|UFDC Home||myUFDC Home | Help|
This item has the following downloads:
1 FORAGING AND FECUNDITY OF Larra bicolor (HYMENOPTERA: SPHECIDAE) A PARASITOID OF Scapteriscus MOLE CRICKETS. By SCOTT L. PORTMAN A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLOR IDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 2007
2 2007 Scott Linus Portman
3 To my loving parents and family who have always accepted and encouraged my fascination with insects
4 ACKNOWLEDGMENTS First I sincerely thank my dearest wife Y ukiko for giving me the courage to pursue my passion for insects and her willingness to provid e all the support that I needed along the way. Secondly, I especially thank Dr. J. Howard Fran k for his guidance, insight and patience during my stay in his lab. His knowledge and experience were invaluable to the success of this project. Next, I thank the other members of my graduate committee (Dr. Norman Leppla and Dr. Robert McSorley) for their support and leadership. Without their direc tion and assistance this thesis would have been impossible. Lastly I thank Dr. Matthew Aubuchon (USDA CMAVE, Gainesville, FL) for helping me to learn how to use SAS.
5 TABLE OF CONTENTS page ACKNOWLEDGMENTS...............................................................................................................4 LIST OF TABLES................................................................................................................. ..........7 LIST OF FIGURES................................................................................................................ .........8 ABSTRACT....................................................................................................................... ..............9 CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW..............................................................11 Introduction to the Mole Cricket Problem in Florida.............................................................11 History of the Mole Cricket Invasion.....................................................................................12 Mole Cricket Basic Biology and Behavior.............................................................................12 Benefits of Mole Cric ket Biological Control..........................................................................13 History of Larra bicolor s Introduction into Florida..............................................................13 Basic Biology and Behavior of Larra bicolor ........................................................................15 Validation of Larra bicolor as an Effective Classical Biological Control Agent..................16 Project Objectives............................................................................................................. ......17 2 EVALUATION OF NATIVE PLANT SPECI ES FOR VIABILITY AS NECTAR SOURCES AND ATTRACTANTS.......................................................................................18 Introduction................................................................................................................... ..........18 Materials and Methods.......................................................................................................... .20 Results........................................................................................................................ .............22 Discussion..................................................................................................................... ..........24 3 EFFECT OF ADULT HOST PLANTS ON FORAGING PATTERNS AND LOCAL DISTRIBUTION................................................................................................................... .29 Introduction................................................................................................................... ..........29 Materials and Methods.......................................................................................................... .32 Results........................................................................................................................ .............33 Discussion..................................................................................................................... ..........34 4 POTENTIAL FECUNDITY AND OVIPOSITION RATE...................................................38 Introduction................................................................................................................... ..........38 Material and Methods........................................................................................................... ..40 Results........................................................................................................................ .............42 Discussion..................................................................................................................... ..........43
6 5 CHARACTERIZATION OF OVARI AL ULTRA-STRUCTURE AND QUANTIFICATION OF INTRASPECIFIC VARIATION...................................................46 Introduction................................................................................................................... ..........46 Materials and Methods.......................................................................................................... .48 Results........................................................................................................................ .............49 Discussion..................................................................................................................... ..........50 6 SUPERPARSITISM AND DEVELOPMENTA L SUCCESS OF SUPERNUMERARY LARVAE......................................................................................................................... .......53 Introduction................................................................................................................... ..........53 Materials and Methods.......................................................................................................... .54 Results and Discussion......................................................................................................... ..54 LIST OF REFERENCES............................................................................................................. ..57 BIOGRAPHICAL SKETCH.........................................................................................................63
7 LIST OF TABLES Table page 2-1 Number (x SEM) of adult males observed on eac h plant species. Values represent counts from all plots (2 locations 3 repli cations). Means with same letters are not significantly (P< 0.05) different according to Duncans MRT on log-transformed data).......................................................................................................................... ..........28 2-2 Number (x SEM) of adult females observed on each plant species. Values represent counts from all plots (2 locations 3 replications) Means with same letters are not significantly (P< 0.05) different according to Duncans MRT on logtransformed data............................................................................................................... ..28 2-3 Number (x SEM) of total wasps (males + females) observed on each plant species.Values represent counts from all plot s (2 locations 3 replications). Means with same letters are not significantly (P< 0.05) different according to Duncans MRT on log-transformed data...........................................................................................28 3-1 Number (x SEM) of mole crickets caught in pitfall traps and pe rcent parasitized by Larra bicolor Values represent monthly sampling means (2005: Oct N=2, Nov N=5, Dec N=2; 2006: Aug N=1, Sept N= 4, Oct N=4, Nov N=6, Dec N=1)..............................36 4-1 Summary statistics for e xperimental variables. Values represent mean SEM, minimum and maximum for wasp weight (N=15), lifespan (N=20), lifetime fecundity (N=20), and oviposition rate (N=20).................................................................44 5-1 Summary statistics for ovarial traits. Values represent x SEM, minimum and maximum for tibia length (N=10), Ovary length (N=10), ovariole length (N=60), mature egg number (N=10), egg length (N=76), and number of oocytes (N=10).............52
8 LIST OF FIGURES Figure page 3-1 Average number of parasites present as a function of the numbe r of mole crickets captured in 10 pitfall traps placed at 20 m intervals extending out from a patch of nectar source plants. (A) Numbers reco rded for 2005. (B) Numbers recorded for 2006........................................................................................................................... .........37 3-2 Average number of parasites present as a function of distance from the adult wasps preferred food source. (A) Numbers of para sitoids recorded in 2005. (B) Numbers of parasitoids recorded in 2006..............................................................................................37 4-1 Average number of eggs oviposited per da y as a function of female weight (N=15).......45 4-2 Total lifetime egg production as a function of female lifespan (N=20)............................45 5-1 L. bicolor ovary stained with neutral red and photographed at 10 original size using the Auto-montage imaging system attached to a stereomicroscope. Ovary was slide-mounted in 30% glycerol..........................................................................................52 6-1 Photographs showing the diffe rent numbers and locations of L. bicolor eggs. A) two eggs adjacent B) three eggs................................................................................................55 6-2 Photographs showing the development of supernumerary L. bicolor larvae. A) larvae opposite sides B) la rvae same side.....................................................................................55
9 Abstract of Thesis Presen ted to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science FORAGING AND FECUNDITY OF Larra bicolor (HYMENOPTERA: SPHECIDAE) A PARASITOID OF Scapteriscus MOLE CRICKETS By Scott L. Portman December 2007 Chair: J. Howard Frank Major: Entomology Studies on important aspects of the biology and behavior of Larra bicolor F. (Hymenoptera: Sphecidae) such as nectar source preferences, host foraging patterns, fecundity and ovary morphology are presented. In July, Au gust and September as many adult wasps were observed on the native plant Chamaecrista fasciculata (Michx.) Greene as were observed on the non-native control plant Spermacoce verticillata L. In the following months most wasps were observed on the control plant and very few if a ny on the native species which had deteriorated later in the year. Possibilities for using these two plants in habitat management are discussed. Studies on the foraging patterns of Larra bicolor revealed that the number of parasitized mole crickets was found to be larger in high density host patches. Additionally, greater numbers of parasitized mole crickets were found in pitf all traps that were closer to a plot of Spermacoce verticillata. This indicated that Larra bicolor females are more abundant near nectar sources. Implications for biological control and populat ion dynamics are discussed. Fecundity studies revealed that female wasps lived an average of 23.5 days and produced an average of 56.05 eggs during their lifetime. Eggs were produced at an average rate of 2.44 eggs per day. The number of eggs that females produce was positively correlate d to their size (weight). The number of eggs that they produce is positively linearly rela ted to their lifespan. Characterization of Larra
10 bicolor s ovaries revealed that ovaries and ovari oles averaged 9.68 mm and 9.23 mm in length, respectively. They carried an average of 7.60 mature eggs and an average of 69.0 developing oocytes. Additionally, ovary length, ovariole le ngth and egg load all corresponded positively with female size and eggload correlated ne gatively with egg length. Observations of superparasitized mole crickets indicated that mole cricket hosts cannot support more than one larval parasite. Superparasitism by Larra bicolor yields two possible outcomes, either both larvae die or only one larva survives while the others perish. The effect of superparasitism on biological control and wasp populations is discussed.
11 CHAPTER 1 INTRODUCTION AND LITERATURE REVIEW Introduction to the Mole Cricket Problem in Florida Mole crickets belonging to the genus Scapteriscus (Orthoptera: Gryllotalpidae) have become the most destructive turf and pasture-grass pests in Flor ida and the southeastern United States (Hudson 1988, Frank 1990). Hosp itable environmental conditions in the region, such as abundant food resources, warm temperatures, sa ndy soil, and the lack of specialist natural enemies, have allowed Scapteriscus populations to grow and sp read rapidly (Castner 1988b, Frank 1994). Bahiagrass ( Paspalum notatum Flgge), commonly grown on beef cattle ranches, and Bermudagrass ( Cynodon spp.), used to sod golf course fairwa ys, are especially susceptible to mole cricket damage (Frank 1999). Mole crickets not only destroy turf-gra ss, they also injure vegetable seedlings and ornament al plants by feeding on roots and shearing off stems at ground level (Frank 1990, 1999). Lastly, areas heavily populated by mole crickets can attract insectivorous mammals, such as raccoons, mole s and armadillos whose foraging activity may cause additional damage to th e local vegetation (Frank1999, 2003). In Florida alone, mole crickets are respons ible for the destruction of several hundred thousand acres of turf each year In 1986, it was estimated that ranchers, golf courses and sod farmers lost a total of $45 milli on as a direct consequence of turf damage caused by mole crickets. Moreover, in 1996, turf-related indus tries reportedly spent approximately $18 million on insecticides or other contro l treatments (Frank 1999). Because turf-grass is crucial to many economically important industries in the southeastern United States, particularly Florida, there is a need for mole cricket control methods that ar e effective, economical an d environmentally safe.
