nica de la importancia de la seleccin de hbitat debido a las condiciones locales que influyen grandemente en el crecimiento y la sobrevivencia. Al igual que sus parientes ecolgicos, las hormigas len (Neuroptera: Myrmeleontidae), las larvas de gusano len son depredadores sentarse-a-esperar que construyen sus trampas cnicas para capturar a sus presas. En este estudio, demuestro el efecto de la humedad en la construccin de las trampas por las larvas para explicar la presencia y los posibles problemas con las agregaciones de los gusanos len. He encontrado que las larvas prefieren las condiciones secas del sustrato, la presencia y el tamao de las trampas se vio afectado negativamente por las grandes cantidades de agua. Los resultados sugieren que las larvas carecen de la fuerza para manipular el suelo compactado, ms el sustrato saturado de agua, tienen que esperar hasta que las condiciones mejoren para poder construir las trampas. Debido a que los espacios secos estn limitados en el bosque lluvioso, las larvas del gusano len estn propensos a vivir en agregaciones densas. Para determinar el efecto de la variacin en la densidad de las larvas, compar el nmero de larvas construidas a concentraciones mayores con las observadas en el campo. Al aumentar la densidad de las larvas, el establecimiento de las trampas decrece. Los menores ndices de acierto en las trampas de las densidades altas pueden ser atribuidos tanto a la competencia en la interferencia directa e indirecta. Las preferencias por los sustratos secos en las larvas de gusano len presenta un aspecto de la historia natural del organismo que no ha sido investigado anteriormente y provee una posible explicacin por la presencia de las agregaciones.
Text in English.
Monteverde Biological Station (Costa Rica)
Seleccin de hbitat
Estacin Biolgica de Monteverde (Costa Rica)
Tropical Ecology Fall 2010
Wormlion (Diptera vermileonidae)
Ecologa Tropical Otoo 2010
Gusanos len (Diptera vermileonidae)
t Monteverde Institute : Tropical Ecology
1 Pit trap construction decisions by W ormlions (Diptera: Vermileonidae): effects of substrate moisture and larval density James H. Muldoon Department of Biologi cal Sciences, Whitman College. ABSTRACT During their sessile larval stage, wormlions (Diptera: Vermileonidae) provide a unique insight into the importance of habitat selection because their local conditions greatly influence their growth and survivorship. Like their ecological relativ es, antlions ( Neuroptera: Myrmeleontidae), wormlion larvae are a sit and wait predators that construct conical pits to capture their prey . In this study, I demonstrate effects of moisture on larval pit trap construction to explain the presence of and potential problems with wormlion aggregations . I found that larvae prefer drier soil conditions, for pit pr esence and size was negatively a ffected by large amounts of water . Results suggest larvae lack the strength to manipulate the compact, more w ater saturated soil and must wait until soil conditions improved to construct their pits. Because dry spaces are limited in rainforests, wormlion larvae are likely to live in dense aggregations . To determine the effects of varying densities on larval aggregations, I compared the number of pits constructed within high concentrations to that observed in the fiel d. As wormlion density increased , pit establishment decreased. Lower pit success in higher densities may be attributed to both direct and indirec t competitive interference. RESUMEN Durante el estado larval sÃ©sil, los gusano leÃ³n (Diptera: Vermileonidae) provee una Ãºnica visiÃ³n en la importancia de la selecciÃ³n de hÃ¡bitat debido las condiciones locales influencian grandemente el crecimiento y sobre vivencia. Como sus relativos ecolÃ³gicos, las hormigas leÃ³n (Neuroptera: and cÃ³nicas para capturar las presas. En este estudio, demuestro el efecto de la humedad en la construcciÃ³n de las trampas por las larvas para explicar la presencia de un problema potencial con la agregaciÃ³n de gusanos leÃ³n. EncontrÃ© que las larvas prefieren condiciones secas del sustrato, la presencia y tamaÃ±o de las trampas se vio ampliamen te afectado con grandes cantidades de agua. Esto sugiere que las larvas carecen de la fuerza para manipular el compacto y mÃ¡s saturado de agua sustrato hasta que las condiciones mejoran para construir las trampas. Debido a que Ã¡reas secas estÃ¡n limitados en el bosque lluvioso, las larvas del gusano leÃ³n son propensos a vivir en agregaciones densas. Para determinar el efecto de la variaciÃ³n en la densidad de las larvas, comparÃ© el nÃºmero de larvas construidas a concentraciones mayores de las observadas en el campo. Al aumentar la densidad de larvas, el establecimiento de trampas decrece. Menor Ã©xito en altas densidades puede ser atribuido tanto a la competencia en interferencia directa e indirecta. Preferencias por sustratos secos en las larvas de gusano leÃ³n presenta un aspecto de la historia natural del organismo no investigado anteriormente y provee una posible explicaciÃ³n por la presencia de las agregaciones.
