ELENA M. RHODES UNIVERSITY OF FLORIDA

Transcription

1 CONTROLLING TWOSPOTTED SPIDER MITE (Tetranychus urticae Koch) IN FLORIDA STRAWBERRIES WITH SINGLE AND COMBINATION TREATMENTS OF Phytoseiulus persimilis Athias-Henriot, Neoseiulus californicus (McGregor), AND ACRAMITE By ELENA M. RHODES A THESIS PRESENTED TO THE GRADUATE SCHOOL OF THE UNIVERSITY OF FLORIDA IN PARTIAL FULFILLMENT OF THE REQUIREMENTS FOR THE DEGREE OF MASTER OF SCIENCE UNIVERSITY OF FLORIDA 25

2 Copyright 25 by Elena M. Rhodes

3 This thesis is dedicated to the Lord and to the ministry of Chapel House.

4 ACKNOWLEDGMENTS I thank Dr. Oscar Liburd, my major professor, for allowing me to be his student and for all of his help on my projects and with the final write up of this thesis. I thank Drs. Robert Meagher and Donald Dickson for all of their hard work to get this in on time. I also thank Dr. William Crowe for sitting in for Dr. Dickson at my exit seminar. My field work would not have been possible without the staff and workers of the Citra Plant Science Research and Education Unit who took care of my strawberries in the field and did the bulk of the harvesting. Many thanks go to them. The staff and students of the Small Fruit and Vegetable IPM Laboratory also deserve my thanks. Of special note are Crystal Kelts who assisted with a lot of the field and microscope work in the 23/24 field season, Jeff White and Carolyn Mullin who spent long hours helping me count mites, and Alejandro Arevalo for statistics help. I thank Marinela Capana and Dr. Ramon Littell from IFAS statistics for helping me with data analysis. I thank my parents and all of my friends at Chapel House for support and for putting up with my whining and frustration. iv

5 TABLE OF CONTENTS page ACKNOWLEDGMENTS... iv LIST OF FIGURES... viii ABSTRACT...x CHAPTER 1 INTRODUCTION LITERATURE REVIEW...6 Twospotted Spider Mite...6 Biology...6 Management...7 Reduced-Risk Pesticides...9 Predatory Mites...9 Phytoseiulus persimilis...11 Neoseiulus californicus...12 Potential as Biological Control Agents and Use in Integrated Pest Management (IPM)...13 Hypothesis...15 Specific Objectives LABORATORY EXPERIMENTS...17 Methods...18 Colony...18 Experiment Experiment Data Analysis...19 Results...2 Experiment Experiment Discussion SINGLE TREATMENT EFFECTS ON TWOSPOTTED SPIDER MITE CONTROL...26 v

6 Methods...26 Sampling...27 Data Analysis...27 Results /24 Field Season /25 Field Season...3 Discussion TREATMENT COMBINATION EFFECTS ON TWOSPOTTED SPIDER MITE AND PREDATORY MITE SPECIES...46 Methods...47 Experiment 1 (23/24 Field Season)...47 Sampling...47 Data analysis...47 Experiment 2 (23/24 Field Season)...48 Experiment 3 (24/25 Field Season)...49 Sampling...49 Data analysis...49 Results...5 Experiment Experiment Experiment Discussion CONCLUSIONS...68 The Cost of Control...68 Future Directions...69 APPENDIX BEHAVIORAL STUDIES...71 Protocol for Laboratory Bioassay...71 Preliminary Experiments...72 Methods...72 Results...72 Egg Consumption Over Time by Predatory Mites...73 Methods...73 Results...73 Adult Consumption Over Time by Predatory Mites...74 Methods...74 Results...74 Discussion...75 LIST OF REFERENCES...79 vi

7 BIOGRAPHICAL SKETCH...85 vii

8 LIST OF FIGURES Figure...page 2-1 Twospotted spider mite and the damage it causes Phytoseiulus persimilis and eggs Neoseiulus californicus and eggs Twospotted spider mite colony Cage construction Greenhouse experimental setup Weekly average TSSM per leaflet in each treatment for Experiment Weekly average TSSM per leaflet in each treatment for Experiment Field experiment setup Weighing the harvest Average TSSM per leaflet for 5 periods of the 23/24 season Weekly average number of TSSM per leaflet in each treatment during the 23/24 season Weekly Average TSSM and predatory mite populations in each treatment during the 23/24 season Average strawberry yield from each treatment for the 23/24 season Average TSSM per leaflet for 5 periods during the 24/25 season Average number of TSSM per leaflet in each treatment for each week in the 24/25 season Weekly average TSSM and predatory mite populations in each treatment during the 24/25 season Treatment layout for 23/24 field season...56 viii

9 5-2 Treatment layout for 24/25 field season Average TSSM per leaflet for 5 periods of the 23/24 season Weekly average number of TSSM per leaflet in each treatment during the 23/24 season Weekly average TSSM and predatory mite populations in each treatment during the 23/24 season Weekly average TSSM and predatory mite populations in each treatment during the 23/24 season Weekly average TSSM and N. californicus populations each week in the Acramite/N. californicus treatment during the 23/24 season Average TSSM per leaflet for 5 periods of the 24/25 season Weekly average number of TSSM per leaflet in each treatment during the 24/25 season Weekly average TSSM and predatory mite populations in each treatment during the 24/25 season Weekly average TSSM and predatory mite populations in each treatment during the 24/25 season...66 A-1 Experimental arena...76 A-2 Experimental setup...77 A-3 Number of eggs laid in each m arena...77 A-4 Average percent reduction of TSSM adults in each treatment...78 A-5 Cumulative average percent reduction of TSSM adults in each treatment...78 ix

10 Abstract of Thesis Presented to the Graduate School of the University of Florida in Partial Fulfillment of the Requirements for the Degree of Master of Science CONTROLLING TWOSPOTTED SPIDER MITE (Tetranychus urticae Koch) IN FLORIDA STRAWBERRIES WITH SINGLE AND COMBINATION TREATMENTS OF Phytoseiulus persimilis Athias-Henriot, Neoseiulus californicus (McGregor), AND ACRAMITE By Elena M. Rhodes August 25 Chair: Oscar Liburd Major Department: Entomology and Nematology Laboratory and field experiments were conducted from 23 to 25 to determine the effectiveness of single and combination treatments of 2 predatory mite species, Phytoseiulus persimilis Athias-Henriot and Neoseiulus californicus (McGregor), and a reduced risk miticide, Acramite 5WP (bifenazate), for control of twospotted spider mite (TSSM) (Tetranychus urticae Koch) in Florida strawberry fields. In one set of laboratory tests, 15 mite-free strawberry plants were selected randomly and infested with known numbers of TSSM. After 1-2 weeks, 1 predatory mites from each species were released onto each plant. Twospotted spider mite populations were recorded before release of predatory mites and once a week for 4 weeks after release of predatory mites. Both species significantly reduced TSSM numbers to below those found in the control. However, TSSM numbers on the P. persimilis plants increased at week four. x