12 History of the Mole Cricket Invasion Three species of Scapteriscus mole crickets arrived into the southern United States between 1899 and 1926 (Walker 1981) It is believed that the s hort-winged mole cricket, S. abbreviatus Scudder, the southern mole cricket, S. borellii Giglio-Tos, and the tawny mole cricket, S. vicinus Scudder, were inadverten tly transported from Argen tina and/or Uruguay in soil which was used as ballast for cargo sh ips (Walker 1981, Castner 1984, Frank 1994, Parkman 1996, Frank 1999). Current distributions of the southe rn mole cricket and the tawny mole cricket extend throughout the southern Coastal Plain fr om eastern Texas to North Carolina (Walker 1981, Frank 1990, Parkman 1996, Henne 2001). Isolated populations of S. borellii have also been reported in Arizona (Nickle 1988, Pa rkman 1996, Henne 2001) and California (Frank 1994, Henne 2001). The distribution of the short-winged mole cricket remains limited to coastal regions in southern Florida, northeast Fl orida and southern Ge orgia (Castner 1988a). Scapteriscus abbreviatus failure to expand its range further most likely stems from its inability to fly (Castner 1988a, Frank 1990,1999). Mole Cricket Basic Biology and Behavior Mole crickets are so named because they tunne l just below the surface of the soil using a pair of enlarged muscular forelegs. Their forelim bs are able to cut through soil rapidly because they bear heavy blade-like projections known as dactyls (Smith 1893, Frank 1994, 1999). The number and arrangement of dactyls are useful indicators for distinguishing species (Frank 2003) Mole crickets are omnivorous; therefore, they readily feed on both plan t and animal material (Matheny 1981). They generally injure grasse s and seedling vegetation by feeding aboveground on the foliage and stems, or belowground on the r oots and tubers. Their bu rrowing activities also harm plants by dislodging the roots causing the plants to desiccate (F rank 1998). Southern mole crickets are primarily carnivores; thus, they typi cally inflict less herbivory damage than either
13 tawny or short-winged mole crickets (M atheny 1981, Brandenburg 2002 ). Southern mole crickets, however, tend to be more active diggers Consequently, the dama ge that they inflict upon vegetation largely results from their t unneling behavior (Brandenburg 2002). Although they are major pests in southeastern Florida, there is generally less concern for the harm caused by short-winged mole crickets because of their limited distribution (Castner 1988a). Benefits of Mole Cricket Biological Control Recurrent application of toxic chemicals, such as chlordane or soil fumigants, has been the prevailing mole cricket control strategy in Florida since the 1 940s (Frank 1990, 1999, Lewis 1997). However, the treatment of cattle grazing la nd with pesticides presents a problem because the insecticidal toxins contained within such trea tments may also pose a risk to livestock (Adjei 2000). Furthermore, a 1994 survey of Florida pe st species exhibiting resistance to commonly used insecticides identified Scapteriscus mole crickets as displaying signs of resistance to chlordane (Leibee 1995). A ban on the agricultural use of chlordane in the United States and a general impetus toward reducing or eliminating th e use of chemical insecticides necessitates the implementation of self-sustaining area-wide bi ological control measures (Leibee 1995, Lewis 1997). Establishment of effective biological cont rol methods, which are capable of restricting pest mole cricket populations to acceptable levels, woul d ultimately benefit turf-grass growers by decreasing their losses due to damage by mole cr ickets as well as reducing their overall maintenance costs. In addition, biological control of these pests will help curtail the release of potentially harmful chemicals in to the environment (Frank 1999). History of Larra bicolor s Introduction into Florida Larra spp. (Hymenoptera: Sphecidae: Larrinae) are specialist parasitoid s of mole crickets (Williams 1928, Smith 1935, Bohart 1976, Pruett 1991, Menke 1992). In the early 20th century, entomologists in Hawaii first recognized Larra s potential as a classical biological control agent
14 for combating an invasive species of mole cricket, Gryllotalpa orientalis Burmeister. After unsuccessful attempts at relocating L. bicolor F. from Brazil and L. amplipennis F. Smith from the Philippines, a population of L. polita F. Smith, acquired from the Philippines, was finally established in Hawaii by 1925 (Frank 1995, 1999). Correspondingly, a population of L. bicolor had been established in Puerto Rico by 1941 in an attempt to control the Changa, Scapteriscus didactylus Latreille, which had become a serious agricultural pest on the island (Wolcott 1941, Frank 1995). In the 1940s, a single attempt to import L. bicolor into Florida from Puerto Rico failed (Frank 1994,1995). With the foundation of the Univer sity of Floridas mo le cricket research program in 1979, a renewed effort to introduce L. bicolor into Florida was organized (Frank 1994, 1995, 1999). In 1981, large numbers of female wasps were captured in Puerto Rico and released in Ft. Lauderdale, Gainesville, and Tampa. In 1982 and 1983, more wasps were released in Bradenton, Ft. Lauderdale and Lakeland. Despite multiple releases at various locations over three years, Ft. Lauderdale became the only release site where L. bicolor managed to survive (Sailer 1985, Frank 1990, 1995). F. D. Bennett hypothesized that establishment of L. bicolor was unsuccessful in northern Florida because the stock, which originated from Br azil, was unable to tolerate the colder winters of the higher latitudes (Castner 1988a). Therefore, researchers collected more living specimens of L. bicolor from Santa Cruz de la Sierra, Bolivia, and released them in Alachua County, Florida, from 1988-1989 (Frank 1990, 1995 Bennett 1991). By 1993, a population of L. bicolor had been established in the Gainesville metr opolitan area; by 2002, the wasps had dispersed at least 30 km to the west and nor theast and 280 km northwest and s outh from their initial release site (Frank 2003).
15 Basic Biology and Behavior of Larra bicolor Larra species are solitary parasi toid wasps which belong to the family Sphecidae (digger wasps) (Smith 1935, Castner 1988b, Menke 1992). Th e genus is distributed worldwide but the majority of the species inhabit tropi cal regions (Bohart 1976, Frank 1995). All Larra species exploit mole crickets (Orthoptera: Gryllotalpid ae) as hosts for their developing larvae (Williams 1928, Bohart 1976, Menke 1992, Frank 1994). Unlike most sphecid wasps, which provision underground nest cells with perm anently paralyzed arthropods, Larra s sting only paralyzes their victims for several minutes (Steiner 1984, Castner 1988b, Frank 1995). The larvae feed and develop as external parasitoids on active mole cricket hosts (Smith 1935, Bohart 1976, Castner 1988b, Frank 1995). Adults primarily feed on nectar. Nectar provi des the wasps with s ugars and other vital nutrients (Smith 1935, Castner 1988b, Pruett 1991). Although thes e wasps are known to feed from the flowers of various plants field tests and observations indica te that they are particularly responsive to Spermacoce verticillata L. (Rubiaceae) (Williams 1928, Castner 1988a, Bennett 1991, Pruett 1991). The wasps preference for S. verticillata most likely stems from the fact that the flowers possess shallow coro llas which are accessible to L. bicolors comparatively short tongue (Arvalo-Rodrguez 2005). It is believed that female wasps are observed on flowers less often than males because females spend most of their time hunting for mole crickets (Castner 1988b). Adult female wasps locate mole cricket tunne ls, known as galleries, by searching along the ground. Once a gallery is located, she digs down in to the soil and forces the inhabiting mole cricket to the surface (Smith 1935, Bohart 1976, Castner 1988b, Bennett 1991). Once the cricket surfaces, the wasp chases it down and pounces on it. The female positions herself above the crickets pronotum, perpendicular to the long itudinal axis of the host; she then bends her
16 abdomen underneath and administers repeated st ings to the ventral base of the crickets prothorax, mesothorax, and cer vix (Smith 1935, Steiner 1984, Castner 1988b). As soon as the paralyzing effect from her stings immobilizes the cricket, the wasp attaches a single egg to the crickets ventral thorax, between its first and second pair of legs (Bohart 1976, Castner 1988b). When paralysis subsides, the cricket immediat ely burrows into the ground and resumes its normal activity (Smith 1935, Boha rt 1976, Steiner 1984, Castner 1988b). Eggs begin to hatch in 6-8 days. First instar s remain attached to the external surface of their hosts. The larvae feed on their hosts he molymph by puncturing the cuticle and inserting their head and mouthparts into the mole cr ickets body cavity (Smith 1923, Castner 1988b). Full larval development consists of five inst ars (Cushman 1935, Smith 1935, Castner 1988b). On completion of the fourth instar, the larvae kill th eir hosts. Fifth-instars feed on the muscular and soft internal tissues of the crickets carcass, leaving behind an indicativ e pile of sclerotized remains (Smith 1935, Caster 1988b). When the grub reaches maturity, it rests for a short time before constructing an ovate cocoon from s ilk and soil particles (Smith 1923, Castner 1988b). Adult wasps begin to eclose 6-8 weeks later. Complete development and metamorphosis takes approximately 2 months at a temperature of 26o 2o C (Castner 1988b). Validation of Larra bicolor as an Effective Classical Biological Control Agent Successful mole cricket c ontrol programs employing Larra wasps in Hawaii and Puerto Rico suggest that L. bicolor could be used as an effec tive biological control agent of Scapteriscus mole crickets in Florida (Frank 1999). Preliminary estimates suggest that L. bicolor parasitizes roughly 70% of the mole cricke ts inhabiting areas where the wasps are abundant (Frank, pers. comm. 2004). Furthermore, L. bicolor is a specialist on Scapteriscus spp. and shows no indication of harming the native northern mole cricket, Neocurtilla hexadactyla Perty or any
17 other orthopteran species (Boha rt 1976, Pruett 1991, Menke 1992). Larra bicolors reproductive potential and specificity makes this insect an ex cellent choice for combating pest mole crickets. Project Objectives Biological control ultimately re duces the costs associated with managing these destructive pests because it establishes a permanent stable contro l system which requires little or no effort to maintain. Moreover, biological control circumvent s many of the drawbacks often associated with the use of chemical pesticides, such as pest resistance and environmental pollution. Unfortunately too many biological control progr ams fall short of accomplishing their goals due to the lack of information concerning the basi c biology and behavior of the selected control species. In the case of L. bicolor significant deficiencies remain in our unders tanding of the wasps feeding preferences, fora ging behavior and fecundity. Th e main idea behind this project was to resolve these deficiencies in order to acquire a more comp lete understanding of Larras biology and behavior. The first objective was to ev aluate the viability of native plants as food sources for L. bicolor by comparing the number of adult wasps attracted to the native plant species with the number attracted to the adventive species Spermacoce verticillata The second objective was to determine the distance L. bicolor females will travel from a food source to forage for suitable hosts. This was accomplished by measuring parasitism levels in mole crickets collected at increasing distances from a large patch of S. verticillata plants, the wasps preferred food source. The third objective was to measure the potential fecundity a nd oviposition rate in female wasps by captive rearing females and c ounting the number of eggs they produced each day. The fourth and final objective was to characterize the structure of L. bicolors ovaries by taking digital photographs of the organs and quantifying important features using image analyzing software.
18 CHAPTER 2 EVALUATION OF NATIVE PLANT SPECIES FOR VIABILITY AS NECTAR SOURCES AND ATTRACTANTS Introduction There is growing concern among ecologists, c onservationists and land managers regarding the economic and ecological costs associated with the introduction of non-indigenous plant species (Gordon 1998, Curnutt 2000). The geography a nd subtropical climate of Florida make it particularly vulnerable to invasi ons by exotic species. Furthermore, periodic natural disturbances in Florida, such as hurricanes and fires, often allow adventive species to become established in areas where they would normally be excluded by biogeographical barriers or other ecological constraints (Gordon 1998, Volin 2004). One of the principal challenge s to implementing state-wide distribution and biological control of mole crickets using Larra bicolor is that the adult wasp s preferred food plant, Spermacoce verticillata is generally consider ed non-indigenous and a potential nuisance in Florida. This plant is native to the Caribbean, Central and Sout h America. However, its current distribution also includes south Fl orida, the Keys and occasional patches as far north as Alachua County in north-central Florida (Wunderlin 1998). Exactly how th is plants range was expanded into the continental United States remains unclear. It seems likely that S. verticillata could have been distributed to other locations by storms or other natural mean s (Arvalo-Rodrguez 2003). Although this plant is considered non-invasive in Florida, many turf growers and farmers are still reluctant to grow S. verticillata despite the fact that it readily attracts many beneficial insects such as L. bicolor The availability of adult food resources can be a critical factor for the success of Hymenoptera parasitoids (Baggen 1998, Ceballo 2004, Rogers 2004). Plant nectar is one of the most widely distributed and eas ily obtainable food sources for wasps and other insects (Baggen