2 INTRODUCTION significantly influences its physiological capacities and ultimately its demographic and ecological performance (Huey 1991). Consequently, an organism will choose to live in the habitat that best suits its needs . While habitat selection of mobile animals, specifically birds, has been thoroughly in vestigated, that of sessile species remains largely unstudied (Farji Brener 2003). Sessile species, however, provide an especially interesting opportunity to study habitat selection because their local conditions completely determine their growth, survivor ship, and reproduction ( Gotelli 1997) . As a result, sessile species are exposed to a higher degree of selective pressure to occupy an adequate habitat ( Farji Brener 2003; McCarthy 2007 ) . Several species of insects including antlions , Myrmelion s p. ( Neuroptera: Myrmeleontidae) , and wormlions, Vermilio s p. (Diptera: Vermileonidae), exhibit a sessile predatory larval stage , imp lementing the sit and wait strategy to capture their prey (McClure 1983 ; Zumbado 2006) . The duration of this larval stage and th e size of the ability to maximize food intake, thus emphasizing the importance of this larval stage on the success of the organism (Got elli 1997). While the ha bitat selection of antlions ( Farji Brener 2003; Hauber 1999 ; McCarthy 2007 ) has been thoroughly investigated, that of wormlions remains largely un studied (Devetak 2008). Although antlion and wormlion larvae differ signifi cantly in morphology, their pit trap s are strikingly similar (CalderÃ³n pers. comm.) . Recent studies reveal that microhabitat preference , including both biotic and abiotic factors, can severely influence pit construction and thus prey capture success (Gotelli 1997; Farji Brener 2003; Zumbado 2006). Although antlion response to a numb er of variables, including temperature, soil grain size, litter abundance, and density has been comprehensively explored (Day & Zalucki 2000, Farji Brener 2003 ; Gotelli 1997; McClure 1976) , w ormlions have only been shown to exhibit a strong preference towa rds finer grain soils (< 2 mm) ( Devetak 2008; McCarthy 2007) . Field observations support the prediction that wormlions prefer to build their traps in dry dusty soil, but the effects of moisture on trap efficiency have yet to be investigated. In this stud y , I simulate varying amounts of rainfall to demonstrate the effects of moisture content on trap efficiency to explain the presence of wormlion aggregations. Secondly, I examine the effects of wormlion density in the aggregations to explore the issue of habitat availability. If wormlions demonstrate the same reaction to an increased density in their aggregations as antlions, the number of pits constructed should be reduced due to increased competitive interference (Day & Zalucki 2000; McClure 197 6 ). METHODS Vermilio sp. larvae and soil were collected at the EstaciÃ³n BiolÃ³gica in Monteverde, Costa Rica (1500 m) . Sandy soil was taken directly from the natural wormlion habitat and sifted to create a uniform grain size substrate of < 2 mm, which larvae have been shown to prefer (Devetak 2008; McCarthy 2007 ). The sampled aggregation was located beneath a patio near the EstaciÃ³n. Fifty nine round plastic containers, 1 3 cm in diameter , were filled
3 to a dep th of 4 cm with the sifted soil (McCarthy 200 7). Containers were stored indoors at room temperature and exposed to even light conditions throughout the study . Study species Vermilio sp . larval stage l asts for approximately one year until the fly pupates and emerges a month later ( Petersen & Baker 2006). Larvae characteristically range from 2 to 20 mm in length and 0.2 to 2.5 mm in width. As their morphology resembles that of a typical worm, parts of their bodies are specialized for prey capture and pit building , including mouthparts and a ps eudopodium (Devetak 2008). To build their pit traps, larvae use their heads and mandibles to scoop up sand and throw it out of the pit (Petersen & Baker 2006). Although the preferred pit angle of wormlions has yet to be investigated, that of antlions is 10 0 degrees an angle shown to increase prey capture efficiency (Griffiths 1986 ) . When prey fall into the pit, their escape is hindered not only by the loose substrate of the pit slope, but also the miniature landslides caused by larvae hurling sand (Devetak 2008, Petersen & Baker 2006 , Pierce 1985 ) . Moisture