11 In the second set of laboratory tests, two other treatments, application of Acramite at the recommended rate and a combination release of P. persimilis/n. californicus, were added to the experiment. Both significantly reduced TSSM numbers below levels found in the control. No differences were recorded between predatory mite species except at week four when there was a significant increase in numbers of TSSM in the P. persimilis treatment. Field studies employed two setups: one looking at only single treatment applications of P. persimilis, N. californicus, and Acramite and the other adding combination treatments of P. persimilis/n. californicus, Acramite/N. californicus, and Acramite/P. persimilis. Both N. californicus and P. persimilis significantly reduced populations of TSSM below numbers recorded in the control plots. In 23/24 P. persimilis took longer than N. californicus to bring the TSSM population under control. Acramite was very effective in reducing TSSM populations below 1 mites per leaflet in 23/24 but not in 24/25, possibly due to late application. Among the combination treatments, the P. persimilis/n. californicus treatment significantly reduced TSSM numbers to below levels found in the control, but was not as effective as N. californicus alone in 23/24. Also, the Acramite/N. californicus, and Acramite/P. persimilis combination treatments significantly reduced TSSM populations to below the control. These findings indicate that N. californicus releases, properly timed Acramite applications, and combinations of Acramite applications and releases of either predatory mite species are promising options for TSSM control in strawberry for growers in north Florida and other areas of the southeast. xi

12 CHAPTER 1 INTRODUCTION Florida produces 1% of the domestically grown winter strawberry crop and ranks second in the United States behind California in terms of the quantity of strawberries produced (Mossler and Nesheim, 22). During 24, the total value of strawberries produced in Florida was approximately $178 million (Florida Agricultural Statistics Service, 24). Approximately 95% of Florida s strawberries are grown in Manatee and Hillsborough County with the remaining being grown in Alachua, Miami-Dade, and several other counties (Mossler and Nesheim, 22). In Florida, strawberries are planted as an annual crop using a raised bed system. The beds are fumigated with a combination methyl bromide + chloropicrin 2 weeks before planting and immediately covered with plastic mulch (Mossler and Nesheim, 22). Transplants are planted in late September through early November. The common strawberry varieties that are grown include Camerosa, Florida Festival, Sweet Charlie, Oso Grande, Rosa Linda, and Selva. The first five are short-day or June bearing; they need temperatures below 6ºF and/or photoperiods under 14 h to initiate flower production (Rieger, 25). Selva is an ever-bearing day-neutral; photoperiod has no effect on flowering (Rieger, 25). Overhead irrigation is used for the first 3 weeks after transplanting to establish the strawberry plants. After this period, drip irrigation is used. Fertilizer is usually applied through the drip irrigation system, a process often called fertigation (Hochmuth and Albregts, 1994). The average harvest period in Florida runs 1

13 2 from late November to early April. Harvesting occurs every 3 to 5 days. The harvest period ends when California strawberries begin to dominate the market. Strawberries are susceptible to many pests and soil borne pathogens. Bed fumigation is used to manage plant parasitic nematodes and many weed species. However, herbicides are used to control weeds in between rows. Most can only be applied twice in a season and cannot be used once harvesting begins because the postharvest interval (PHI) is too long (Mossler and Nesheim, 23). Removal of weeds by cultivation is often necessary (Stall, 23). Birds can also be a problem in some years (Mossler and Nesheim, 23). Most of the foliar diseases affecting Florida strawberries are caused by fungi. Anthracnose rots (Colletotrichum acutatum Simmonds, C. fragariae Brooks, and C. gloeosporioides (Penz.) Penz. and Sacc.), gray mold (Botrytis cinerea Pers.), phytophthora crown rot (Phytophthora cactorum (Lebert et Cohn) Schröter and P. citricola Sawada), and powdery mildew (Sphaerotheca macularis (Wallr.:Fr.) Lind.) are the main fungal diseases affecting strawberries in Florida, although others can also occur (Mossler and Nesheim, 23; Roberts, 23). The principle fungicides used are Captan and Thiram and a rotation of the two is the backbone of all strawberry growers fungus control program (Mossler and Nesheim, 23). Angular leaf spot, Xanthomonas fragariae Kennedy and King, is the only major disease not caused by a fungus. This bacterium causes an infection on plants that flourishes under cold, wet conditions (White and Liburd, 25; Mossler and Nesheim, 23). Insect pests of Florida strawberries include several Lepidoptera species namely: fall armyworm, Spodoptera frugiperda (J. E. Smith), southern armyworm, S. eridania

14 3 Cramer, and budworm, Helicoverpa zea (Boddie), sap beetles, flower thrips, and aphids. The Lepidoptera species are early season pests. If present during the establishment period, these insects grow to mature larvae before any treatment can be applied. Since Bacillus thuringiensis (Bt) (Dipel ) is ineffective against mature larvae, growers must resort to methomyl (Lannate ), a compound highly toxic to beneficial organisms including predatory mites, for control. However, many growers now apply spinosad (SpinTor ), which is much less toxic to predatory mites and other beneficials (Mossler and Nesheim, 23). Many species of sap beetles (family Nitidulidae) feed on strawberries (Mossler and Nesheim, 23; Mossler and Nesheim, 22). They prefer rotting fruit, but they will sometimes lay their eggs in fresh fruit. Sap beetles are only a problem in fields where it is not possible to remove all fresh and rotting fruit from the fields (Mossler and Nesheim, 23). Western flower thrips (Frankliniella occidentalis (Pergande), and strawberry and melon aphids {Chaetosiphon fragaefolli (Cockerell) and Aphis gossypii Glover} are sporadic pests often controlled by natural enemies if toxic broad-spectrum insecticides are not used continually (Mossler and Nesheim, 23). Insecticides are used to control these insects if outbreaks occur. In Florida, twospotted spider mite (TSSM) Tetranychus urticae Koch is the key arthropod pest effecting strawberries. High densities of TSSM significantly reduce photosynthesis, transpiration, productivity, and vegetative growth (Sances et al., 1982). Twospotted spider mite feeding causes injury to chlorophyll containing mesophyll cells within the leaf tissue and increases stomatal closure, which results in a decrease of photosynthetic capacities of infested leaves (Sances et al., 1982). Electron micrograph pictures of tissues heavily injured by spider mites show misshapen cells that contain

15 4 homogeneous protoplasts with only vestiges of necrotic chloroplasts visible within them (Kielkiewicz, 1985). Twospotted spider mite populations can build up to damaging levels very quickly. Development from egg to mature adult takes about 19 days and females can lay up to 1 eggs (Mitchell, 1973). This high fecundity also allows TSSM to quickly become resistant to acaracides traditionally used to control them (Trumble and Morse, 1993). Traditional control strategies for TSSM have required several applications of key pesticides (= acaricides) during the strawberry production season. In many cases, miticides have been applied as a preventative measure or on a calendar basis, resulting in high control costs of up to $989 per hectare (Prevatt, 1991). This has resulted in a significant reduction in efficacy due to development of acaricide resistance in the TSSM population (Trumble and Morse, 1993). In addition, widespread concerns over food safety, human health, and the environment have led the U. S. Environmental Protection Agency (EPA) to announce cancellations, restrictions, and tolerance reassessments on many major pesticides available for strawberry pest management. Control of TSSM with releases of Phytoseiulus persimilis Athias-Henriot has been fairly successful in south-central Florida (Decou, 1994). However, the introduction of P. persimilis has not sufficiently suppressed high populations of TSSM in more northern areas of the state. As a result, growers have relied on a conventional miticide program. There is also evidence that P. persimilis may be less tolerant to cooler temperatures that northern Florida growers experience during the winter months (White and Liburd, 25).

16 5 Another predatory mite that has been released in Florida to control TSSM in strawberries is Neoseiulus californicus (McGregor) (Acari: Phtoseiidae). Neoseiulus californicus has traits of both a type II selective predator and a Type III generalist predator (Croft et al., 1998). It has also been reported to be more cold tolerant than P. persimilis (White, 23). These characteristics may allow N. californicus to provide better control of TSSM in north Florida where the more specialized P. persimilis is ineffective. The objectives of this study were 1) to determine if N. californicus can provide effective control of TSSM in north Florida strawberry fields and 2) to compare the effectiveness of N. californicus with P. persimilis as well as a reduced-risk miticide for control of twospotted spider mite.