19 1998). In the 1920s, S. verticillata was reported as being an im portant nectar source for L. bicolor (Williams 1928). Many synovigenic parasitoids, including L. bicolor rely on nectar, and pollen to provide the essential nutrients required for the metabolic demands of flight activity and ovigenesis (Leius 1960, 1967a, Takasu 1995, Le wis 1998, Eliopoulos 2003). Adult feeding positively affects parasitoid fitness parameters su ch as longevity, fecundity, and activity levels (Lewis 1998, Baggen 1998, Eliopoulos 2003, Rogers 2004). Acquiring more nourishment allows them to sustain longer periods of activity and provision eggs more qui ckly then nutritionallydeprived individuals (Takasu 1995, Lewis 1998, Wckers 2004, Wanner 2006). In addition, the physiological condition of female parasitoids infl uences the likelihood that they will search for food or for hosts (Lewis 1998, Fadamiro 2001, R ogers 2004). Well-fed parasitoids should spend more time searching for hosts instead of fora ging for food. Consequently, higher levels of parasitism are expected to occur in areas where rich nutrient resources an d suitable host species are both readily available (Leius 196 7a, Lewis 1998, Rogers 2004, Vattala 2006). Recent studies lend support to this concept by suggesting that the availability of highquality nectar and pollen sources attracts greater numbers of natural enemies to the target area which leads to an increase in their efficiency as biological controls. Su itable nectar resources may also facilitate the introduction and estab lishment of new parasitoid species (Lewis 1998, Rogers 2004, Wanner 2006). However, the mere avai lability of flowering plants may not be sufficient to guarantee a suitable nectar supply for adult parasitoids such as L. bicolor (Wckers 2004). Current efforts have concentrated on modify ing agro-ecological systems using select host plants which are known to be preferentially attrac tive to particular paras itoid species (ArvaloRodrguez 2003, Wckers 2004).This concept can be readily applied to agro-ecosystems where effective biological control of pest mole crickets is desired. Usi ng preferred host-pl ant species in
20 habitat management systems could be crucial to L. bicolor s success at controlling pest mole crickets The primary objective of this study was to id entify and evaluate the viability of native species of flowering plants as suitable nectar sources for L. bicolor Integrating harmless native plant species into our mole cricke t biological control strategy ensu res that the strained ecological balance in Florida is preserved. Growing and maintaining preferred na tive flowering plants should increase local populations of L. bicolor which will result in higher mole cricket mortality rates. This environmentally fr iendly alternative to planting S. verticillata is expected to win approval from cattle ranchers, golf course superintendents and sod farmers. Materials and Methods Four experimental native plant species were selected for evaluation: partridge pea, Chamaecrista fasciculata (Michx.) Greene [Fabaceae], golden rod, Solidago fistulosa Michx. [Asteraceae], woodland false buttonweed, Spermacoce remota Lamarck [Rubiaceae] and prostrate false buttonweed, Spermacoce prostrata Aubl. [Rubiaceae] (Bell & Taylor 1982, Taylor 1992, Wunderlin 1998, Garland 2005 pers. comm.). Chamaecrista fasciculata and S. fistulosa were chosen on the basis of historical observations of L. bicolor feeding from these plants in the field (Smith 1935, Hudson 2001, Arvalo-Rodrguez 2005 pers. comm). Spermacoce remota and S. prostrata were included because they are indigenous Spermacoce species which occur throughout Florida (Wunderlin 1998). In December 2005, a combination of seed stock and wild-collected seedlings of each plant species C. fasciculata, S. fistulosa, S. re mota, S. prostrata and S. verticillata (control), were collected from various locations around the Gain esville metropolitan area. From these seeds and seedlings, >40 specimens of each plant species were potted in commercial potting soil and grown up in a greenhouse behind the Entomology and Nemato logy building at University of Florida in
21 Gainesville, FL. Immediately foll owing planting, the seeds and seed lings were watered regularly and fertilized once per month with a solution of Peters All Purpose Plant Food 20-20-20 (NP2O5-K2O) (United Industries, St. Louis, MO). In March 2006, 15 plots that were m 1 m in size were established at two separate sites around the Gainesville metropolitan area: The Univer sity of Floridas Beef Research Unit (BRU) and Horse Teaching Unit (HTU). Each site contai ned three groups of 5 treatment plots arranged in randomized complete blocks. Blocks were sp aced > 30 m apart. Individual treatment plots were laid out in a side-by-side parallel array with 4 m of spacing between each plot. To prevent weeds and grasses from taking over the plots, eac h plot was covered with a 50 cm 100 cm 0.015 cm sheet of black polyethyl ene plastic sheeting. Plants we re grown through four evenly spaced 15 cm diam. holes, cut through the plas tic. Each plot was bordered by 1.0 cm diam. yellow, nylon rope which was supported by wooden stakes positioned at the corners. Observations of adult wasps began in July 2006, when the wasps became sufficiently abundant, and continued through November 2006. Wasp counts were conducted semi-weekly around 1:00 PM at the HTU and 2:00 PM at the BRU (peak activity time). Both male and female wasps observed visiting the plants in each indivi dual plot were counted for 1 min. A timer was used to standardize the amount of time spent at each plot. Plant condition, time of day, local temperature and basic weather conditions (sunny, partly cloudy, overcast or rainy) were also recorded for each count day. Data were analyzed by multilevel analysis of variance (ANOVA) to determine main effects from farm, month, treatment, and their interactions using Proc GLM of the Statistical Analysis System (SAS) 9.1 (SAS Institute, Cary, NC). Count values were log + 1 transformed to achieve normality. Means and standard errors for numbers of males, females and total wasps observed on
22 each plant were calculated for each month. Within each month, data were pooled across farms and treatment effects were analyzed by one-w ay ANOVA followed by mean comparisons using Duncans Multiple Range tests (MRT). Analysis was performed on log-transformed data, but untransformed means and standard e rrors are presented in the tables. Results Multi-level Analysis of Variance (ANOVA), i ndicated significant (P < 0.05) effects on males, females, and total wasps from treatment, month and treatment month interactions, but not from farm (P< 0.01). During the month of Ju ly, the greatest number of male wasps was observed on the native C. fasciculata and the control plant S. verticillata attracting averages of 22.67 11.07 and 15.00 5.60 males respectively (Table 2-1). The highest average number of females was found on C. fasciculata which attracted an average of 14.83 8.45 female wasps Table 2-2). The next largest number of female wasps was observed on the control plants, S. verticillata which had an average of 2.17 1.25 females. The native S. fistulosa plants attracted an average of 0.5 0.5 females but no males. No wasps were observed visiting either of the other two native plant species ( S. prostrata and S. remota ) in July. Patterns observed in abundance of total wasps (males + females) over tim e (Table 2-3) were similar to those observed for male wasps (Table 2-1). Similar numbers of male wasps were observed on the control plants, S. verticillata and the native C. fasciculata plants during the month of August, attracting averages of 30.17 13.96 and 37.33 18.15 male wasps respectively. However, the average number of females seen visiting the two species differed, with the native species attracting significantly more females. Small numbers of male and female wasps were also found on S. fistulosa An average of 0.33 0.21 males was observed on the native S. prostrata plants, but no males or females were seen on S.
23 remota in August. Average numbers of males and females recorded for S. fistulosa, S. prostrata and S. remota were not statistically different from zero. Comparable numbers of male and female wasp s were observed visiting the control plants and the native C. fasciculata plants in September. Spermacoce verticillata attracted averages of 23.50 6.85 males and 2.17 1.22 females; C. fasciculata attracted averages of 18.67 11.62 males and 2.83 1.05 females. A small number of males, 0.5 0.5, but no females were also seen on the native S. prostrata plants. No wasps were detected on either S. fistulosa or S. remota in September. During the month of October, the greatest numbers of both male and female wasps were observed on the control plants, which attract ed averages of 39.00 22.52 males and 1.67 1.09 females. The only other plant vi sited by wasps in October was S. fistulosa which attracted an average of 0.83 0.65 males but no female s. The number of males observed on S. fistulosa in October was significantly less than the number of males recorded for the control plant. The control plant, S. verticillata continued to categorically outperform the four native plant species in November, attracting an aver age of 30 23.18 male wasps. In comparison, 2.00 2.00, 2.00 1.30, and 0.10 0.10, males were recorded visiting C. fasciculata, S. fistulosa and S. remota respectively. November was the only month that wasps were seen on S. remota Despite attracting a large number of males, the relatively low number of females found on the S. verticillata plants was not statistically different from the numbers of females seen on the four native plant species. An average of 1.0 0.63 fe males was observed on the control plant while C. fasciculata and S. fistulosa averaged 0.40 0.40 and 0.40 0.24 females respectively. No wasps were found on S. prostrata in November.
24 Discussion This study investigated the viab ility of using native plants to attract the mole cricket hunting wasp L. bicolor In the past, patches of non-native S. verticillata have been established throughout Florida in efforts to increase and enhance local popul ations of this beneficial parasitoid. Unfortunately, many tu rf growers and land owners are reluctant to use non-native plants for habitat management. Therefore, I eval uated the potential of four native species by comparing the numbers of wasps observed on the native plants, to the numbers of wasps found on adjacent S. verticillata plants. The five plant species displayed extensive dive rsity in their basic phe nology, growth rates and number of nectar sites (both floral and ex tra-floral). Hence, no effort was made to standardize plant size or flower number among species. Attempti ng to artificially standardize these variables across the different species w ould have significantly altered their primary structures and number of availa ble nectaries which could negativel y affect the plants ability to attract L. bicolor Therefore, the different plants were allowed to grow and develop naturally because we wanted to accurately imitate natural growth conditions. My results indicate that only one of the four native plant species ( C. fasciculata ) attracted a significant number of wasps. In fact, these pl ants attracted at least as many wasps as the S. verticillata control plants during the m onths of July, August and Sept ember. However, during the months of October and November S. verticillata continued to attract large numbers of wasps while C. fasciculata attracted very few if any. Many wasps were observed on Spermacoce verticillata during the final two months of the study because this plant continued to produce flower blossoms, where as C. fasciculata began to display signs of senescence in early October. By November all the C. fasciculata plants had withered and died. It is interesting to note that a
25 few wasps were observed on this plant in Novemb er. However, it seems likely that the wasps were simply taking shelter in the plants tangle of dried branches. Contributing to the mortality of C. fasciculata was the plants apparent sensitivity to cooler temperatures. During the 2nd week of October, the evening temperatures around the Gainesville metropolitan area descended to 10-15oC and by the 4th week, temperatures fell below 10oC (NCDC 2007). Coinciding with the decrease in temperature was a conspicuous deterioration in the condition of C. fasciculata In contrast, the control plants were able to persist well through December that year. Exposure to cooler temperatur es in October could have triggered premature mortality in C. fasciculata which would explain why these plants ceased to attract L. bicolor in the final two months of the study. There is some ambiguity regarding the lifespan of C. fasciculata According to R. P. Wunderlins Guide to the Vasc ular Plants of Florida, C. fasciculata blooms only during spring (March-May) and summer (June-S eptember) (Wunderlin 1998). Richard Weaver from the Division of Plant Industry in Ga inesville, FL, believes that C. fasciculata plants typically live 2224 weeks (Weaver 2007 pers. comm). However my pl ants lived approximately 36 weeks. It is unclear why my plants lived longer than their prev iously reported lifespan. My plants resilience suggests that C. fasciculata s persistence is probably influenced more by temperature and weather conditions rather than a predetermined duration period. The lack of wasps observed on S. prostrata was unexpected given its similarity in appearance to S. verticillata In fact, the two species can on ly be distinguished by a single characteristic. The calyx lobes of S. prostrata have green centers with white margins, whereas the calyx lobes of S. verticillata are uniformly green (Wunderlin 1998). The lack of wasps observed on S. prostrata could be explained by the fact that S. prostrata did not flourish as well