preference Thirty six wormlion larvae were added to the separate soil contain ers and divided into three groups of twelve two treatments, light rain and heavy rain , and o ne dry control . They were allowed to construct their pit traps over night, and traps were subsequently measured the following morning to ensure the viability of all larvae. Traps were destroyed after measuring. With a water sprayer, each co ntainer of the light rain treatm ent was sprayed with 13 ml of wat er and each container of the heavy rain treatment was sprayed with 30 ml of water to test the impact of varying amounts of water . The soil was sprayed with a gentle misting pressure at in the afternoon before the larvae wer e able to construct their pit traps . Soil of the control group were left dry to provide a controlled comparison. All containers were left undisturbed for three days to simulate the effects of an isolated precipitati on event on initial construction of the p it trap . Pit trap depth and diameter were measured each day with a caliper to determine the change in pit trap size over time . Assuming a conical sha pe, pit trap angle was calculated a s = tan 1 (diameter/ (2* depth)) ( McCarthy 2007 ) ( Figure 1). W hen the soil had completely dried after three days of no additional water exposure , all traps were destroyed. Thirteen ml and 30 ml of water were reapplied to the soil of their respective groups, light rain and heavy rain . For the next four consecutive days, the same amounts of water (13 and 30 ml) were rea pplied to the same containers to simulate the effects of persistent daily precipitation on trap construction. Pit trap depth and diameter were measured each day to observe changes in trap size over time . T rap an gle was subsequently calculated . Wormlions were released following the experiment. Density effects Average p it trap density of the sampled larvae aggregation was measured in the field and translated to the expected average density of the circular conta iners. Five quadrats (466 cm 2 ) were placed in flat test areas of the aggregation. The number of pit traps within the quadrat and the distance s of each trap to the clos est other trap were measured . For the area of the container, it was concluded that three l arvae was the natural density observed
4 in the aggregation. One hundred and forty additional wormlion larvae were collected from the aggregation and divided into five sets, each including four containers of different degrees of larvae density one, three (control), five, and ten larvae per container . Three containers , each containing fifteen larva e , were later added to increase the maximum density variable . N umber of traps constructed within each container and distance s of each trap to the closest other trap were measured the following day. After measuring, traps were destroyed, allowed to reform at night, and measured the next aftern oon. This process was repeated for a period of four days . Wormlions were released following the experiment. Results Moist ure preference The number of pits constructed per day diff ered amongst treatments for the single exposure experiment (Figure 2). All 36 larvae constructed pits on the first day of observation , that is, the day before the water was implemented . On day two, however, after the containers were exposed to water, 83% of larvae of the control, 92% of larvae of the light rain treatment, and only 33% of larvae of the heavy rain treatment constructed pit traps. On day three, the seco nd day after water e xposure, 100% of larvae of the control, 92% of larvae of the light rain treatment, and 67 % of larvae of the heavy rain treatment established pits. Trap count of the forth day revealed more uniform numbers 100% larvae of the control and 92% of larvae of bo th the light rain and heavy rain treatment s built their traps. Because container soil typically dried unevenly, larvae tracks on the surface of the sand were commonly observed leading to the most arid part of the container where a pit was present. 0 2 4 6 8 10 12 14 1 2 3 4 Number of pit traps constructed Day None 13 ml 30 ml Figure 2. Number of wormlion ( Vermilio sp.) pit traps constructed over time single rain treatment. Pit number greatly decreased following the heavy rain (30 ml) treatment and steadily rose over the following days. Pit number remained relatively static foll owing the light rain treatment and the control.