17 CHAPTER 2 LITERATURE REVIEW Twospotted Spider Mite Early season infestations of twospotted spider mites (TSSM) cause reductions in photosynthesis and transpiration at a much lower population level than the population level that causes the same level of injury later in the season (Sances et al., 1981). The high levels of stress caused by large populations of TSSM decrease the quality and quantity of mature fruit and can lead to a reduction in flower development and vegetative growth (Sances et al., 1982). Biology The TSSM life cycle progresses through five stages: egg, six-legged larvae, protonymph, deutonymph, and adult. Each of the three intermediate stages feed and grow for only a short time before entering a quiescent state (Mitchell, 1973). Development from egg to mature adult takes about 19 days (Mitchell, 1973), although this time can be as short as 5 days (Krantz, 1978). Optimal conditions for development are high temperatures (up to 38 C) and low humidity (Krantz, 1978). Satoh et al., (2) noted that male TSSM use both precopulatory and postcopulatory guarding to increase their paternity. Twospotted spider mite adults are oval and.5 mm in length. They are usually light greenish-yellow in color with two large, dark spots on their abdomens. However, their coloration may include brown, red, orange, and darker green forms (Figure 2-1). Although T. urticae is made up of two distinct lineages, both the red and green forms are 6

18 7 found in both lineages, which indicates that they are the same species (Hinomoto et al., 21). Navajas et al., (2) notes that T. urticae is a relatively homogeneous species, although reproductive incompatibility does occur between conspecific populations coming from different host plants or locations. The eggs, which are spherical and clear to tan in color, are usually laid on the underside of leaves (Figure 2-1). They are approximately.2 mm in diameter. Twospotted spider mites usually infest the undersides of leaves. They spin a fine web on the leaves of strawberries and other host plants (Mossler and Nesheim, 22). This webbing provides protection from predators. It may also help to maintain favorable environmental conditions for mite development on the leaf surface (Krantz, 1978). Twospotted spider mites collect in large numbers on leaf edges as a response to overcrowding or poor host plant condition and are dispersed by wind (Boykin and Campbell, 1984). Even at low infestation levels, TSSM can be dispersed by winds of only 8 km/h (Boykin and Campbell, 1984). Twospotted spider mites can also disperse by walking (ambulatory dispersal), but they cannot cover great distances in this manner. Management Three cultural control practices that are important in the management of TSSM in strawberries are: 1) close monitoring of transplants, 2) sanitation, and 3) irrigation techniques. Planting transplants that are as mite-free as possible is extremely important because chemical sprays are washed off of the plant by the overhead irrigation that is used to establish strawberry transplants. In addition, predatory mites can also be washed off the plants by overhead mists. Sanitation practices such as removing clipped runners and other debris from the field eliminates potential reservoirs of TSSM.

19 8 Managing soil moisture level is also critical in regulating insect pests and diseases. Recent investigation by White and Liburd (25) found that low soil moisture promoted TSSM development in the laboratory. Similarly, in field studies TSSM numbers were significantly higher in low soil moisture treatments. Further investigation showed that excessive irrigation involving drip and overhead was found to increase the incidence of angular leaf spot, Xanthomonas fragaria disease in strawberries (White and Liburd, 25). There has been no economic threshold established for TSSM in north-central Florida. Twospotted spider mite is usually monitored by collecting a known number of leaf samples and counting the number of mites and eggs present on the leaves (White, 23). In south-central Florida, a control action (release of predatory mites or use of a miticide) is taken when 5% of the sampled leaves are infested with TSSM (Vrie and Price, 1994). Abamectin (Agri-Mek ), a compound derived from a soil bacterium, and hexythiazox (Savey ), a non-systemic miticide with ovicidal, larvicidal, and nymphicidal activity, and (more recently) bifenazate (Acramite 5WP ) are the predominant miticides used on Florida strawberries. Bifenthrin (Brigade ), a general insecticide, and fenbutatin-oxide (Vendex ), an organotin compound are used to a lesser extent (Mossler and Nesheim, 22; Mossler and Nesheim, 23). Their high fecundity and short life cycle allow TSSM to quickly become resistant to miticides. Using greenhouse and laboratory experiments, Price et al., (22) found that TSSM in Florida strawberry fields show a ten-fold resistance to abamectin when

20 9 compared with spider mites in a laboratory colony taken from the field two years prior to the experiments. Reduced-Risk Pesticides Recently, several new classes of pesticides have been introduced for use in small fruit crops, largely in response to the potential loss of and/or restrictions on older pesticides due to the 1996 Food Quality and Protection Act (FQPA) regulations. Applications of reduced-risk miticides may hold potential for control of TSSM. To be classified as reduced-risk, a pesticide must have at least one of the following characteristics: it must pose a low risk to human health, have low toxicity to non-target organisms, have a low potential to contaminate the environment, and/or enhance the use and reliability of integrated pest management (IPM) (Price, 22b). Acramite 5WP is one such compound that has shown promising results in strawberries. This miticide has a low toxicity toward beneficials and carries a Caution precautionary statement. Acramite can only be applied twice in a season at.85 to kg product per hectare with applications at least 21 days apart (Mossler and Nesheim, 23; Price, 22a). In laboratory studies using leaf disks, White (23) recorded a higher rate of TSSM mortality using Acramite 5WP compared with the conventional miticide Vendex. Predatory Mites Many species of phytoseiid mites are important biological control agents and form an integral part of pest management programs in strawberries and other crops (McMurtry and Croft, 1997). Generally, phytoseids are larger than TSSM, pear shaped, and have

21 1 longer legs. They range in color from pale to reddish depending on species. Phytoseiid eggs are larger than TSSM eggs and elliptical in shape (Henn et al., 1995). Phytoseiid mites are attracted to host volatiles released by plants when they are attacked by spider mites (Dicke and Sabelis, 1988). Margolies et al., (1997) found that there is a genetic component in predator response to herbivore-induced plant volatiles by testing selected and unselected phytoseiid mites in a Y-tube olfactometer. Predatory mites use volatiles produced by adult prey (possibly alarm pheromones) to avoid patches with conspecifics. This prevents the formation of predator aggregations on prey patches (Janssen et al., 1997). Phytoseiid mites can be placed into four categories based on their food habits that include: Type I specialist predatory mites that feed exclusively on Tetranychus mites, Type II specialists that consume Tetranychus mites and species in other genera that produce webbing, Type III generalists that feed on species in many different tetranychid genera and will also consume small insects and even pollen, and Type IV generalists that feed on mites but prefer pollen (McMurtry and Croft, 1997). Larval feeding types (nonfeeding, facultative-feeding, or obligatory-feeding) are not associated with the degree of prey specialization (Schausberger and Croft, 1999). Schausberger and Croft (2) examined whether specialist and generalist phytoseiid mites differ in aggressiveness and prey choice in intraguild predation. They found that aggressiveness in intraguild predation, species recognition, and preferential consumption of heterospecifics when given a choice is common in generalist but not in specialist phytoseiid mites. Like TSSM, phytoseiid mites disperse using both ambulatory and aerial means. Some phytoseiids have behavioral control of take-off but the aerial dispersal itself is