26 as S. verticillata Although these native plants prospere d from March through May, they did not seem to cope well with the higher temperatures and lack of water in the summer months. By the end of June many of the plants appeared st ressed and ceased growing or producing flower blossoms. This plant is typically found in wet fl atwoods and floodplain fore sts, indicating that it prefers habitats with moist soils and partial sunlight (Wunderlin 1998). These wildflowers were planted in horse and cow pastures which have drier, sandy soils and receive full sunlight. These conditions probably made it difficult for S. prostrata to grow successfully and produce enough floral nectar to attract L. bicolor Spermacoce remota, formerly named S. assurgens (Ruiz & Pav n), also failed to attract many wasps. Interestingly, this species was lis ted by James Castner as a native plant which L. bicolor fed from frequently in both the laborat ory and the field (Castner 1988a). Castners findings differ sharply from my results. The explanation fo r this discrepancy remains unknown. However, plants that typically attract L. bicolor also tend to attract many other species of insects particularly Hymenoptera, Dipter a, and Coleoptera. During my surveys I noted a conspicuous lack of other insects visiting this plant. This observation supports the hypothesis that S. remota is not an important nectar source for L. bicolor Another native plant reported by Castner as being attractive to L. bicolor is Chamaesyce hirta (L.) Millsp. (Castner 1988a). In this instance, my own field observations confirm Castners assessment. Chamaesyce hirta grows throughout the state of Fl orida and is commonly found in pastures and disturbed sites. According to W underlin, this plant flowers all year (Wunderlin 1998). However, my observ ations indicate that L. bicolor could only be found on this plant during the fall and winter months Although this species may have potential as a nectar source
27 for L. bicolor it was excluded from my evaluations, because this plant propagates so rapidly and aggressively that it often become s a nuisance for turf growers. Based on my evaluations of four nativ e plants species, I conclude that C. fasciculata could be used successfully to enhance local populations of L. bicolor particularly in the summer months. However, due to the decline in C. fasciculata s ability to attract large numbers of wasps during October and November, this plant may be less suitable for use in habitat management systems in regions of Florida which experience cooler temperatures during these months. The prospects for this plant are more promising farther south where C. fasciculata persists year-round by continually reseeding itself (W eaver 2007. pers. comm.). In southern Florida, this native wild flower should attract large numbers of wasps th roughout the year. In addition to its ability to attract beneficial insects, the plants numerous bright yellow fl owers make it rather visually pleasing. Its aesthetic app eal would allow this plant to be us ed in broader c ontexts such as decorative landscapes and flower ga rdens. Turf growers and landowne rs who seek to control pest mole crickets by attracting L. bicolor but are concerned about ma intaining non-native plants, should be encouraged to grow C. fasciculata instead of S. verticillata
28 Table 2-1 Number (x SEM) of adult males observed on each plant species. Values represent counts from all plots (2 locations 3 repli cations). Means with same letters are not significantly (P< 0.05) different according to Duncans MRT on log-transformed data). Table 2-2 Number (x SEM) of adult females observed on ea ch plant species. Values represent counts from all plots (2 locations 3 repl ications) Means with same letters are not significantly (P< 0.05) different according to Duncans MRT on log-transformed data. Table 2-3. Number (x SEM) of total wasps (males + females) observed on each plant species.Values represent counts from all plot s (2 locations 3 replications). Means with same letters are not significantly (P < 0.05) different according to Duncans MRT on log-transformed data. Treatment July August September October November Spermacoce verticillata 15.00 5.60 A 30.1713.96 A 23.50 6.85 A 39.0022.52 A 30.2023.18 A Chamaecrista fasciculata 22.6711.07 A 37.3318.15 A 18.6711.62 A 0.00 B 2.00 2.00 B Solidago fistulosa 0.00 B 0.67 0.67 B 0.00 B 0.83 0.65 B 2.00 1.30 AB Spermacoce prostrata 0.00 B 0.33 0.21 B 0.50 0.50 B 0.00 B 0.00 B Spermacoce remota 0.00 B 0.00 B 0.00 B 0.00 B 0.10 0.10 B Treatment July August September October November Spermacoce verticillata 2.171.25 A 1.330.80 A 2.171.22 A 1.671.09 A 1.0 0.63 A Chamaecrista fasciculata 14.838.45 A 10.505.48 A 2.831.05 A 0.00 B 0.40 0.40 B Solidago fistulosa 0.500.50 B 0.500.50 B 0.00 B 0.00 B 0.40 0.24 AB Spermacoce prostrata 0.00 B 0.00 B 0.00 B 0.00 B 0.00 B Spermacoce remota 0.00 B 0.00 B 0.00 B 0.00 B 0.00 B Treatment July August September October November Spermacoce verticillata 17.17 6.54 A 31.5014.57 A 25.67 7.86 A 40.6723.60 A 31.2023.46 A Chamaecrista fasciculata 37.5019.25 A 47.8323.62 A 21.5012.38 A 0.00 B 2.40 2.40 B Solidago fistulosa 0.50 0.50 B 1.17 1.17 B 0.00 B 0.83 0.65 B 2.40 1.47 AB Spermacoce prostrata 0.00 B 0.33 0.21 B 0.50 0.50 B 0.00 B 0.00 B Spermacoce remota 0.00 B 0.00 B 0.00 B 0.00 B 0.10 0.10 B
29 CHAPTER 3 EFFECT OF ADULT HOST PLANTS ON FORAGING PATTERNS AND LOCAL DISTRIBUTION Introduction Parasitoids make up more than 10 % of all met azoan species. Most belong to three orders of insects: Hymenoptera, Dipt era and Coleoptera (Hassell 200 0, Borror and Delong 2005). Adult parasitoid females usually oviposit one or more eggs on, i n, or near a single invertebrate host (Comins 1996, Hassell 2000). After hatching from their eggs, parasito id larvae subsist by consuming the tissues of their hos ts. All the nutritional resources, necessary to complete larval development and achieve maturity, come fr om their hosts (Rivero-Lynch 1997, Heimpel 1998, Jervis 2001). The larvas feeding activity typicall y weakens and eventually kills the host. Unlike typical predators, which consume their prey i mmediately, parasitoid larvae must delay killing their hosts until the parasitoid larvae are fully developed (Rivero-Lynch 1997, Heimpel 1998, Hassell 2000, Jervis 2001). In addition to their importance as biological controls, parasitoids are also useful models for studying ecological theories of competition and predator-p rey distribution (Mattiacci 1999, Hassell 2000, Darrouzet-Nardi 2006). Competition i nvolves the struggle to obtain the necessary resources for survival and reproduction (Price 1997, Molles 2 002). Synovigenic hymenopteran parsitoids, such as Larra bicolor require two essential resour ces in order to survive and reproduce successfully: suitable hosts and food (Lewis 1998). Typically food and hosts are arrayed in uneven or patchy distributions. In addi tion they are often not simultaneously present at the same general location (Desouhant 2005, Wanner 2006). A divergent distribution of food and hosts is typical for agricultural and ornament al landscapes because these methods produce a mosaic of simple monocultures bordered by more complex habitats. Furthermore, foraging
30 distances for predators and parasitoids tend to be greater in simple monocultures compared with complex landscapes (Tscharntke 2004). When adult female parasitoids are challeng ed with environments where hosts and food are found in different locations, th e optimal decision to search fo r hosts or food strongly depends on the energy reserves of each female and the probability of pinpointing new hosts (Baggen 1998, Lewis 1998). When a females energy reserves are high she should search for more hosts. Conversely, when her energy reserves are low, she should disrupt host foraging in order to locate food. At extreme probabilities of finding food, female parasitoids should continue to search for hosts (Desouhant 2005). Consequently, landscape f eatures in spatially patchy environments, such as the presence of flowering plants, can consid erably influence the population and behavioral dynamics of natural enemies and their prey (Comins 1992). Female parasitoids generally travel from one resource patch to another by flight (Desouhant 2005). Insect flight is very energy-d emanding. (Desouhant 2005, Wanner 2006). During flight, an insects metabo lic rate increases 50-100 fold compared with its metabolism at rest (Wanner 2006). Due to their substantial energy require ments, adult wasps must locate food frequently to avoid starva tion (Lewis 1998, Lavandero 2005). This implies that they must periodically interrupt their host foraging activities to find food, but flying long distances from host patches to search for food is costly in terms of energy loss and mortality risks (Baggen 1998, Lewis 1998). To minimize these costs, adult parasitoids should travel shorter distances between host patches and food sources (Wanner 2006). Resource availability may be more important than habitat area to population survival and biodiversity (Tscharntke 2004). Applied studies have revealed th at the presence of food plants affects habitat preferences by parasitoids by attrac ting and retaining these natural enemies to the
31 target areas (Takasu 1995, Lewis 1998). Additionally, the presence of food sources have been correlated with increased levels of parasitism (Mattiacci 1999, Rogers 2004). Conversely, parasitoid populations and para sitism rates in target patche s without food were significantly reduced (Lewis 1998). Feeding experience in the field also appear s to improve foraging efficiency since parasitoids tend to spend more time searching for hosts near the food source (Takasu 1995, Lewis 1998, Rogers 2004). Parasitoid foraging activity is an important be havioral characteristic for species used in biological control (Wanner 2006). Female foraging and oviposition decisions dictate their dispersal and spatial distribution within landscapes. Their decisions may, in turn, be based on the distribution patterns of resources such as food and hosts (Chow 2000). Parasitoid females foraging in low-quality resource patches can increas e their odds of encount ering a higher-quality patch by dispersing (King 2005). However, migration between pa tches can have a detrimental effect on population stability because life history tr aits are often correlat ed with other factors imposing reproductive constraints on dispersers (Comins 1992, Desouhant 2003). Therefore population dynamic models emphasize the importan ce of density dependence and low levels of dispersal (Darrouzt-Nardi 2006). Habitat quality is important in biological cont rol because the lack of resources can trigger dispersal by parasitoids, thereby reducing th eir effectiveness at managing pests (King 2005, Wanner 2006). Moreover, the effects of habitat arrangement can be even more pronounced in simpler agricultural la ndscapes (Tscharntke 2004). Curre ntly it is not well understood how habitat manipulations will effect the local distributions of L. bicolor females. For example, it is unclear how far female wasps will travel from nectar sources while hunting for suitable hosts. The second objective of this proj ect addresses these uncertainties by measuring the total number
32 and the percentage of mole crickets parasitized by L. bicolor collected in a seri es of pitfall traps located at increasing distances from a rich nect ar source. Ultimately, in formation regarding the host foraging range of the wasp will provide turf growers with useful guidelines for establishing patches of nectar source plants Moreover, insight gained rega rding seasonal changes in the levels of parasitism could help maximize th e efficiency of other control treatments. Materials and Methods In August 2005, a plot 1 m 3 m in size, containing 16 S. verticillata plants, was established at a private horse fa rm (Duncan Farm), located about 10 km north of the town of Hampton in north-central Florida. The plants were grown and maintained using the same methods outlined in Chapter 2. To preven t weeds and grasses from crowding the Spermacoce plants, the soil surface of the plot was covered wi th a sheet of black polyethylene plastic 1 m 3 m 0.015 cm in size. The plants grew through two rows of evenly spaced 15 cm diam. holes cut through the plastic. The plot wa s surrounded with barbed wire to discourage the horses from destroying the plants. A linear array of 10 pitfall traps, spaced at 20 m intervals, extended out from the long edge of the plot. Each trap was constructed from four, 3 m sec tions of split 7.62 cm PVC pipe laid out at approximately 90o angles in a cross-shaped pattern (L awrence 1982). All the pipes were capped at the far end and buried at gr ound level. Buried where the pipe s intersect, were 20 L plastic buckets, with 4 large holes bored into the sides to accommodate the 4 sections of pipe. Smaller inner buckets filled with moist sand were placed inside the 20 L buckets. Shorter 10.5 cm lengths of thin-walled 7.62 cm PVC pipe were fitted in to the larger pipes and extended into the 20 L buckets. These extender tubes were removable so that the smaller buckets could be easily removed and cleaned out. Mole crickets, trappe d in the pipes would make their way along until they dropped off the end of the extender tubes into the inner bucket.
33 All traps were monitored 9 times from Oc tober-December 2005 and sixteen times from August-December 2006. Mole crickets caught by the traps, were take n back to the University of Floridas Entomology and Nematology building in Gainesville, FL, and inspected under a stereomicroscope for the presence of L. bicolor larvae or eggs. Numbers of S. borellii and S. vicinus mole crickets caught, their corresponding pronot al lengths and presence of parasites were recorded for each trap. Parasitoid larval instars were also noted. Average monthly means and standard errors we re calculated for total cricket numbers and percent parasitism. Yearly means a nd standard errors for numbers of mole crickets, parasites, and percent parasitism were calculated for all distances from the S. verticillata plot, Statistical Analysis System (SAS) 9.1 (SAS Institute, Ca ry, NC). Correlation coefficients (r) were calculated between all variables to determine an y relationships between them. Using the Proc Reg procedure of (SAS) 9.1, regression analysis determined the exact linear pattern of each significant (P< 0.05) relationship. Results The average monthly mole cricket catc h increased by 42% from 2005 to 2006. The highest levels of parasitism for both years were observed in the months of November and December (Table 3-1). Due to wide variation betw een samples, the numbers of mole crickets and parasitoids recorded from each trap were averag ed over yearly periods. The average numbers of mole crickets and parasitoids recorded fo r 2005 ranged from 2.56 1.23 to 7.67 2.03 crickets and 0.11 0.11 to 2.00 0.90 parasitoids per sample In 2006, the average numbers ranged from 4.44 0.88 to 20.56 4.84 crickets and 0.50 0.18 to 2.56 0.58 parasitoids per sample. Correlations between the number of crickets caugh t in the traps and the number of associated parasitoids were highly significant (r =0.84, P< 0.003, N=10) for 2005 and (r=0.84, P=0.002, N=10) for 2006.