5 One way ANOVA analyses of s ingle rain results demonstrate varying degrees of differ ence between water treatm ents over time . Mean measurements ranged from 8.5 to 28.6 mm in pit diameter, 1.4 to 13.5 mm in pit depth, and 73.6 to 97.2 degrees in angle . On day one before the containers had undergone water treatment, average pit trap sizes in diameter, depth, and angle were not sig nificantly different acr oss the control and both degrees of water treatments ( F 2,33 = 2.1, p=0.14 ; F 2,33 = 1.4, p=0.27 ; F 2,33 = 0.12, p=0.92 ) (Figures 1 & 3; Table 1 ). Pit trap diameter from days two through four did include significant differences ( F 2,33 = 8.9 , p=0.0008 ; F 2,33 =13 .6 , p<.0001 ; F 2,32 =12.2 , p=0.0001 ) (Figure 3) . Differences were present between the heavy rain treatment and the others . In heavy rain treated soil, larvae constructed pits with much smaller diameters. Analyses also indicated differences in pit trap depth from days two through four ( F 2,33 =14.7 , p<0.0001 ; F 2,33 =9. 5 , p=0.0006 ; F 2,32 =13.1 , p<0.0001 ) (Figure 3) . Similarly, differences sourced from the discrepency between the heavy rain treatment and the others. In the heavy rain treated soil, larvae built much shallower pits. Analyses of pit trap angle never reveale d significant differences (Table 1) . Pit trap angle remained relatively static (about 87 degrees) throughout the four days of observation regardless of treatment (Figure 1). Figure 1. Measurements and comparisons of wormlion ( Vermilio sp) pit traps across single and persistent water treatments. Figures are drawn to scale. Each triangle represents relative mean size and shape of the pit trap for the specific treatment. Angle stays relatively constant (Table 1) in single treatment. Angle bec omes steeper in the persistent exposures.
6 Treatment Day 1 Day 2 Day 3 Day 4 None 8 6 .6 Â±11.7 85.9 Â±10.0 90.7 Â±24.5 95.9 Â±16.5 Light rain (13 ml) 84.4 Â±15.3 73.6 Â±16.8 80.0 Â±16.0 83.4 Â±14.4 Heavy rain (30 ml) 85.0 Â±8.2 90.7 Â±44.4 90.1 Â±17.0 97.3 Â±16.6 ANOVA F 2,33 =2.1, p=0.14 F 2,22 =1.4, p=0.27 F 2,28 =0.97, p=0.39 F 2,32 =2.7, p=0.08 P it traps constructed varied even more under the persistent water treatments (Figure 4). Throughout the five days of observation , 100% of wormlion larvae in the control group constructed pit traps. Larvae exposed to the light rain treatment ( 13 ml water per day ) typically built their traps. On days one through five, 83%, 92%, 83%, 92%, and 92% of larvae established traps , respectively . Those exposed to the persistent heavy rain treatment co nstructed far less traps. Only 17% of larvae built traps on day one, and just one larva built a trap on day two. No traps were created in the f inal three days of observation. Table 1. Mean ( Â± SD) wormlion ( Vermilio sp.) angle measurements (degrees) per day across water treatments single rain treatment. Amount of water in both treatments did not affect pit angle, as angle remained relatively static regardless of treatment. Significant differences in angle measurement s between treatments are absent throughout the observation period (ANOVA, p>0.05). Figure 3. Mean ( + SD) wormlion ( Vermilio sp.) pit trap dimensions over a four day period (mm) single rain treatment. On day 1 the water had not been applied, so there are no significant differences between treatments (p>0.05 ANOVA) for neither variable, diameter (A.) or depth (B.). Different let ters indicate significant differences between means each day (p<0.05, post hoc Tukey test).
7 S ignificant differences (t tests) within mean pit diameter, d epth, and angle across the five day observation period between the control and t he light rain treatment were found . D ue to the absence of pits , the heavy rain treatment was not included in these analyses . Mean measu rements ranged from 10.0 to 28.8 mm in pit diameter , 5.5 to 11.6 mm in pit depth, and 61.5 to 95.9 degrees in angle. Analyses reveal ed no significant differences between the light rain treatment and the control in pit trap diameter over the five days of observation ( t= 0.70, df=20, p=0.49 ; t= 1.090 , df=20, p=0.29 ; t= 1.7 , df=20, p=0.11 ; t= 1.8 , df =21, p=0.08 ; t= 1.8, df= 21, p=0.08 ) (Figure 5). Pit trap diameters of the two treatments were relatively the same size. Also, analyses illustrate d no differences in pit trap depth over the observation period ( t=0.28, df=19.9, p=0.78 ; t=1.5 , df= 20, p=0.14 ; t=1.4, df=20, p=0.18 ; t=1.4, df=21, p=0.18 ; t=1.4, df=21, p=0.17 ) (Figure 5) . Regardless of the treatment, larvae constructed pit traps of similar depths. Analyses of angle, however, did reveal some significant differences (Figure 5). On day one, no significant differences bet ween the light rain treatment and the control were present pits exhibited similar angles ( t= 1.7, df=20, p=0.10 ). Pit trap angle was different in days two through five ( t= 4.2 , df=20, p=0.0004 ; t= 4 .9 , df=20, p<0.0001 ; t= 5.9 , df=21, p<0.0001 ; t= 5.6 , df =21, p<0.0001 ). The angles of the light rain treated pit traps much steeper than those of the control pits (Figure 1) . 0 2 4 6 8 10 12 14 1 2 3 4 5 Number pf pit traps consrtucted Day None 13 ml 30 ml Figure 4. Number of wormlion ( Vermilio sp.) pit traps constructed over time per water treatment (none, light rain, heavy rain) persistent rain treatment. Larvae subject to the heavy rain treatment did not establish pits after the second day of observation. Larvae subject to light rain treatment always constructed ten or more of their pits. Larvae in the control group always created their pits.