22 11 mostly passive and, as far as is known, they have no control of landing (Jung and Croft, 21a). Auger et al., (1999) showed that food deprivation and high temperature increase ambulatory dispersal in N. californicus. Generally, selective predatory mites tend to have higher rates of dispersal then generalists (Jung and Croft, 21b). Phytoseiulus persimilis Phytoseiulus persimilis is a Type I specialist that feeds exclusively on tetranychid mites (Figure 2-2). Adult female P. persimilis prefer TSSM eggs to larvae, but will feed on all life stages (Blackwood et al., 21). Walzer and Schausberger, (1999a) found that adult female P. persimilis can survive for a limited period by cannibalism and interspecific predation, but they cannot reproduce on this diet of conspecifics and heterospecifics. Juveniles, however, can reach adulthood when provided either conspecifics or heterospecifics. They appear to lack the ability to distinguish between conspecifics and heterospecifics and will feed equally on both if given a choice (Walzer and Schausberger, 1999b). Phytoseiulus persimilis has a short developmental time, a nonfeeding larval stage, and a high rate of fecundity (McMurtry and Croft, 1997). Friese and Gilstrap (1982) found that P. persimilis has a fixed potential fecundity, and that within this potential, as the ovipositional rate increases the ovipostion period decreases. In the presence of excess prey, P. persimilis has a shorter developmental time for active stages, a greater reproductive longevity, and kills more prey and a greater number of prey/h both as an immature and as a reproductive adult than N. californicus (Gilstrap and Friese, 1985). This allows P. persimilis populations to increase very quickly. The population increases (booms) when spider mites are abundant and crashes (busts) when the spider mite populations have been decimated. P. persimilis often disperse in patches

23 12 and overexploit their prey before emigrating to a new area. These boom-bust cycles often make several releases necessary to effectively manage TSSM. Neoseiulus californicus Neoseiulus californicus has traits of both a Type II specialist and a Type III generalist. They prefer spider mites as food but can subsist on other sources of food such as thrips and pollen when mite populations are low (Figure 2-3). They will also prey upon other predatory mite species (Gerson et al., 23). Palevsky et al., (1999) observed that in adult females, egg predation on heterospecific eggs is significantly higher than cannibalism of conspecific eggs implying that N. californicus can distinguish its eggs from those of other species. When given a choice, N. californicus appear to be able to distinguish between conspecifics and heterospecifics and prefer the latter (Walzer and Schausberger, 1999b). Neoseiulus californicus females can sustain oviposition when preying upon other phytoseiid mites, but not when preying upon conspecifics. Juveniles can reach adulthood preying upon both conspecifics and heterospecifics (Walzer and Schausberger, 1999a). Neoseiulus californicus larvae are facultative feeders (Schausberger and Croft, 1999). The presence of spider mite prey increases larval walking and intraspecific interactions of N. californicus larvae (Palevsky et al., 1999). Overall, N. californicus has a longer development time and a lower dispersal rate than P. persimilis (Jung and Croft, 21b). Like P. persimilis, N. californicus has a fixed potential fecundity, within which, as the ovipositional rate increases the oviposition period decreases (Friese and Gilstrap, 1982). Walzer et al. (21) found that N. californicus reared on detached bean leaf arenas with diminishing TSSM prey survived three to five times longer after prey depletion than P. persimilis whether they were reared alone or in combination. When reared together, N.

24 13 californicus eventually displaced P. persimilis. This suggests that N. californicus can persist in a field for a longer period of time when TSSM populations are low and may give better season-long control when introduced into the field earlier in the season. Potential as Biological Control Agents and Use in Integrated Pest Management (IPM) Both P. persimilis and N. californicus have been used to control TSSM on strawberry and other crops throughout the world. Phytoseiulus persimilis has been successfully used to control TSSM on strawberries grown in greenhouses in Korea (Kim, 21), in walk-in plastic tunnels in England (Cross, 1984; Port and Scopes, 1981), and in the field in many parts of the world (Charles, 1988; Cross et al., 1996; Decou, 1994; Easterbrook, 1992; Oatman et al., 1977a; Oatman et al., 1976; Oatman et al., 1968; Oatman et al., 1967; Trumble and Morse, 1993; Waite, 1988). A combination of releases of P. persimilis and N. fallacis (Garman), a type II specialist, controlled TSSM in Taiwan (Lee and Lo, 1989). Phytoseiulus persimilis has also been used to control TSSM on dwarf hops (Barber et al., 23), on raspberry in New Zealand (Charles et al., 1985), and on Rhubarb in California (Oatman, 197). Neoseiulus californicus has also been successfully used to control TSSM on strawberry, although to a lesser extent than P. persimilis. Garcia-Mari and Gonzalez- Zamora (1999) noted that N. californicus is the main predator controlling TSSM on strawberries in Valencia, Spain. Neoseiulus californicus has also shown potential for use in the UK; controlling TSSM on potted strawberry plants in a gauze-sided glasshouse (Easterbrook et al., 21). Greco et al., (1999) reported that N. californicus is a promising established natural enemy for controlling TSSM on strawberry in greenhouses in Argentina. In the U.S., N. californicus has also been used to successfully control TSSM

25 14 on strawberry in southern California (Oatman et al., 1977b). Liburd et al., (23) also recorded good results from preliminary studies involving releases of N. californicus in strawberry fields in north Florida. Neoseiulus californicus has also been used to control TSSM on other crops, for example on dwarf hops (Barber et al., 23). A combination of one or two miticide sprays and the release of predatory mites may provide better control than either alone. Trumble and Morse (1993) used weekly yields and control costs to calculate the economic benefit of controlling TSSM with several chemicals, P. persimilis releases, and combinations of both. The best returns were generated by abamectin in combination with releases of P. persimilis. Their data also indicated that neither fenbutatin-oxide nor hexythiazox are compatible with P. persimilis. Easterbrook (1992) noted that to prevent an early build up of TSSM it may be necessary to reduce TSSM numbers with an application of a miticide early in the season before a later release of P. persimilis. The use of the Pest in First (PIF) technique may enhance the effectiveness of P. persimilis releases. In PIF, a crop is artificially seeded with a pest so that there will be sufficient prey present for the predator to establish and suppress developing populations when it is introduced some time later. This technique has been used in strawberry fields in Queensland, Australia for six seasons (1995-2). The predatory mites provided season-long control for a cost roughly equivalent to one spray of abamectin (Waite, 22). Chemical control in these strawberry fields requires at least two sprays of abamectin. Phytoseiulus persimilis is used to control TSSM on only 3% of grower fields in south Florida. There are two big impediments to predatory mite adoption. Many growers

26 15 desire clean plants (no mites at all). The second barrier is that the liquid formulation of Captan, one of the main fungicides used on strawberries, is highly toxic to predatory mites (Mossler and Nesheim, 23). Hypothesis Our hypothesis is that N. californicus will provide better control of TSSM in strawberries compared with P. persimilis. Furthermore, N. californicus and the reducedrisk miticide, Acramite 5WP will significantly reduce populations of TSSM in strawberries below untreated (control) plots. Specific Objectives The objectives of this thesis are as follows: 1. To conduct controlled laboratory experiments comparing the effectiveness of the predatory mites P. persimilis and N. californicus as well as a combination of the two and Acramite for control of TSSM. 2. To conduct field experiments to compare and evaluate the effectiveness of P. persimilis and N. californicus for control of TSSM and to compare predatory mite releases with periodic applications of the reduced-risk miticide, Acramite for control of TSSM. 3. To study competition between the two predatory mite species as well as their interaction with Acramite and the effects of these treatment combinations on TSSM control. 4. To conduct behavioral studies of P. persimilis and N. californicus to determine the rate at which they consume both TSSM adults and eggs.