34 Numbers of parasitoids present at a partic ular location increased positively with the numbers of mole crickets also presen t in 2005 (0.2874X 0.1614) and in 2006 (0.1104X + 0.2731) (Figure 3-1). The relationship was highly significant for 2005 (R2=0.7114, P< 0.003) and 2006 (R2=0.718, P< 0.003). In fact, the relations hip was nearly identical for both years. Correlation analysis indicated a negative correlation be tween distance from the plants and the number of parasitized mole crickets per trap for 2005 (r= -0.54, P< 0.11, N=10) and 2006 (r=0.64, P< 0.05, N=10). The correlation was stronger in 2006 than in 2005. Regression analysis (Figure 3-2) shows a general declin e in the number of parasitized mo le crickets as their distance from the host plants increased in 200 5 (-0.0056X + 1.7556) and 2006 (-0.0067X + 2.0042). The relationship was weaker for 2005 (R2=0.29, P<0.11) compared to 2006 (R2=0.41, P< 0.05). These results suggest that the probability of findi ng parasitized mole crickets diminishes with greater distance from L. bicolors food source. Discussion Previous studies found strong correlations between nearby flowering vegetation and greater parasitism ra tios (Lewis 1998, Rogers 2004, Lee 2006, Wanner 2006). In the case of the mole cricket parasitoid, L. bicolor it was not known how a nearby nectar source would affect local parasitism levels or the wasps distri bution patterns. These effects were examined by measuring the proportion of mole cr ickets which were parasitized by L. bicolor Crickets were collected with a series of pitfall traps, pos itioned at increasing distances (up to 200m) from a patch of S. verticillata plants, the preferred host plant of L. bicolor The data indicate that the mole cricket popula tion in the area was not evenly distributed. This was not surprising given that most insect s tend to exhibit clum ped distributions (Price 1997). Parasites also exhibited a clumped distribution which corre lates with the distribution of mole crickets. These results (Figure 3-1) suggest a strong positive relationship between the
35 number of mole crickets caught and the number of associated parasites. This indicates that greater numbers of parasitized mole crickets ca n be found in areas with higher cricket population concentrations. Additional analysis revealed th at the percentages of parasiti zed mole crickets caught at each location were highly variable and did not si gnificantly correlate with cricket abundance or distance from the nectar source. These findings imply that L. bicolor females are not concentrating their efforts in areas where mole cr ickets are more prevalent, rather their foraging patterns appear to be more random. When females forage randomly, the probability of encountering suitable hosts is dependent on the de nsity of mole crickets inhabiting the foraging area. The likelihood of encounteri ng hosts increases in high density patches and deceases in low density patches. Consequently, more parasites sh ould be present in high density host patches. Furthermore if females are foraging randomly, th e percentage of mole crickets parasitized in each patch should vary considerably depending on the number of individuals present and the success rate of the wasps. If L. bicolor females were concentrating their search efforts on hostrich patches, parasitism levels would correlate more closely with host density. However, the results did not support this so the hypothesis that L. bicolor females focus their host searching activity on high density patc hes can be rejected. The data also indicated that the distribution of parasitized mole crickets was affected by their distance from the nectar producing plants. Th e second set of results (Figure 3-2) shows that the number of parasitized mole crickets declines significantly as their distance from the wasps host plants increases. As previously mentioned, no significant correlation exists between trap distance and the number of mole crickets captured. Therefore I reject the hypothesis that the decline in parasitism leve ls, relative to distance, was due to a corresponding decrease in mole
36 cricket abundance. These results suggest that L. bicolor s host searching patterns are not completely random, but rather they spend significa ntly more time searching for hosts close to familiar nectar sources. By integrating the conclusions from both sets of results, we recogni ze that parasitism levels are influenced by two as pects of mole cricket population demographics; the populations regional density and its proximity to high-quality nectar sources. Both factors have important implications concerning habitat management and biological control. These results suggest that the most effective use of habitat management wo uld be to focus on areas that are inhabited by dense populations of mole crickets. Nectar source patches established in close proximity (< 100 m) to these areas could increase and enhance the resident L. bicolor population, thereby maximizing the number of mole crickets encountered by the wasps. This should result in a rise in parasitism levels and a proportional re duction in the mole cricket populations. Table 3-1 Number (x SEM) of mole crickets caught in pitfall traps and pe rcent parasitized by Larra bicolor Values represent monthly sampling means (2005: Oct N=2, Nov N=5, Dec N=2; 2006: Aug N=1, Sept N=4, Oct N=4, Nov N=6, Dec N=1). 2005 2006 month Total Crickets Crickets/Trap % Parasitism Total Crickets Crickets/Trap % Parasitism August 57.0 0.0 5.7 0.0 5.3 0.0 September 48.3 7.4 4.8 0.7 5.8 2.2 October 35.0 13.0 3.5 1.3 6.4 1.9 92.5 53.8 9.3 5.4 9.8 4.0 November 42.8 10.1 4.3 1.0 31.5 6.4 118.8 31.8 11.9 3.2 18.7 1.5 December 57.0 44.0 5.7 4.4 26.9 6.4 70.0 0.0 7.0 0.0 17.1 0.0
37 AR2 = 0.7114 P < 0.0030.0 0.5 1.0 1.5 2.0 2.5 2345678Number of mole cricketsNumber of parasites BR2 = 0.718 P < 0.0030.0 0.5 1.0 1.5 2.0 2.5 3.0 468101214161820Number of mole cricketsNumber of parasites Figure 3-1 Average number of pa rasites present as a function of the number of mole crickets captured in 10 pitfall traps placed at 20 m intervals extending out from a patch of nectar source plants. (A) Numbers reco rded for 2005. (B) Numbers recorded for 2006. AR2 = 0.2884 P < 0.110.0 0.5 1.0 1.5 2.0 2.5 020406080100120140160180200220Distance from nectar source (m)Number of parasites BR2 = 0.4095 P < 0.050.0 0.5 1.0 1.5 2.0 2.5 3.0 020406080100120140160180200220Distance from nectar source (m)Number of parasites Figure 3-2 Average number of parasites present as a function of distance from the adult wasps preferred food source. (A) Numbers of para sitoids recorded in 2005. (B) Numbers of parasitoids recorded in 2006.
38 CHAPTER 4 POTENTIAL FECUNDITY AND OVIPOSITION RATE Introduction Fecundity and oviposition behavior are fundame ntal aspects of an insects lifehistory, ecology, and population dynamics. Fecundity refers to the number of eggs or offspring (for species that bear live young) produced by individual females (Minkenberg 1992, Molles 2002). This life history trait is part icularly important for parasito ids because of their frequent application in biological control (Mills 2000). Parasitoids that have rela tively high fecundity tend to be more successful at reducing pests because they have the ability to reproduce rapidly in direct response to increases in pest abundan ce (Ceballo 2004). Hence, a clear understanding of the reproductive strategies of parasitoid natura l enemies is necessary for improving the success and efficiency of these insects as biological controls (Eliopoulos 2003, Zhang 2004). Reproductive success of female parasitoids is de termined by the number of eggs that they oviposit in their lifetime and the survivorship of the offspring (Minkenb erg 1992). A parasitoids overall reproductive performance can be characteri zed by three elements which are related to their total reproductive output: maximum fecundit y, potential fecundity a nd realized fecundity. Maximum fecundity represents the greatest numbe r of eggs that an individual female can produce under optimal conditions. Potential fecundity denotes the average number of eggs that a population of females can generate under optimal conditions. Realized fecundity is the average egg output by the corresponding population under a sp ecific set of field conditions (Mills 2000). Maximum fecundity is a genetically determin ed characteristic which may vary among species but does not vary on average within a species. A populations po tential fecundity is reduced from the species maximum by factors infl uencing the quality and quantity of resources acquired during juvenile development. Realized fecundity of a population is further diminished
39 below its potential fecundity by factors which aff ect the longevity of the females, oogenesis and oviposition behavior (Mills 2000). A females total egg output is limited by a combination of factors, such as the number of mature eggs that she can produce during her lifetime, the number of suitable hosts she can locat e, and her oviposition rate. Due to the importance of fecundity as a life history and population characteristic, it is surprising that there is a substant ial lack of data on potential f ecundity and even fewer estimates of realized fecundity (Leath er 1998, Mills 2000). Many estimate s of insect fecundity are calculated from easily measured parameters lik e female body size, egg load or number of ovarioles. Although such related parameters may produ ce accurate assessme nts of fecundity, these associations can be invali dated by other interfer ing factors which were overlooked (Leather 1998, Mills 2000). For instance, adult weight may be used to estimate the total number of eggs contained within an individual female, but this quantity may have little bearing on the actual number of eggs she deposits (Mills 2000). Consequently, estimates of both potential and realized fecundity based on such shortcuts should be regarded with skep ticism (Leather 1998). The best method for attaining a better understanding of a parasitoids pote ntial and realized fecundity is to measure its lifetime reproductive rate directly, by counting the number of eggs each female lays beginning the day she ecloses until her death (Sweetman 1936). Although the values obtained by these methods represent to tal fecundity under ideal conditions, the biology and behavior of most parasitoids make obtaining fecundity data from field populations extremely difficult. The aforementioned situation is analogous for the mole cricket hunting parasitoid L. bicolor Although L. bicolor has been used for almost 20 years to combat Scapteriscus mole crickets in Florida, significant uncertainty re mains regarding the mean number of eggs these wasps lay daily or their tota l lifetime reproductive output. The objective of this study was to
40 measure L. bicolors potential fecundity and da ily oviposition rate. This study will also help to elucidate egg-laying patterns w ithin days and across days th roughout the wasps lifetime. Larras reproductive output will be measured by allo wing captive-reared females to oviposit on an abundance of host mole crickets. By dividi ng potential fecundity with the developmental success ratio for the eggs, we will be able to more accurately estimate L. bicolors practical impact on mole cricket populations. Material and Methods Beginning in the fall of 2006, Scapteriscus mole crickets parasitized by L. bicolor were collected in pitfall traps. The parasitoids were allowed to grow and deve lop on their mole cricket hosts. In order to prevent the crickets from pr eying on one another, single crickets were housed in separate vials filled with moist sterilized sa nd. To keep the mole crickets alive long enough for the larvae to complete their development, the crickets were fed FRM Cricket & Worm Feed (Flint River Mills Inc. Bainbridge, GA, USA) twi ce per week and water was added to the sand as needed. Environmental conditions were standardi zed by keeping the vials containing parasitized crickets in a Florida Reach-In e nvironmental chamber (Walker 1993) (27oC, humidity 55%, 16/8 hr L/D). Once the wasp larvae construc ted their cocoons, they were carefully removed from the vials. Cocoons were sorted and separated based on size in order to keep the sexes separate. The large cocoons were presumed to house females and small cocoons contained males. Female and male cocoons were placed into separate clear polystyrene observation boxes filled halfway with moist sterilized sand. The observation boxes were also kept in a Florida Reach-In chamber (27oC, humidity 55%, 12 hr L/D) and monitored daily for adult activity. Upon eclosing, adult wasps were removed from their plastic observation boxes and placed into a plastic shell vial. Wasps were first weighe d inside the plastic vials then released into
41 oviposition arenas. Single females were placed in an arena with one or more males. Each oviposition arena was made by placing 2.5 cm of moist sand on the bottom of a 30 cm screened insect cage. A small potted S. verticillata plant and an artificial nectary were also placed inside each cage to provide food for the adu lt wasps. Artificial nect aries were constructed from small glass vials filled with a 50/50 mixture of honey and 20 % sucrose solution topped with an absorbent cotton wick. A ll oviposition arenas were kept in a temperature-regulated (85o C) greenhouse behind the Univer sity of Floridas Entomology and Nematology building in Gainesville, FL. Seven Scapteriscus mole crickets were added to e ach oviposition arena. Typically 4-5 S. vicinus and 2-3 S. abbreviatus were offered to each female. Adult S. vicinus were collected from the field using sound traps. Adult S. abbreviatus were acquired from the Department of Entomology and Nematologys mole cricket rearing facility. Before being offered to the females, field caught mole crickets were thoroughly inspected to ensure that no previous L. bicolor eggs or larvae were present. Cricke ts were left in the cages fo r approximately 24 hr. Because Larra wasps are diurnal, their feeding, ovipos ition and flight activities would ceased 2 hr before sunset. Therefore, all crickets were remove d from each cage around 6:00 pm and checked for eggs. Eggs were carefully detached from cricke ts using forceps and the crickets were placed back into clean vials with mois t sand and food. Previously parasiti zed crickets were allowed to recover for several days before being offered to the wasps again. Parasitized crickets were always replaced by fresh Scaptericus specimens. This cycle was repeated daily until the female wasps died. Analyses were performed using Statistical Analysis System (SAS) 9.1 (SAS Institute, Cary, NC). Means and standard errors were calc ulated for the following va riables: wasp weight,
42 lifespan (days), total number of eggs and eggs per day. Correlation coefficients (r) were calculated between each pair of variables to dete rmine if any relationships existed between them. Regression analysis was used to determine the exact pattern linear equa tion of each significant (P< 0.05) relationship. Results The relative size of each female wasp was quan tified by its weight. Often hind tibia length is used to quantify inter-specific size variation in insects. However, it was established that female weight corresponded very strongly with tibia length for this sp ecies (data not shown)(r=0.89, P< 0.0005, N=10). Average, minimum and maximum va lues for wasp weight, lifespan (days), lifetime fecundity, and oviposition rate (eggs/day ) are shown on table 4-1. The average lifespan for this sample of females was 23.5 1.87 days. The sample consisted of six females that stayed alive for 30 days, seven that lived 20-29 days, six th at lived 10-19 days and 1 that survived < 10days. The shortest and longest lived individuals ranged from 8 to 40 days respectively, a difference of 32 days. The females produced an average 56.05 4.38 eggs during their lifetime. Their minimum and maximum lifetime output range d from 17 to 91 eggs respectively. The average daily oviposition rate for this group of females was 2.44 0.14 eggs per day with a range difference of 10 eggs per day. Their aver age daily oviposition ra te ranged from 1.32 to 3.77 eggs per day (data not shown). Correlation analysis revealed a significant pos itive relationship between female weight and oviposition rate (r=0.67, P< 0.006, N=15). Ther e was also a strong correlation between female lifespan and lifetime fecundity (r=0.80, P< 0.001, N=20). Daily oviposition rate (y) increases linearly (y= 0.0121x + 1.2027) as female weight (size) (x) in creases (R2=0.4576, P< 0.006 N=15) (Figure 4-1). This result indicates th at larger females are capable of ovipositing successfully more often than smaller females. Lifetime fecundity of L. bicolor (y) is linearly
43 related (y=1.8751+ 11.985) to the number of days that the fe male survives (x) (R2=0.643, P< 0.001, N=20) (Figure 4-2). This result suggests th at females which survive longer will produce larger clutches than short-lived individuals. Discussion Fecundity and reproductive success are key f acets of female parasitoid biology. Despite L. bicolor s importance for biological control of Scapteriscus mole crickets, very little was known regarding its reproductive capab ilities and limitations. By rearing 20 adult females and maintaining them in a controlled environmen t, I was able to observe and measure their oviposition rates and total egg out puts. Although a few individuals were relatively short-lived, the majority (about 2/3) of the wasps lived more than 20 days. Based on these results, I conclude that L. bicolor survives for at least 3-4 weeks as adults. My results show two significant relationships related to female fecundity. The first relationship indicates that larger females are ovi positing more frequently than smaller females. The precise reason for this inequality is unclear Perhaps larger female s have the ability to mature eggs faster than smaller females. This situation appears plausible if we assume that all females lay eggs which are roughly the same size and the process of yolk provisioning is nutrient-limited and not rate-limited. Additionally venom is physiologically expensive to manufacture and is probably depleted rapidly be cause these wasps administer multiple stings to their prey. Large females should be equipped with proportionally bigger crops than small females. If large females are ingesting greater vol umes of nectar, they po tentially consume more nutrients than lesser females. Greater quantities of nutrients would allow large females to be able to provision eggs and replenish th eir venom supply faster than sm all females. As a result, large females will be ready to oviposit mo re often than small females.