8 Density effects Wormlion density influenced the percentage of larvae per container that constructed pit traps ( Figure 6). Th r oughout all five days of observation , nearly 100% of larvae in densities 1, 3, and 5 built their traps only once did a larva of a density 3 group fail to p roduce a trap. On average, between 80 to 88% of larvae in density 10 constructed their pit traps over the five day period. Density 15 experienced an even lower mean pit establishment only 62 to 75% of larvae typically built pits. Pit trap counts remained r elatively uniform th r oughout the observation period . However, the low pit trap count of density 15 showed a steady increase as time progressed , rising from 62 to 75% . Pits of density 10 and 15 changed location after being destroyed on a daily basis. Kruskal Wallis analyses highlight significant differences in mean trap distances across density treatments ( Figure 7). Dens ity 1 was not included in the DNN comparisons due to the impossibility of distance comparison. Mean pit trap distances ranged from 28 to 58 mm. Mean trap distance did not differ significantly among density treatments on Figure 5. Mean ( + SD) wormlion ( Vermilio sp.) pit trap dimensions over a five day period (mm) persistent rain treatment. On day 1 the water had not been applied, so there are no significant differences between treatments (p>0.05 ANOVA) for any of the variables, diameter (A.) or depth (B.), or an gle (C.). Different letters indicate significant differences between means each day (p<0.05, post hoc Tukey test). There are no significant differences in either diameter or depth.
9 day one ( KW: 2= 4.7 , d f =2, p= 0.093 . On day two t he mean distance between traps of the control density was significantly greater than that of all other density treatments ( KW: 2= 11.3, df=3, p=0.01 ) . Li kewise, mean distance of density 15 was significantly smaller than that of the others. On day three m ean trap distance of the control density was significantly larger than that of all other density variables ( KW: 2= 14.5, df=3, p=0.002 ). Although smaller in that of the control density, mean trap distance of density 5 was significantly greater than that of densities 10 and 15. This trend continues on days 4 ( KW: 2 =14.9, df=3, p=0.002 ) and 5 ( KW: 2= 14.9, df=3, p=0.0 02 ) (Fig. 7). 50 60 70 80 90 100 1 2 3 4 5 Percent (%) Day 1WL 3WLs 5WLs 10WLs (9) 15WLs (11) Figure 6. Percentages of wormlion ( Vermilio sp.) pit trap success across density treatments (1,3,5,10, and 15 larvae per container) per day. Density 15 variable is absent from day 1 observations as it was not yet added to the study. Number s in parentheses represent mean numbers of wormlions that constructed pit traps for the corresponding density treatments.