27 16 A B Figure 2-1. Twospotted spider mite and the damage it causes. A) twospotted spider mite and eggs, B) strawberry plants damaged by twospotted spider mites A B Figure 2-2. Phytoseiulus persimilis and eggs. A) P. persimilis and B) its eggs (larger, ovoid eggs) shown with TSSM eggs (smaller, spherical eggs) for comparison. A B Figure 2-3. Neoseiulus californicus and eggs. A) N. californicus adult female and B) eggs.

28 CHAPTER 3 LABORATORY EXPERIMENTS In the presence of excess prey, Phytoseiulus persimilis has a shorter developmental time for active stages, a greater reproductive longevity, and kills more prey and a greater number of prey/h both as an immature and as a reproductive adult than Neoseiulus californicus (Gilstrap and Friese, 1985). Both predatory mite species have been found to effectively control TSSM populations on strawberries and other crops in various parts of the world. Phytoseiulus persimilis does not perform well in north Florida because of the colder temperatures found there (White, 23). The purpose of these experiments was to compare the efficacy of both species of predatory mite on TSSM control under controlled greenhouse conditions. Under such conditions, P. persimilis would be expected to perform as well or better than N. californicus because temperature effects are not a factor. In the first experiment, an untreated control was compared with treatments where each individual predator was released. In the second experiment, releases of each individual species were compared to two other treatments: Acramite applied at the recommended rate and a combination release of P. persimilis and N. californicus. Acramite is known to effectively control TSSM populations (White, 24). Lee and Lo (1989) found that a combination of P. persimilis and N. fallacies (Garman) released weekly or biweekly in November effectively controlled TSSM on strawberry in Taiwan for the season. The purpose of adding these two treatments was to compare the efficacy of N. californicus and P. persimilis releases on TSSM control to Acramite application 17

29 18 and to determine whether using a combination of the two species is an effective control strategy. Methods Colony A TSSM colony reared on strawberries was maintained in the laboratory to ensure that only TSSM predisposed to strawberries were used in the experiments (Fig. 3-1). The colony consisted of mite-infested strawberry plants that were screened periodically for the presence of TSSM. The colony was kept under 14:1 photoperiod at a temperature of ~27 C with 65% relative humidity. Plants were watered twice weekly. Experiment 1 Fifteen mite-free strawberry plants var. Festival were placed into previously constructed mite-free cages. Each plant was placed into an individual cage. Cages were constructed of nylon fabric. Velcro was placed on three sides (Figure 3-2A). Each cage was attached to a pot using a pull cord sown into the bottom (Figure 3-2B). Cages were used to keep both TSSM and predatory mites from dispersing between plants. Ten TSSM were released onto each plant and allowed to multiply for one to two weeks (This varied depending on when the predatory mite shipment arrived). Prior to the release of predatory mites, a leaflet was collected from each plant. The number of TSSM motiles and eggs on each leaflet was counted and the average number per leaflet calculated. Experimental design was a completely randomized block with 3 treatments. Each treatment was replicated 5 times. Treatments included: 1) 1 P. persimilis released per infested plant, 2) 1 N. californicus released per infested plant, and 3) untreated (control) plants (Fig. 3-3).

30 19 Each week the population of predators and TSSM was sampled by taking one leaflet from each plant (5 leaflets from each individual treatment) and counting the numbers of TSSM as well as predatory mites (motiles and eggs). Samples were recorded for 4 weeks. This experiment was repeated three times: once in Feb./Mar. 24, again in Dec. 24/Jan. 25, and finally in Mar./Apr. 25. Experiment 2 The protocol for Experiment 2 resembles Experiment 1 except that 5 treatments involving 25 mite-free strawberry plants were tested. Treatments included: 1) 1 P. persimilis released per infested plant, 2) 1 N. californicus released per infested plant, 3) 5 P. persimilis and 5 N. californicus released per infested plant, 4) Acramite sprayed onto each infested plant at the recommended rate, and 5) untreated (control) plants (Fig. 3-3). Experimental design was a completely randomized block with 5 treatments. Each treatment was replicated 5 times. Each week the population of predators and TSSM was sampled by taking one leaflet from each plant (5 leaflets from each individual treatment) and counting the numbers of TSSM as well as predatory mites (motiles and eggs). Samples were taken for 4 weeks. This experiment was repeated twice: once in Dec. 24/Jan. 25, and again in Mar./Apr. 25. Data Analysis Data were subjected to statistical analysis using the SAS program (SAS Institute, 22). TSSM motile and egg data from both experiments were log transformed. Average TSSM per leaflet was compared each week across treatments using an ANOVA and

31 2 means were separated using a LSD test. Predatory mite data from Experiment 2 were compared using a Student s t-test or a Satterthwaite t-test depending on whether or not the variances were equal. Results Experiment 1 There were no significant differences in TSSM motile and egg populations between treatments when the initial sample was taken (week ) (motiles: F = 1.1, df = 2,24, p =.3638; eggs: F =.94, df = 2,24, p =.446) and one week after predatory mite release (motiles: F =.8, df = 2,24, p =.9257; eggs: F =.45, df = 2,24, p =.6441) (Figure 3-4). Two weeks after release, P. persimilis significantly reduced TSSM motile and egg numbers below numbers in the control (motiles: F = 3.4, df = 2,24, p =.513; eggs: F = 4.4, df = 2,24, p =.23). However, TSSM numbers on plants where N. californicus was released were intermediary (Figure 3-4). Both species of predatory mite significantly reduced numbers of TSSM motiles and eggs below the control by week 3 (for motiles: F = 6.2, df = 2,24, p =.68; for eggs: F = 6., df = 2,24, p =.75). TSSM motile and egg numbers on plants where P. persimilis was released began to increase at week 4. At this time, there were significantly fewer motiles per leaflet in the N. californicus treatment compared with the other two treatments (motiles: F = 4.3, df = 2,24, p =.256), however, the difference in egg numbers was of low significance (eggs: F = 1.9, df = 2,24, p =.1717). Experiment 2 As in experiment 1, there were no significant differences in TSSM motile and egg numbers amoung treatments at week (initial sample) (motiles: F =.7, df = 4,2, p =.5983; eggs: F =.25, df = 4,2, p =.956) or week 1 (motiles: F =.5, df = 4,2, p =

32 ; eggs: F =.81, df = 4,2, p =.539) (Figure 3-5). All four treatments significantly reduced TSSM numbers at week 2 compared to the control (motiles: F = 5.8, df = 4,2, p =.28; eggs: F = 5.1, df = 4,2, p =.56) (Figure 3-5). Numbers of TSSM motiles and eggs remained low in all four management treatments at week 3 and all management treatments had significantly lower numbers of TSSM than did the control (motiles: F = 5.5, df = 4,2, p =.36; eggs: F = 3.8, df = 4,2, p =.191). At the end of the experiment, numbers of TSSM motiles in the N. californicus and Acramite treatments were significantly lower than those in the control (F = 4.2, df = 4,2, p =.13). Numbers of TSSM eggs were significantly lower than the control in the N. californicus and P. persimilis/n. californicus treatments (F = 2.6, df = 4,2, p =.696). However, TSSM motile and egg numbers in the P. persimilis treatment increased greatly at week 4. Both TSSM motile and egg numbers increased slightly but not significantly in the Acramite treatment at week 4. There were no significant differences in numbers of P. persimilis motiles and eggs between the P. persimilis and P. persimilis/n. caifornicus treatments (motiles, t =.16, df = 71, p =.871; eggs, t =, df = 78, p = 1). There were also no significant differences in numbers of N. californicus motiles and eggs between the N. californicus and P. persimilis/n. caifornicus treatments (motiles, t =, df = 78, p = 1; eggs, t = -.92, df = 63.1, p =.3619). Discussion Both experiments indicated that N. californicus controls TSSM more effectively than P. persimilis, although both species significantly reduce TSSM numbers below those found in the control. Neoseiulus californicus continuously suppressed TSSM populations. Unlike N. californicus, there were a few failures for P. persimilis especially at week 4