44 The second relationship suggests that female s that survive for longer periods of time produce a greater number of eggs than shorter-lived females. The idea that total egg output is determined by longevity is not surprising given that L. bicolor females are synovigenic and therefore only capable of producing a finite numbe r of eggs per day (typically 2-3). In addition, the average egg rate for my sample of females (2.44 eggs per day) showed only minor variation (SEM=0.14) which indicates that there is little population variation in the daily oviposition rate of this species. Therefore I hypot hesize that the main factor aff ecting the lifetime fecundity of L. bicolor females is the duration of their lifespan. These results have significant implications for managing this species as a biological control of pest mole crickets. At present, th ere is ample evidence suggesting that parasitoid fitness parameters, such as l ongevity and fecundity are positiv ely affected by adult feeding (Baggen 1998, Eliopoulos 2003, Ceballo 2004, Rogers 2004). Hence, habitat management could be used to maximize L. bicolor s ability to parasitize pest mo le crickets. By providing these insects with plentiful nutrient resources such as nectar, we may be able to boost the cricket killing capabilities of local wa sp populations by increasing female longevity. This could substantially reduce local mole cricket populations and minimi ze the damage that they cause. Table 4-1 Summary statistics for experimental variables. Values represent mean SEM, minimum and maximum for wasp weight (N=15), lifespan (N=20), lifetime fecundity (N=20), and oviposition rate (N=20). Mean SEM Minimum Maximum Wasp Weight (mg) 107.43 8.71 58 185 Lifespan (days) 23.5 1.87 8 40 Lifetime Egg Output 56.05 4.38 17 91 Daily Egg Output 2.44 0.14 0 10
45 R2= 0.4576 P < 0.0061 1.5 2 2.5 3 3.5 4 406080100120140160180200Wasp weight (mg)Eggs per day Figure 4-1 Average number of e ggs oviposited per day as a func tion of female weight (N=15). R2= 0.643 P < 0.000110 20 30 40 50 60 70 80 90 100 110 51015202530354045Wasp lifespan (days)Lifetime egg output Figure 4-2 Total lifetime egg production as a function of female lifespan (N=20).
46 CHAPTER 5 CHARACTERIZATION OF OVARIAL ULTR A-STRUCTURE AND QUANTIFICATION OF INTRASPECIFIC VARIATION Introduction Parasitoid females lay their eggs on, in or near invertebrate hosts. After hatching from the eggs, parasitoid larvae must feed on the hosts tissues in order to obtain the nutritional resources required to reach maturity. Larval feeding activit y typically weakens and eventually kills their hosts (Rivero-Lynch 1997, Heimpel 1998, Jervis 2001). The larvae from many parasitoid wasps, such as Larra bicolor live as koinobionts (C astner 1983, 1989). Koinobion t parasitoid larvae feed and develop on active hosts. In this type of host-parasitoid associa tion, the parasitized hosts continue to grow and develop (as immatures) or even reproduce (a s adults) before to being killed by the parasitoid (Jervis 2001). Koinobiont parasitoid wasps can be categori zed as either pro-ovi genic or synovigenic based on their method of egg production (Fla nders 1950, Heimpel 1998, Jervis 2001). Before oviposition, pro-ovigenic parasito ids already possess a fixed numb er of mature eggs. The egg capacity of pro-ovigenic parasitoids is determ ined by the amount of nutrients acquired by the larval stages. Conversely, synovi genic parasitoids, such as L. bicolor continue to provision and produce mature eggs throughout their reproduc tive life (Flanders 1950 Jervis 2001). Synovigenic species typically obtain the nutrients necessa ry for egg production from feeding on nectar, honeydew or host tissues (Heimpel 1998, Jervis 2001, Eliopoulos 2003). Synovigenic koinobiont parasitoids are especially valuab le as biological controls becaus e they can attack the adult or immature stages of a host. This enables these pa rasitoids to develop succe ssfully under a greater range of host population stru ctures (Ceballo 2004). The reproductive strategies and capabilitie s of parasitoids are constrained by their evolutionary history and anatomy (i.e., mor phology of their reproducti ve organs) (Jervis 2001,
47 Ohl 2002). Hymenopteran ovaries are formed from multip le, elongated tube-like, functional units called ovarioles. Each ovariole contains a prog ressive series of deve loping oocytes which are individually enclosed in ep ithelial follicle s (Nation 2002). All Hymenoptera bear meroistic polytrophic ovarioles; meaning, ea ch developing oocyte possesses a cluster of associated nurse cells (meroistic). Nurse cells are located either within the same follicle as the oocyte or in an adjacent follicle (polytrophic) (Na tion 2002, Simiiczyjew 2002, Martins 2004). Ovarioles are subdivided into three basic stru ctural regions: the terminal filament, the germarium, and the vitellarium. The terminal filame nt, located at the proximal tip of the ovariole, acts as an attachment point that links the ova rioles to supporting connective tissue (Martins 2004). The germarium is the small region adjacent to the terminal filament where the process of oocyte production (ovigenesis) occurs. The remaining bulk of the ovariole consists of the vitellarium. Oocyte development, yolk uptake and vitelligenesis take place in this region (Nation 2002, Martins 2004). Maximum rate of ovigenesis influences th e lifetime reproductive output of synovigenic parasitoids (Rosenheim 2000, Eliopoulos 2003). Ovigenesis is a nutrient-limited process and adult wasps require nutrients for both metabolic maintenance and reproduction. As a result, egg maturation rate and reproductive output are often directly related to th e amount of nutritional resources acquired by the female (Jervis 2001, Ceballo 2004). Another ovarial characteristic wh ich also affects egg maturation rate and realized fecundity is egg load (Rosenheim 2000, Ceballo 2004). Egg lo ad is the maximum number of mature eggs that a female wasp can store in her ovaries. Egg load is the end result of a dynamic process involving input from egg produc tion and output from oviposition (Minkenberg 1992). Egg load influences behaviors such as searching effici ency and motivation to oviposit (Eliopoulos 2003).
48 Wasps have been shown to maintain relative ly constant egg loads by adjusting their egg maturation rates (Rivero-Lynch 1997, Jervis 2003). Op timal fecundity models predict that clutch size should correlate positively with egg load and the rate of egg production (Minkenberg 1992). Many studies that examine parasitoid re production tend to focus mainly on their reproductive output and behavior. As a consequence, associations with the insects internal anatomy are often overlooked. Despite L. bicolor s importance as a biol ogical control of Scapteriscus mole crickets, very little informati on was available on the anatomy of its reproductive system. The objective of this study was to characterize L. bicolor s ovary morphology and measure the variability of each tr ait. Knowledge of this beneficial parasitoids reproductive anatomy, combined with information regarding its potential fecundity, will improve our understanding of this in sects overall reproductive cap abilities and limitations. Materials and Methods Female wasps were collected from the Univer sity of Floridas Beef Research Unit (BRU) and immediately transporte d back to the lab where they were chilled in a refrigerator. Once the wasps were immobilized, the leng th of their right hind tibia wa s measured. Next, their abdomens were severed at the petiole, using small spring scissors. Ovaries were removed from the loose abdomens by grasping the last abdominal segment with a forceps and gen tly pulling it away from the rest of the segments; bri nging along with it th e wasps entire reproductive tract. Once removed, the reproductive organs were placed on a 60 mm Petri dish filled with ice cold 1% phosphate buffer saline (PBS) solution (pH 7.0). Th e ovaries were quickly separated from the rest of the wasps reproductive organs and rins ed in a new Petri dish with fresh PBS. After rinsing, the ovaries were put into a small vial filled with 30% gl ycerol in PBS and placed in the refrigerator to equilibr ate overnight. The following day the ov aries were moved to a watch-glass filled with neutral red staining so lution. After 2 min the ovaries were removed from the stain and
49 washed for 5 min in a 5 ml beaker of ice cold PBS. After washing, the ov arioles were carefully separated from one another and each ovary was detached from the common oviduct. Next, each ovary was temporarily slide-mounted in 30% glycerol solution. After slide-mounting, the ovari es were photographed using a stereomicroscope outfitted with the Auto-Montage digital-photography sy stem (Leica Microsytems, CA). Photographs were taken at 10 magnification. Graphics editing software (Photoshop 5.5) was used to remove noise and enhance the quality of the im ages. Auto-Montage image processing software was used to scale the images and make measur ements of the different ovarial components. Means, standard errors and ranges were calculated (Proc means) for tibia length, ovary length, ovariole length, mature egg number, matu re egg length, oocyte number, largest oocyte length and smallest oocyte length. To determ ine any relationships between each pairs of variables, correlation coefficients were comput ed using Statistical Analysis System (SAS) 9.1 (SAS Institute,Cary, NC). Results All wasps had two ovaries (right and left) a nd each ovary consisted of three ovarioles. No variation in the basic structure of the ovaries was observed. Various parts of the ovaries such as, the germarium, oocytes, nurse ce lls and mature eggs, stain diffe rently. For instance, no color could be detected within the germarium. Unde veloped oocytes stained weakly and therefore appeared pink. More intense staining of nurse cells caused them to look reddish. Mature eggs typically appeared pallid and showed lit tle indication of si gnificant staining. Female tibia length averaged 3.01 0.08 mm, with minimum and maximum values of 2.70 mm and 3.45 mm respectively (Table 5.1). Ovar ies had an average length of 9.68 0.37 mm with the smallest ovary measuring 7.70 mm a nd the largest ovary measuring 11.53 mm. Average ovariole length for the wasps was 9.23 0.38 mm. The shortest ovariole measured 7.21 mm and
50 the longest ovariole measured 11.10 mm. Averag e egg load was 7.60 0.63 eggs. The lowest number of eggs recorded from a single female was 4 and the most was 10. The average length for mature eggs was 1.63 0.03 mm. The smallest egg measured 1.50 mm and the largest egg measured 1.80 mm. Females possessed an av erage of 69.0 3.30 developing oocytes. The lowest number of oocytes counted was 54 and the highest number counted was 83. Correlation analysis determined that several of the traits were related. Tibia length corresponded positively with ovary length (r= 0.67, P< 0.04, N=10), ovari ole length (r=0.68, P< 0.03, N=10) and egg load (r=0.68, P< 0.03, N=10). Ov ary length closely co rresponded with egg load (r=0.76, P= 0.01, N=10). Ovariole length al so corresponded strongly with egg load (r=0.81, P< 0.004, N=10). Lastly, egg load corresponded nega tively with average e gg length (r=-0.70, P< 0.03, N=10). Discussion To date, information regarding the structural organization of Larra bicolor s internal reproductive organs is especially limited. It is im portant to be familiar with the basic anatomy of parasitoids because the physical characteristic s of their reproductive systems can be key determinants of their reproductive capabiliti es (Jervis 2001, Ohl 2002). Ovarial traits including ovariole number, numbers of mature eggs, and matu re egg size closely correlate with the habits of female parasitoids (Iwata 1955, Jervis 2001). In the case of some parasitoids such as ichneumonids (Hymenoptera) and tachinids (Diptera), fecundity strongly correlates with the number of ovarioles per ovary (Price 1997). However, the number of ovarioles per ovary does not vary between individuals of L. bicolor This makes it more difficult to estimate Larras potential fecundity because thei r fixed ovariole number cannot acc ount for observed variations in egg output (see Chapter 4).