10 DISCUSSION Pit trap count s and sizes following the single water treatment support t he field observation that wormlion larvae prefer dry dusty soil to construct their pits ( Devetak 2008). In fact, results indicate that exposure to water even hinders trap con s truction and efficiency. Not only did fewer larvae construct traps when exposed to the most water , but also when they constructed their pits they were significantly sm aller in depth and diameter. Due to pit size reduction, potential prey capture is lessened and net energy return is lost (Griffiths 1986). Size reductions may act as a behavioral adaptation to avoid the impa cts of additional rain fall less rain is likely to fall in to a smaller trap , and thus less energy is required to fix the trap after it becomes inundated. Reduced number and size of water exposed pi t traps also suggest a higher degree of difficulty in constr ucting traps in saturated soil. Because larvae must throw and manipulate surrounding soil to build pits, they are likely too weak to operate in the heavier, more compact water saturated soil (Petersen & Baker 2006 ; Pierce 1985) . As the soil dried, larvae were more likely able to move it. P it angle , however, remained relatively static throughout the observation period, even though pit diameter and depth decreas ed. This suggests that Figure 7. Mean distance of wormlion ( Vermilio sp.) pit traps across density treatments (3, 5, 10, and 15) per day. Different letters indicate significant differences between means each day (p<0.05, post hoc Tukey test) Treatments connected by the same letter are not significantly different (ANOVA p>0 .05). Density 15 variable is absent from day 1 observations as it was not yet added to the study.
11 wormlion larvae prefer a pit angle of approxima tely 87 degrees, and that this specific pit shape is more important in prey capture and efficiency than depth and diameter (Griffiths 1986) . Interestingly, mean depth was grea ter in the light rain treated pits than in the control pits. This could be explai ned by the malleability of the semi damp soil that allowed the larvae to dig a deeper pit . However, success of traps constructed in damp soil is questionable. Because larvae rely on loose soil substrate to prevent prey from escaping the pit, they may exper ience a po or prey retention rate (Pierce 1985) . As damp soil is more compact than dry soil, prey could more easily escape due to increased surface traction of the substrate, and larvae could not as effectively induce miniature landslides to hinder evasion. Furthermore, the surface tracks atop the water treated containers suggest a preference for dr ier soils. Larvae sought out the drier regions of soil , indicating they are able to differentiate among soil moisture content to select a suit able habitat . Although the 13 ml treatment induced a slightly lower mean diameter, most other significant differences resulted from the 30 ml treatments, suggesting that a significant isolated precip itation event is required to a ffect pit trap construction . Effects of heavy moisture on pit trap construction are further supported by persiste nt water treatment results. Because traps were exposed to water each day, soil moisture content was consistent as it was unable to dry out. The absence of pits in the 30 ml treated containers suggests that the larvae was insufficient to move the heavily saturated soil. As the soil became increasingly wet, they were likely trapped beneath the compact soil, unable to reach the surface to seek out drier conditions. Curiously, mean angle of the 13 ml treatment, approximately 70 degrees, was significantly steeper than that of the control, 87 degrees, originating from a greater depth and smaller diameter. Essentially, pit Larvae in the persistently wet soil strayed away from their preferre d angle of 87 degrees, which they exhibit in dry soil. This likely reflects a behavioral adaptation designed to thwart negative effects of damp soil on prey capture. Although damp compact soil may promote prey evasion, a steeper pit may compensate for unde sirable moisture conditions, in that a steeper pit face would be harder to climb. Absence of this angle adaptation in the single rain treatment results further highlights the importance of an appropriate dry habitat larvae need not change their foraging st rategies in dry conditions. Low numbers of established pits and likely behavioral changes further suggest that wormlion larvae prefer dry soil conditions for pit trap construction, and that persiste nt precip i tation events are detrimental to larvae feeding strategies. In tropical areas like Monteverde, Costa Rica that receive 3.5 m of rainfall per year , dry space is very limited, especially during the rainy season (Clark et al. 2000). Because wormlion larvae require dry conditions to facilitate their foraging strategies, they must seek out these cryptic dry spots in any place possible, like areas adjacent to buildings or under overhanging logs and rocks ( Day & Zalucki 200 0; Gotelli 1997; Zumbado 2006) . Due to limited habitat availa bility and their restricted dispersal ability, worml ion larvae tend to live together in aggregations . Results illustrate that larvae density within an aggregation can effect pit trap construction, as seen in studies of antlions (Day & Zalucki 2000; McClu re 1976). The lower numbers of pits formed at higher densities can be explained by several scenarios of competitive interaction . Some larvae may have postponed pit construction in an effort to achieve an optimal spatial arrang e ment within their aggregation. Antlion studies demonstrate that larvae construct pits as far apart from one another as possible to optimize chances of prey ca pture, suggesting a preference for uniform distribution