33 22 where high numbers were recorded in two replicates of the Dec.4/Jan.5 repeat of each experiment. Most likely, the P. persimilis adults did not establish on these two plants for unknown reasons. Acramite appears to be highly effective in controlling TSSM populations. However, the slight increase in numbers at 4 weeks suggests that it does not kill 1% of the TSSM and that its affects eventually wear off. Since it can only be sprayed twice in a season, timing of Acramite applications is critical, especially if it is the only mite control strategy employed. Also, the application of N. californicus or P. persimilis following Acramite sprays may be an important management strategy for TSSM control in Florida. The P. persimilis/n. californicus combination treatment significantly reduced TSSM numbers. This strategy appears to be more effective than releasing P. persimilis alone and slightly less effective than releasing N. californicus alone, although these trends were not statistically significant. Releasing both species in combination does not appear to be an economical strategy since it is not any better than using the predatory mites individually. Numbers of predators between the single and combination treatments were not statistically different. This is due to the fact that so few numbers of predators were found. Therefore, no firm conclusions can be drawn from my experiments on the effects of competition when both species of predatory mite are released together. Relatively low numbers of TSSM were recorded on many plants as well. Therefore, experiments need to be repeated before any firm conclusions are drawn.

34 23 A B Figure 3-1. Twospotted spider mite colony. A) a colony heavily infested with TSSM, B) close-up of a heavily infested strawberry plant. A B Figure 3-2. Cage construction. A) diagram showing placement of Velcro and B) a cage laid flat on a laboratory bench. A B Figure 3-3. Greenhouse experimental setup. A) the greenhouse setup for a replicate of experiment 2, B) close-up of several caged plants.

35 24 Average TSSM motiles per leaflet in each treatment TSSM motiles per leaflet a a a b ab b b b a C N P Week after predatory mite release A Average TSSM eggs per leaflet in each treatment TSSM eggs per leaflet a a ab a b b b b a C N P Week after predatory mite release B Figure 3-4. Weekly average TSSM per leaflet in each treatment for Experiment 1. A) TSSM motiles per leaflet and B) TSSM eggs per leaflet. Week is the initial sample taken before predatory mites were released. Error bars represent standard error of the mean. (C = control, P = P. persimilis, and N = N. californicus).

36 25 Weekly average TSSM motiles per leaflet TSSM motiles per leaflet a a a a b b b b b b b b b ab b C P N P/N A Week after predatory mite release A Weekly average TSSM eggs per leaflet 8 TSSM eggs per leaflet a a b a b a b b b bb b b b ab C P N P/N A Week after predatory mite release B Figure 3-5. Weekly average TSSM per leaflet in each treatment for Experiment 2. A) TSSM motiles per leaflet and B) TSSM eggs per leaflet. Week is the initial sample taken before predatory mites were released. Error bars represent standard error of the mean. (C = control, P = P. persimilis, N = N. californicus, P/N = P.persimilis/N. californicus combination, and A = Acramite).

37 CHAPTER 4 SINGLE TREATMENT EFFECTS ON TWOSPOTTED SPIDER MITE CONTROL Both P. persimilis and N. californicus have been shown to effectively control TSSM on strawberry and other crops in various parts of the world. In Florida, P. persimilis has been fairly successful in controlling TSSM populations in south-central Florida (Decou, 1994). However, releases of P. persimilis have not been effective in northern Florida, possibly because it cannot survive colder temperatures during winters found in this area. As a more generalist predator, N. californicus may be able to withstand these conditions and control TSSM populations where P. persimilis cannot. Preliminary studies by Liburd et al. (23) indicate that N. californicus can effectively control TSSM in north Florida strawberries. The purpose of this field study was to compare management of TSSM using either releases of N. californicus or releases of P. persimilis or applications of Acramite. Strawberry production was evaluated to determine if management affected crop yield. Methods This experiment was conducted at the University of Florida, Plant Science Research Unit, Citra. Strawberry plants (var. Festival) were planted in plots 7.3 m x 6.1 m consisting of six rows.5 m wide with.5 m row spacing. The 24 plots were arranged in a 4 x 6 grid and were spaced 7.3 m apart (Figure 4-1A). The experiment was a completely randomized block design with six replicates. Four treatments were evaluated and included: 1) releases of P. persimilis (P), 2) releases of N. californicus (N), 3) application of the reduced-risk miticide Acramite 5WP at the rate of kg product 26

38 27 per hectare (A), and 4) an untreated control (C) (Figure 4-1B,C). During both seasons, each treatment was applied twice. In the 23/24 field season, predatory mites were released on December 11 and February 11 and Acramite was sprayed on December 18 and February 14. In the 24/25 field season, all treatments were applied on December 9 and March 1. Sampling Sampling was initiated once the plants had established. Each week, 6 leaves per plot (24 leaves per treatment) were collected and brought back to the laboratory where the number of TSSM motiles and eggs on each leaf were counted under a dissecting microscope. After predators were released, the numbers of predators and their eggs were also counted. Yield data were collected beginning in early January in both seasons. Strawberries were harvested weekly, and those from the four inner rows were weighed. The two outer rows served as border rows and the yield from these rows was discarded (Figure 4-2). Data Analysis Data were subjected to statistical analysis using the SAS program (SAS Institute, 22). The TSSM motile and egg data were separated into 5 periods based on treatment application dates and time during the season. In the 23/24 field season these periods were: 1) pretreatment (3 weeks prior to any treatment), 2) early-season (post treatment to week 7), 3) mid-season (week 8 to the second application at week 12), 4) early-late season (week 13 to when the second treatment was applied in the 24/25 season at week 16), and 5) late season (weeks 17-2). The late season was split into two periods because of the difference in timing of the second application between seasons. For instance, during the 23/24 season, the second application occurred at the beginning

39 28 of the late season whereas in the 24/25 field season, the second application occurred in the middle of the late season. In the 24/25 field season, the periods were: 1) pretreatment (3 weeks prior to the first treatment), 2) early-season (weeks 4-8), 3) mid-season (week 9 to when the second treatment had been applied in the previous field season at week 13), 4) early-late season (week 14 to when the second treatment was applied on week 16), and 5) late season (week 17 to the end of the season). The late season was split into two periods because of reasons discussed above. In each period, data were log transformed and then treatments were compared using an ANOVA and means were separated using a LSD test. Yield data in each replicate were totaled over the season. The average total yields for each treatment were compared using an ANOVA. A LSD test was used to separate the means for the yield data. This was done for both seasons. Results 23/24 Field Season There were no significant differences in motile and egg numbers between treatments in the pretreatment period (motiles: F =.8, df = 3,15, p =.541; eggs F =.7, df = 3,15, p =.585) or in the early-season (motiles: F =.7, df = 3,15, p =.5511; eggs F = 1.4, df = 3,15, p =.2755) (Figure 4-3). The N. californicus and Acramite treatments had significantly fewer motiles and eggs per leaflet than the control plots during the mid-season (motiles: F = 3.9, df = 3,15, p =.297; eggs F = 5., df = 3,15, p =.137) and during the early-late season (motiles: F = 5.1, df = 3,15, p =.121; eggs F = 4., df = 3,15, p =.28) (Figure 4-3). During these two periods, TSSM numbers in the P. persimilis treatment were fairly high but were not significantly different from TSSM numbers in the N. californicus treatment. Also, numbers of TSSM in the P.