51 My results show several significant correlati ons between variables. The first set of correlations suggest that the magnit ude of certain traits, such as the average length of ovaries and ovarioles and the number of mature eggs carried by females, are at least partially determined by the size of the wasp. This is understandable, cons idering that mature e gg size and largest ooctye size (data not shown) displaye d relatively minor variation (SEM= .03) and (SEM= .04) respectively. These correlations also help to expl ain the differences in the number of mature eggs carried by the females. We can assume that if ma ture eggs have a relati vely constant volume and abdomen size is proportional to tibia length, than large females will be able to hold more eggs because their abdomens have higher storage capacit ies then small females. Furthermore, females that are carrying more mature eggs will have increased opportunities to oviposit than females carrying fewer eggs. Another significant correlation indicates that egg load varies inversely to the size of the eggs. This result suggests that a parental investment trade-off exists within the context of available egg storage space. Therefore females must either produce fewer large eggs or a greater number of small eggs. Evolutionarily, both strate gies have potential to increase female fitness levels. Producing larger eggs incr eases the probability that the la rvae which hatch will be healthy and have better survival rates. Producing greater nu mbers of eggs increases the probability that at least some the resulting larvae w ill live to adulthood, despite lower survival rates. Currently it is uncertain what factors are important for determining egg size and number for L. bicolor A comprehensive analysis of r vs K strategy for pa rasitoids is beyond the scope of this thesis, but would be an interesting subject for future work.
52 Figure 5-1 L. bicolor ovary stained with neutral red and photographed at 10 original size using the Auto-montage imaging system attached to a stereomicroscope. Ovary was slidemounted in 30% glycerol. Table 5-1. Summary statistics for ov arial traits. Values represent x SEM, minimum and maximum for tibia length (N=10), Ovary length (N=10), ovariole length (N=60), mature egg number (N=10), egg length (N=76), and number of oocytes (N=10). Mean SEM Minimum Maximum Tibia Length (mm) 3.01 0.08 2.70 3.45 Ovary Length (mm) 9.68 0.37 7.70 11.53 Ovariole Length (mm) 9.23 0.38 7.21 11.10 Mature Eggs 7.60 0.63 4 10 Egg Length (mm) 1.63 0.03 1.50 1.80 Developing Oocytes 69.0 3.30 54 83
53 CHAPTER 6 SUPERPARSITISM AND DEVELOPMENTA L SUCCESS OF SUPERNUMERARY LARVAE Introduction Host selection has major implications for parasi toid fitness. Generally, it is advantageous for females to select the highest quality hosts for their offspring (Keasar 2006). Acceptance of hosts often depends upon the hosts physical a nd physiological character istics. One important characteristic is whether the host has been pr eviously parasitized (P lantegenest 2004). For many species of solitary wasps, including Larra bicolor only one larva is able to develop successfully per host. The presence of supernumerary larv ae leads to competi tion (Plantegenest 2004, Darrouzet 2007). Typically, older larvae out-com pete and eventually eliminate younger rivals (Plantegenest 2004). There are two types of superparasitism. Consp ecific-superparasitism occurs when a female oviposits on a host already parasitized by another female of the species. Self-superparasitism occurs if a female oviposits on a host which she had previously pa rasitized (Jaramillo 2007, Darrouzet 2007). Generally, parasito ids are reluctant to superparasitize. The factors involved in a parasitoids motivation to superp arasitize could be mediated by envi ronmental conditions such as low host availability or by the physiological conditi on of the female, such as a large egg load or nearing the end of her reproductive period (K easar 2006, Jaramillo 2006). Despite female hostdiscriminating abilities and i ndividual fitness disadvantages superparasitism is commonly observed in nature and in the laboratory (Plantegenest 2004, Jaramillo 2006). Superparasitism by female Larra spp. was first described for Larra analis F. by Charles Smith in 1935. Smith (1935) recorded th e fate of four mole crickets ( Neocurtilla hexadactyla ) carrying two eggs each. According to his description, two of the cric kets died before the larvae hatched from the eggs. For the other two cricke ts, both larvae developed to maturity, although
54 they were smaller than average (Smith 1935). L. bicolor females typically remove previous eggs when they encounter them. However, Scapteriscus mole crickets occasionally become superparasitized (Castner 1988a and pers. field ob servs.). The following is a brief description of the fate of Scapteriscus mole crickets and their L. bicolor supernumerary parasites. Materials and Methods Superparasitized mole crickets we re encountered during experiments on L. bicolor fecundity (see chapter 4). Mole crickets found to possess two or more eggs were placed individually into vials filled with moist clean sa nd. Vials were then pla ced in a Florida ReachIn environmental chamber (Walker 1993) (27oC, humidity 55%, 16/8 hr L/D). Crickets were fed FRM Cricket & Worm Feed (Flint River Mills. Bainbridge, GA) at least twice per week, and water was added to the substrate as needed. Cricket condition and parasite development were monitored by removing the crickets or peering th rough the walls of the vial. Efforts were made not to disturb the crickets and parasites frequently because this seemed to adversely effect their survival. Results and Discussion Superparasitized mole crickets appeared to behave normally. The number and general location of the eggs were recorded. Results show some variation in these characteristics. Ten crickets were found with two eggs positioned on opposite sides of the mole cricket. Nine crickets were found with two eggs positioned on the same side and adjacently (Fig 6-1A). One cricket was found with three eggs, two on one side and on e opposite (Fig 6-1B). Because the crickets were exposed to single females, these are all exam ples of self-superparasitism. However, it could not be determined whether the eggs were attach ed at the same time or the eggs resulted from separate attacks.
55 Figure 6-1 Photographs showing the different numbers and locations of L. bicolor eggs. A) two eggs adjacent B) three eggs. The larvae that hatched from each egg remained attached to the mole crickets in the same location as the egg (Fig 6-1A & B). Adjacent larvae made no attempts to distance themselves from one another. Despite each egg being laid with in 24 hrs, larvae displayed disparity in their developmental rates (Fig 6-2B). Contrary to earlier conjectu re regarding fighting between conspecific larvae (Castner 1988a), no indications of injury or death as a result of physical conflict were observed. The principal level of competition appears to be that of nutrient acquisition. Figure 6-2 Photographs showing th e development of supernumerary L. bicolor larvae. A) larvae opposite sides B) larvae same side.
56 Twenty instances of supernumerary larvae and their mole cricket hosts were monitored and recorded. For the 10 instances where mole cricke ts were found with two eggs on opposite sides, half of the time, only one larva per cricket was able to develop to maturity. For the rest, both larvae and their host crickets died. Larvae exhib ited a 25% survival ratio. Of the nine crickets found with two adjacent eggs, in four instances, one larva out of two managed to complete its development. The remaining five crickets and thei r parasites died. In this case larvae exhibited a mere 22.2% survival ratio. Only one larva develo ped fully on the lone cricket found with three eggs. The total percent survival for all the larvae was 24.4%. This is a much lower proportion compared with their normal 97.6 % surv ival ratio (Cabrera-Mireles 2002). These results indicate that a single mole cricket cannot succes sfully support the complete development of more than one L. bicolor larva. Furthermore, superparasitism does not contribute to increasing the death rate of mole crickets. One larva is sufficient to kill a mole cricket. More importantly, the drastic reduction in larval surv ival suggests that superparasitism lowers the reproductive fitness of females. By lowering la rval survival ratios, superparasitism should negatively affect future adult parasitoid populations.
57 LIST OF REFERENCES Adjei, M. B. (2000). Relative tolerance of warm seas on pasture grasses to mole crickets. Florida Cattleman and Livestock Journal 64 (9): 52. Arvalo-Rodrguez, H. A. (2003) Recognizing and selecting vari ous species of plants that can be used as nectar sources by adults of Larra bicolor F. (Hymenoptera: Sphecidae) in Florida M.S. dissertation University of Florida, Gainesville, FL. Arvalo-Rodrguez, H. A., Frank, J. H. (2005). Observations on the behavior of Larra bicolor (Hymenoptera: Sphecidae) in the fiel d, with emphasis on feeding behavior. Florida Entomol 88 : 146-151. Baggen, L. R., and Gurr, G. M. (1998). Influence of food on Copidosoma koehleri (Hymenoptera: Encyrtidae), and th e use of flowering plants as a habitat management tool to enhance biological cont rol of potato moth Phthorimaea operculella (Lepidoptera: Gelechiidae). Biol. Cont. 11 : 9-17. Bell, C. R., and Taylor, B. J. ( 1982). Florida wild flowers and ro adside plants. North Carolina: Laurel Hill Press. Bennett, F. D., and Pruett, C. J. (1991) Observations on copulat ion in Florida and on the behavior of male and female wasps of the genus Larra in Santa Cruz, Bolivia. Sphecos 21: 16-17. Bohart, R. M., and Menke, A. S. (1976). Sphecid wasps of th e world. A generic revision. California: University of California Press. Brandenburg, R. L., Xia, Y., and Schoeman, A. S. (2002). Tunnel architectures of three species of mole crickets (Orthoptera: Gryllotalpidae). Florida Entomol 85: 383-385. Cabrera-Mireles, H. (2002). Relationship between temper ature and the development of the ectoparasitoid Larra bicolor (Hymenoptera: Sphecidae) and the endoparasitoid Ormia depleta (Diptera: Tachinidae) Ph.D. dissertation. University of Florida, Gainesville, FL. Castner, J. L. (1984). Suitability of Scapteriscus spp. mole crickets as hosts of Larra bicolor (Hymenoptera: Sphecidae). Entomophaga 29: 323-329. Castner, J. L. (1988a). Evaluation of Larra bicolor as a biological control agent of mole crickets. Ph.D. dissertation. University of Florida, Gainesville, FL. Castner, J. L. (1988b). Biology of the mole cricket parasitoid Larra bicolor (Hymenoptera: Sphecidae), pp 131-139 In Gupta V. K. (ed) Advances in parasitic Hymenoptera research Brill Leiden.
58 Ceballo, F. A., and Walter, G. H. (2004). Life history paramete rs and biocontrol potential of the mealybug parasitoid Coccidoxenoides peregrinus (Timberlake) (Hymenoptera: Encyrtidae): asexuality, fecund ity and ovipositional patterns. Biol. Cont 29: 235-244. Chow, A. (2000). Influence of host distribution on foraging behaviour in the hyperparasitoid wasp, Dendrocerus carpenteri. Entomol Exp. Appl. 97: 57-66. Comins, H. N., Hassell, M. P., and May, R. M. (1992). The spatial dynami cs of host-parasitoid systems. J. Anim. Ecol 61: 735-748. Comins, H. N., and Hassell, M. P. (1996). Persistence of multispecies host-parasitoid spatially distributed models with local dispersal J. Theor. Biol 183: 19-28. Curnutt, J. L. (2000). Host-area specific climatic-m atching: Similarity breeds exotics. Biol. Conser 94: 341-351. Cushman, R. A. (1935). A study of the larvae of Larra analis Fabricius. Proc. Ent. Soc. Wash. 37: 82-85. Darrouzet, E., Bignon, L., and Chevier, C. (2007). Impact of mati ng status on egg-laying and superparasitism behavior in a parasitoid wasp. Entomol Exp. Appl 123: 279-285. Darrouzet-Nardi, A., Hoopes, M. F., Walker, J. D., and Briggs, C. J. (2006). Dispersal and foraging behaviour of Platygaster californica : hosts can't run, but they can hide. Ecol. Entomol 31: 298-306. Desouhant, E., Driessen, G., Lapchin, L., Wielaard, S., and Bernstein, C. (2003). Dispersal between host populations in field conditions : navigation rules in the parasitoid Venturia canescens Ecol. Entomol 28: 257-267. Desouhant, E., Driessen, G., Amat, I., and Bernstein, C (2005). Host and food searching in a parasitic wasp Venturia canescens : a trade-off between current and future reproduction? Anim. Behav. 70: 145-152. Eliopoulos, P. A., Harvey, J. A., Athanassiou, C. G., and Stathas, G. J. (2003). Effect of biotic and abiotic factors on reproductive pa rameters of the synovi genic endoparasitoid Venturia canescens. Physiol. Entomol 28: 268-275. Fadamiro, H. Y., and Heimpel, G. E. (2001). Effects of partial sugar deprivation on lifespan and carbohydrate mobilization in the parasitoid Macrocentrus grandii (Hymenoptera: Braconidae). Ann. Entomol. Soc. Am 94: 909-916. Frank, J. H. (1990). Mole crickets a nd other arthropod pests of turf and pastures. pp. 131-139. in Classical Biological Control in the Southern United States. Southern Coop. Series Bull 355: 1-viii 1-197.