12 (McClure 1976) . Because wormlion larvae wo uld not construct their traps within 28 mm of one another , it is possible that some wer e unable to locate a spot sufficiently removed from the others . Higher densities may also encourage direct construction interference between individuals. Because larvae must throw sand to construct their pit traps, it is possible that nearby traps would be subject to incoming substrate a process resulting in the destruction of proxi mal pits (Petersen & Baker 2006; Pierce 1985). Low pit establishment may also result from c annibalistic mortality. Although larvae survivorship was not documented at the end of the observation period , nearby larvae in high densities may have encroached too close upon one another and instigated a fatal confrontation. This beha vior has been observed in antlions but has yet to be revealed in wormlions (Day & Zalucki 2000). Moreover, it is interesting to note that in both densities 3 and 5 all larvae constructed their pits nearly 100% of the time. This suggests that larvae could n aturally live in a hig her mean density than they do in the sampled aggregation an ability that could potentially increase available habitat for colonization, expanding the carrying capac ity for a given dry location. Aggregation age suggests a possible expl anation the density of the sampled area may be a product of how long wormlions have colonized that location. Calculating mean de nsity in the same aggregation several years from now may reveal higher results. This study highlights the effects of moisture on wormlion pit trap size and construction to explain the presence of larvae aggregation s . It demonstrates that soil moisture content plays a major role in larval habitat selection, and highlights potential consequences of living in high densities. Like their ecological equivalents , antlions, wormlion larvae suffer from increased competitive interference in higher densities . To expand upon this study, documenting survivorship of larvae in high densities would help explain the absence of pit traps. Al so, increasing the maximum larvae density would provide a stronger backing to density limitation conclusions. While this study reveals novel information about the understudied Vermilio spp., future studies will uncover additional knowledge about this speci ACKNOWLEDGEMENTS I cannot express enough thanks to the TEC staff for providing me with a comprehensive education of tropical systems and the experience of a lifetime. I would like to extend special thanks to my project advisor, Pablo Allen, for his extensive support and enthusiasm during my study witho ut Pablo, this study would have never been possible , let alone conceived . Thank you, Isabel Brenes Salzaar, for letting me into your wonderful family and supplying me with fried nourishment throughout the course of my homestay. Thank you, EstaciÃ³n BiolÃ³gic a , for the use of your beautiful station. And lastly, thank you , wormlions, for being easier to catch than dragonflies. LITERATURE CITED Clark, K. L, R. O. Lawton, P. R. Butler. 2000. Monteverde ecology and conservation of a tropical cloud forest, pp. 15 17. Oxford University Press, New York, New York. Day, M. D. and M. P. Zalucki. 2000. Effect of density on spatial distribution, pit formation and pit diameter of Myrmeleon acer Walker, (Neuroptera: Myrmeleontidae): patterns and processes. Austral Ecolo gy. 25: 58 64.
13 Devetak, D. 2008. Substrate particle size preference of wormlion Vermileo vermileo (Diptera: Vermileonidae) larvae and their interaction with antlions. Eur. J. Entomol. 105: 631 635. Fafji Brener, A. G. 2003. Microhabitat selection by antli on larvae, Myrmeleon crudelis : effect of soil particle size on pit trap design and prey capture. Journal of Insect Behavior . 16: 783 796. Gotelli, N. J. 1997. Competition and coexistence of larval ant lions. Ecology . 78(6): 1761 17 73. Griffiths, D. 1986. Pit construction by ant lion larvae: a cost benefit analysis. Journal of Animal Ecology 55: 37 59. Hauber, M. E. 1999. Variation in pit size of antlion ( Myrmeleon carolinus ) larvae: the importance of pit construction. Physiological Entomology. 24: 37 40. Huey, R. B. 1991. Physiological consequences of habitat selection. The American naturalist. 137 : 91 115 McCarthy, A. 2007. Selection for survival: soil grain size preference and body size in wormlion (Vermilio spp.) lar vae affect pit trap construction. In CIEE; Fall, Monteverde, pp . 61 69 McClure, M.S. 1983. Myrmeleon (Hormiga Leon, Ant Lions). In: Costa Rican Natural History. D. H. Janzen (ed). The University of Chicago Press, Chicago, IL, pp. 742. Petersen, W. H. and C. W. Baker. 2006. First record of wormlions (Diptera: Vermileoidae) in Idaho. Journal of the Idaho Academy of Science. 42 (1). 1 5. Pierce, J.D. 1985. A review of tool use in insects. The Florida Entomologist . 69(1): 95 104 .