40 29 persimilis treatment were not significantly different from those in the control with the exception of the early-late season in terms of the numbers of motiles (Figure 4-3). There were no significant differences in TSSM motile and egg numbers between treatments in the late season (motiles: F = 1.2, df = 3,15, p =.3461; eggs F =.68, df = 3,15, p =.5777). The same trends can be observed looking at the TSSM motile and egg populations on a weekly basis (Figure 4-4). From early January to mid-march (the mid and early-late seasons), populations of TSSM motiles and eggs were consistently the highest in the control, peaking at 112 ± 44 motiles and 27 ± 85 eggs per leaflet on Feb. 2 (early-late season). The P. persimilis treatment had the second highest TSSM population during this period, peaking at 35 ± 12 motiles on Feb. 11 and 14 ± 71 eggs on Jan. 2 (mid-season). In the N. californicus treatment, TSSM motile populations never exceeded of 1 ± 7 motiles per leaflet or 37 ± 24 eggs per leaflet after the first application. The Acramite treatment also had low TSSM populations, with no more than 11 ± 8 motiles per leaflet and 33 ± 31 eggs per leaflet occurring after the first application. Some interesting trends can be seen when treatments are examined individually. Towards the end of the season predatory mites dispersed into the control plots, causing the TSSM population in these plots to decline (Figure 4-5A). A few predator motiles were found in the Acramite plots after each release, but no eggs were found (Figure 4-5B). The P. persimilis population in the P. persimilis plots peaked at the same time as the TSSM populations (Figure 4-5C). In contrast, the in the N. californicus plots, the predatory mite population peaked about two weeks after the TSSM population (Figure 4-5D).

41 3 The average yield from the P. persimilis treatment was not significantly different than the control (Figure 4-6). The P. persimilis plots averaged 83 ± 5 kg and the control plots averaged 83 ± 8 kg. The N. californicus plots averaged a significantly higher yield of 98 ± 4 kg. The Acramite plots had an average yield between the two of 94 ± 5 kg. 24/25 Field Season In the 24/25 field season, the TSSM population peaked much later. There were no significant differences in TSSM motile and egg numbers in the pretreatment period (motiles: F =, df = 3,15, p = 1; eggs F =, df = 3,15, p = 1) or in the early-season (motiles: F =.47, df = 3,15, p =.773; eggs F =.57, df = 3,15, p =.6432) (Figure 4-7). There were also no significant differences in TSSM motile numbers in the mid-season (F = 1.9, df = 3,15, p =.1699). However, TSSM egg numbers showed a trend of being higher in the Acramite treatment when compared to the N. californicus treatment (F = 2.4, df = 3,15, p =.1129) during the mid-season (Figure 4-7). The control treatment had significantly higher numbers of TSSM motiles and eggs than the N. californicus and P. persimilis treatments in the early-late season (for motiles: F = 11.1, df = 3,15, p =.4; for eggs F = 13., df = 3,15, p =.2) and the late season (motiles: F = 14.6, df = 3,15, p =.1; eggs F = 14.6, df = 3,15, p =.1) (Figure 4-7). During the early-late season, numbers of TSSM motiles in the Acramite treatment did not differ significantly from those found in the control. However, there were significantly less TSSM eggs in the Acramite treatment during this period than in the control. The Acramite treatment had significantly higher numbers of TSSM motiles and eggs than both predatory mite treatments in the early late season. In the late season, Acramite treatment had significantly more TSSM motiles than both predatory mite treatments and significantly more TSSM eggs than the N. californicus treatment (Figure 4-7).

42 31 As with the previous field season, similar trends can be seen when the TSSM population levels are observed on a weekly basis (Figure 4-8). The population of TSSM in the control peaked at 44 ± 23 motiles and 13 ± 47 eggs per leaflet on March 9, 25 (early-late season). TSSM numbers in the Acramite treatment peaked at 38 ± 18 motiles per leaflet on March 9 and 131 ± 56 eggs per leaflet on March 1 (early-late season). In the P. persimilis treatment, TSSM numbers peaked at 7 ± 4 motiles per leaflet and 36 ± 28 eggs per leaflet on March 9 (early-late season). TSSM populations in the N. californicus treatment never exceeded 1 ± 1 motile per leaflet or 7 ± 6 eggs per leaflet after the first application. Examining each treatment individually shows some similarities to and differences from the previous season. Predatory mites again dispersed into the control plots towards the end of the season (Figure 4-9A). This season, they also dispersed into the Acramite plots (Figure 4-9B). As in the previous season, the P. persimilis population peaked at the same time the TSSM population peaked (Figure 4-9C). This occurred in mid-march. A peak in N. californicus population also occurred at this time (Figure 4-9D). There were no significant differences in yield between the four treatments (F = 2.6 df = 3,15 p =.97). Average yield per treatment was much lower than the previous season. Yields averaged 4 ± 1 kg in the control plots, 39 ± 2 kg in the Acramite plots, 35 ± 1 kg in the N. californicus plots, and 37 ± 1 kg in the P. persimilis plots. Discussion In both seasons there were fewer TSSM in the N. californicus treated plots than in the P. persimilis treated plots, although the difference in the 24/25 season was not significant. This suggests that N. californicus is better at suppressing populations of

43 32 TSSM than P. persimilis. Also, P. persimilis may effectively control TSSM when released in seasons when initial TSSM numbers are relatively low. Based on data obtained from the weekly averages, it appears that one release of N. californicus would be sufficient to control TSSM populations, whereas two releases of P. persimilis maybe required. Walzer et al. (21) came to a similar conclusion when they found that N. californicus survived longer than P. persimilis when kept on bean leaves with diminishing prey. Further research would be needed to substantiate this hypothesis. Acramite was highly effective in the 23/24 season but did not adequately control TSSM numbers in the 24/25 season. This was primarily a result of timing. In the 23/24 season, the first spray knocked the TSSM population down and the second spray in early February kept the numbers low. During the 24/25 season, in contrast, there were no detectable TSSM populations when Acramite was first sprayed. By the time TSSM populations began to increase, there was little residual activity left in the Acramite plots. It is possible that if the Acramite plots were sprayed again in February like in the previous season, the results from 24/25 would have been similar to the 23/24 season. In applying Acramite the fact that only 2 applications can be made per field season was taken into consideration. However, the second application of all treatments was delayed until March so TSSM populations could have a chance to increase. Since the Acramite plots were essentially controls (due to delayed application), the TSSM population in these plots exploded and was so high that the second application of Acramite could not reduce TSSM levels. This illustrates how important proper timing of Acramite applications is for growers as well as the importance of using Acramite in combination with other management tactics (predatory mites) to control TSSM.