59 Frank, J. H. (1994). Biological control of pest mole crickets, pp. 343-352. In Pest management in the subtropics: biological control a Flor ida perspective. Andover, U.K: Intercept. Frank, J. H., Parkman, J. P., and Bennett, F. D. (1995). Larra bicolor (Hymenoptera: Sphecidae) a biologi cal control agent for Scapteriscus mole crickets (Orthoptera: Gryllotalpidae), establishe d in northern Florida. Florida Entomol 78: 619-623. Frank, J. H., Fasulo, T. R., Short, D. E., and Weed, A. S. (2003). Mole Cricket Knowledgebase and Tutorials UF/IFAS. http ://www.molecrickets.ifas.ufl.edu/mcri0001.htm. (07/30/2004). Frank, J. H., and Parkman, J. P. (1999). Integrated pest manage ment of pest mole crickets with emphasis on the southeastern United States. Integ. Pest Man. Rev 4: 39-52. Gordon, D. R. (1998). Effects of invasive, non-i ndigenous plant species on ecosystem processes: Lessons from Florida. Ecol. Appl 8: 975-989. Hassell, M. P. (2000). Host-parasitoid population dynamics J. Anim. Ecol 69: 543-566. Heimpel, G. E., Mangel, M., and Rosenheim, J. A. (1998). Effects of time limitation and egg limitation on lifetime reproductive succe ss of a parasitoid in the field. Am. Nat 152: 273289. Henne, D. C., and Johnson, S. J. (2001). Seasonal distribu tion and parasitism of Scapteriscus spp. (Orthoptera: Gryllotalpidae ), in southern Louisiana. Florida Entomol 84: 209-212. Hudson, W. G., Frank, J. H., and Castner, J. L. (1988) Biological control of Scapteriscus spp. mole crickets (Orthoptera: Gr yllotalpidae) in Florida. Bull. Entomol. Soc. Am 34: 192-198. Jaramillo, J., Borgemeister, C., Setamou, M. (2006). Field superparasitism by Phymastichus coffea a parasitoid of a dult coffee berry borer, Hypothenemus hamperi. Entomol. Exp. Appl 119 : 231-237. Jervis, M. A., Heimpel, G.E., Ferns, P.N., Harvey, J.A., and Kidd, N. A. (2001). Life-history strategies in parasitoid wasps: a comparative analysis of `ovigeny'. J. Anim. Ecol 70: 442458. Jervis, M. A., Ferns, P. N., and Heimpel, G. E. (2003). Body size and the timing of egg production in parasitoid wasp s: a comparative analysis. Funct. Ecol. 17: 375-383. Keasar, T., Segoli, M., Barak, R., Steinberg, S., Giron, D., Strand, M., Bouskila, A., and Harari, A. (2006). Costs and consequences of superparasitism in the polyembryonic parasitoid Copidosoma koehleri (Hymenoptera: Encyrtidae). Ecol. Entomol 31: 277-283. King, B. H., and Ellison, J. H. (2006). Resource quality affects restlessness in the parasitoid wasp Nasonia vitripennis Entomol Exp. Appl 118: 71-76.
60 Lawrence, K. O. (1982). A linear pitfall trap for mole crickets and other soil arthropods. Florida Entomol 65: 376-377. Leather, S. R. (1988). Size, reproductive potential and f ecundity in insects: Things aren't as simple as they seem. Oikos 51: 386-389. Leibee, G. L., and Capinera, J. L. (1995). Pesticide resistance in Florida insects limits management options. Florida Entomol 78: 386-399. Leius, K. (1960). Attractiveness of different foods a nd flowers to the adults of hymenopterous parasites. Can. Entomol 92: 369-376. Leius, K. (1967a). Food sources and preferen ces of adults of a parasite Scambius boulianae (Hymenoptera: Ichneumonidae) and their consequences. Can. Entomol 99: 865-871. Lewis, W. J., Stapel, J. O., and Takasu, K. (1988). Understanding how parasitoids balance food and host needs: importanc e to biological control. Biol. Cont 11: 175-183. Lewis, W. J., van Lenteren, J. C., Phatak, S. C., and Tumlinson, J. H. (1997). A total system approach to sustainable pest management. Proc. Nat. Acad. Sci 94: 1243-1248. Martins, G. F., and Serrao, J.E. (2002). A comparative study of the spermatheca in bees (Hymenoptera; Apoidea). Sociobiology 40: 711. Matheny, E. L., Jr. (1981). Contrasting feeding habits of pest mole cricket species. J. Econ. Entomol. 74: 444-445. Mattiacci, L., Hutter, E., and Dorn, S. (1999). Host location of Hyssopus pallidus a larval parasitoid of the codling moth, Cydia pomonella Biol. Cont 15: 241-251. Menke, A. S. (1992). Mole cricket hunters of the genus Larra in the New World (Hymenoptera: Sphecidae; Larrinae). J. Hymen. Res 1: 175-234. Mills, N. J., and Kuhlmann, U. (2000).The relationship between egg load and fecundity among Trichogramma parasitoids. Ecol. Entomol 25: 315-324. Minkenberg, O. P. J. M., Tatar, M., and Rosenheim, J. A. (1992). Egg load as a major source of variability in insect fo raging and oviposition behavior. Oikos 65: 134-142. Nation, J. L. (2001). Insect physiology and bioc hemistry. Florida: CRC Press. National Climatic Data Center (2007) Daily temperatures for Florida. in Climatological Data, Florida 110: North Carolina. Nickle, D. A., and Frank, W. (1988). Pest mole crickets, Scapteriscus acletus (Orthoptera: Gryllotalpidae), reported established in Arizona. Florida Entomol 71: 90-91.
61 Ohl, M., and Linde, D. (2003). Ovaries, ovarioles, and oocyt es in apoid wasps, with special reference to cleptoparasitic species (Hymenoptera: Apoidea: 'Sphecidae'). J. Kans. Entomol. Soc 76: 147-159. Olson, D. M., Fadamiro, H., Lundgren, J. G., and Heimpel, G. E. (2000). Effects of sugar feeding on carbohydrate and lipid me tabolism in a parasitoid wasp. Physiol. Entomol 25: 1726. Parkman, J. P., Frank, J. H., Walker, T. J., and Schuster, D. J. (1996). Classical biological control of Scapteriscus spp. (Orthoptera: Gr yllotalpidae) in Florida. Environ. Entomol 25: 1415-1420. Plantegenest, M., Outreman, Y., Goubault, M., and Wajnberg, E. (2004). Parasitoids flip a coin before deciding to superparasitize. J. Anim. Ecol 73: 802-806. Pruett, C. J., and Bennett, F. D. (1991). Behavior of two species of Larra in Santa Cruz, Bolivia. Sphecos 21: 15-16. Rivero-Lynch, A. P., and Godfray, H. C. J. (1997). The dynamics of egg production, oviposition and resorption in a parasitoid wasp. Funct. Ecol 11: 184-188. Rogers, M. E., and Potter, D. A. (2004). Potential for sugar sprays and flowering plants to increase parasitism of white grubs (Col eoptera: Scarabaeidae) by tiphiid wasps (Hymenoptera: Tiphiidae) Environ. Entomol 33: 619-626. Sailer, R. I. (1985). Natural enemies. pp. 23-32. in Mole Crickets in Florida. Florida Agric. Exp. Bull. 846: (1984) 54 p. Smith, C. E. (1935). Larra analis Fabricius, a parasite of the mole cricket Gryllotalpa hexadactyla Perty. Proc. Entomol. Soc. Wash 37: 65-82. Steiner, A. L. (1984). Why can mole crickets stung by Larra wasps (Hymenoptera: Sphecidae; Larrinae) resume normal activities? The evol ution of temporary pa ralysis and permanent deactivation of the prey. J. Kans. Entomol. Soc 57: 152-154. Sweetman, H. L. (1936).The biological control of insects. New York: Comstock Publishing Co. Takasu, K., and Lewis, W. J. (1993). Hostand food-foraging of the parasitoid Microplitis croceipes : learning and physiol ogical state effects. Biol. Cont. 3: 70-74. Taylor, W. K. (1992).The Guide to Florida Wild Flow ers. Texas: Taylor Publishing Co. Tscharntke, T., and Brandl, R. (2004). Plant-insect interacti ons in fragmented landscapes. Annu. Rev. Entomol 49: 405-430.
62 Vattala, H. D., Wratten, S. D., Phillips, C. B., and Waeckers, F. L. (2006). The influence of flower morphology and nectar quality on the lo ngevity of a parasitoid biological control agent. Biol. Cont 39: 179-185. Volin, J. C., Lott, M. S., Muss, J. D., and Owen, D. (2004). Predicting rapid invasion of the Florida Everglades by old world climbing fern ( Lygodium microphyllum ) Divers. Distrib. 10: 439-446. Wckers, F. L. (2004). Assessing the suitabil ity of flowering herbs as parasitoid food sources: flower attractiveness and nectar accessibility. Biol. Cont 29: 307-314. Walker, T. J., Gaffney, J. J., Kidder, A. W., and Ziffer A. B. (1993). Florida Reach-Ins: Environmental chambers for entomological research. Am. Entomol 39: 177-182. Walker, T. J., and Nickle, D. A. (1981). Introduction and spread of pest mole crickets: Scapteriscus vicinus and S. acletus re-examined. Ann Entomol. Soc. Am 74 : 158-163. Wanner, H., Gu, H., Dorn, S. (2006). Nutritional value of floral nectar sources for flight in the parasitoid wasp, Cotesia glomerata Physiol. Entomol 31: 127-133. Wanner, H., Gu, H., Hattendorf, B., Guenther, D., and Dorn, S. (2006). Methodological insights: Using the stable isotope marker s uper(44)Ca to study dispersal and host-foraging activity in parasitoids. J. Appl. Ecol 43: 1031-1039. Williams, F. X. (1928) Studies in tropical wasps their hos ts and associates (with descriptions of new species) Hawaii. Sug. Plrs. Assoc. Exp. Stn. Entomol. Ser. Bull 19: 1-179. Wolcott, G. N. (1941). The establishment in Puerto Rico of Larra americana Saussure. J. Econ. Entomol 34: 53-56. Wunderlin, R. P. (1998). Guide to the vascular plants of Florida. Florida: University of Florida Press.
63 BIOGRAPHICAL SKETCH Scott Linus Portman was born on Febuary 21, 1972 in St. Louis, Missouri. He earned his Bachelor of Science in Biol ogy from Southeast Missouri State University in 1996. After graduation he began a job at Incy te Genomics Inc, St. Louis, MO, and worked on a project to map the human genome. After the project was term inated in 1999, he was supervising technician in the DiAntonio Lab at the Washington Univers ity School of Medicine, St. Louis, MO. Scott spent four years working with Aaron DiAntonio on the devel opment and functioning of nerve synapses at the Drosophila neuron-muscular junction. Scott alwa ys held a certain fascination for insects, so in 2004 he was accepted into the Department of Entomology and Nematologys graduate program at the University of Florida. After completing his master of science at the University of Florida, Scott will begin wo rking toward his Ph.D. at Pennsylvania State University, under the sponsorship of Dr. James Marden.