44 33 There were several differences in the TSSM populations between the two field seasons. In the 23/24 season, TSSM were present in the plots from the beginning, began to increase when the weather got warmer, and died off in all plots by the last few weeks of the season. In contrast, no detectable numbers of TSSM were found in the 24/25 season until mid-january and numbers did not increase greatly until late February. Both TSSM and predatory mite numbers were much lower in the 24/25 season. White and Liburd (24) found that TSSM prefer hot, dry conditions. Late summer and fall of 24 were incredibly wet because of Hurricanes Francis and Jeanne. This could have knocked back the population of TSSM leaving smaller numbers to disperse into the strawberries when they were planted. There was one more difference in the TSSM populations between seasons. The green form of the TSSM was much more prevalent than the red form in 23/24. This trend reversed itself in 24/25. The red form may survive better in wetter conditions than the green form. This would be an interesting topic to research further. Yield was substantially lower in 24/25 than in 23/24 in all treatments. This was primarily due to a greater amount of fungal problems, bird damage, and weed infestation in 24/25. The difference in TSSM population trends between the two seasons may explain why there was a significantly greater yield in the N. californicus treatment in 23/24 but not in 24/25. Sances et al. (1981) found that higher numbers of TSSM are needed to cause a similar level of damage when infestation occurs later in the season. The TSSM population in 24/25 peaked at a much lower number later in the season than the population in 23/24. I suspect that the numbers of TSSM were not high enough to affect the quantity of strawberries produced, although I did

45 34 observe that the quality of fruit harvested from the predatory mite plots looked better than that harvested from the control and Acramite plots. However, quality of marketable yields was not measured in this study. It appears that N. californicus is not as affected by the difference in TSSM population trends as P. persimilis. This is not surprising because it can survive and reproduce on other prey whereas P. persimilis cannot. Acramite is very effective if applications are timed and applied properly. Combinations of Acramite and a release of one of the two predatory mite species could be highly effective in managing populations of TSSM in north Florida strawberries.

46 35 A B Figure 4-1. Field experiment setup. A) Picture of several field plots, B) treatment layout for 23/24 season, and C) treatment layout for 24/25 field season. C

47 Figure 4-2. Weighing the harvest. 36

48 37 Average TSSM motiles for five periods during the 23/24 season 6 a TSSM motiles per leaflet a ab bc c b b b C P N A pretrt early mid early-late late Period A Average TSSM eggs for five periods during the 23/24 season TSSM eggs per leaflet a ab bc c a ab b b C P N A pretrt early mid early-late late Period B Figure 4-3. Average TSSM per leaflet for 5 periods of the 23/24 season. A) TSSM motiles per leaflet and B) TSSM eggs per leaflet (C = control, P = P. persimilis, N = N. californicus, and A = Acramite).

49 38 Weekly average TSSM motiles in each treatment Number of motiles C A P N 11/24/23 12/8/23 12/22/23 1/5/24 1/19/24 2/2/24 2/16/24 3/1/24 3/15/24 3/29/24 Date A Weekly average TSSM eggs in each treatment Number of eggs C A P N 11/24/23 12/8/23 12/22/23 1/5/24 1/19/24 2/2/24 2/16/24 3/1/24 3/15/24 3/29/24 Date B Figure 4-4. Weekly average number of TSSM per leaflet in each treatment during the 23/24 season. A) TSSM motile populations, B) TSSM egg populations. Arrows indicate dates when predators were released; triangles indicate dates when Acramite was sprayed. (C = control, P = P. persimilis, N = N. californicus, and A = Acramite)

50 39 Mites in control plots Number TSSM per leaflet /24/23 12/1/23 12/26/23 1/7/24 1/2/24 2/2/24 2/2/24 Date 3/3/24 3/17/24 3/3/ Number predatory mites per leaflet TSSMm TSSMe Predm Prede A Mites in Acramite treated plots Number TSSM per leaflet /24/23 12/1/23 12/26/23 1/7/24 1/2/24 2/2/24 2/2/24 Date 3/3/24 3/17/24 3/3/ Number predatory mites per leaflet TSSMm TSSMe Predm Prede B Figure 4-5. Weekly Average TSSM and predatory mite populations in each treatment during the 23/24 season. A) In the control, B) in the Acramite treatment, C) in the P. persimilis treatment, D) in the N. californicus treatment. Arrows indicate predatory mite release dates, triangles indicate dates Acramite was sprayed. (TSSMm = TSSM motiles per leaflet, TSSMe = TSSM eggs, Predm = predatory mite motiles, Prede = predatory mite eggs, Pm = P. persimilis motiles, Pe = P. persimilis eggs, Nm = N. californicus motiles, Ne = N. californicus eggs)

51 4 Mites in P. persimilis treated plots Number TSSM per leaflet /24/23 12/1/23 12/26/23 1/7/24 1/2/24 2/2/24 2/2/24 Date 3/3/24 3/17/24 3/3/ Number P. persimilis mites per leaflet TSSMm TSSMe Ppm Ppe C Mites in N. californicus treated plots Number TSSM per leaflet /24/23 12/1/23 12/26/23 1/7/24 1/2/24 2/2/24 2/2/24 Date 3/3/24 3/17/24 3/3/ Number N. californicus mites per leaflet TSSMm TSSMe Ncm Nce D Figure 4-5. Continued.

52 41 Average total yield per treatment Yield (kgs) a ab b b C A Nc Pp Treatment Figure 4-6. Average strawberry yield from each treatment for the 23/24 season. (F = 3.36, df = 3, 15, p =.376)

53 42 Average TSSM motiles in five periods during the 24/25 season TSSM motiles per leaflet pretrt early mid early-late late a b b a a c c b C P N A Period A Average TSSM eggs in five periods during the 24/25 season 1 a b TSSM eggs per leaflet ab ab b a c a c bc c pretrt early mid early-late late b C P N A Period B Figure 4-7. Average TSSM per leaflet for 5 periods during the 24/25 season. A) TSSM motiles per leaflet and B) TSSM eggs per leaflet (C = control, P = P. persimilis, N = N. californicus, and A = Acramite).

54 43 Weekly average TSSM motiles per treatment Number of TSSM motiles per leaflet /22/24 12/6/24 12/2/24 1/3/25 1/17/25 1/31/25 2/14/25 2/28/25 3/14/25 3/28/25 C A P N Date A Weekly average TSSM eggs per treatment Number of TSSM eggs per leaflet /22/24 12/6/24 12/2/24 1/3/25 1/17/25 1/31/25 2/14/25 2/28/25 3/14/25 3/28/25 C A P N Date B Figure 4-8. Average number of TSSM per leaflet in each treatment for each week in the 24/25 season. A) TSSM motile populations, B) TSSM egg populations. Arrows indicate dates when treatments were applied. (C = control, P = P. persimilis, N = N. californicus, and A = Acramite).

55 44 Mites in control plots Number TSSM per leaflet /22/24 12/6/24 12/2/24 1/5/25 1/19/25 2/2/25 2/16/25 Date 3/1/25 3/16/25 3/3/ Average predatory mites per leaflet TSSMm TSSMe Predm Prede A Mites in Acramite treated plots Number TSSM per leaflet /22/24 12/6/24 12/2/24 1/5/25 1/19/25 2/2/25 2/16/25 Date 3/1/25 3/16/25 3/3/ Average predatory mites per leaflet TSSMm TSSMe Predm Prede B Figure 4-9. Weekly average TSSM and predatory mite populations in each treatment during the 24/25 season. A) In the control, B) in the Acramite treatment, C) in the P. persimilis treatment, D) in the N. californicus treatment. Arrows indicate dates treatments were applied (TSSMm = TSSM motiles per leaflet, TSSMe = TSSM eggs, Predm = predatory mite motiles, Prede = predatory mite eggs, Pm = P. persimilis motiles, Pe = P. persimilis eggs, Nm = N. californicus motiles, Ne = N. californicus eggs).

56 45 Mites in P. Persimilis treated plots Number of TSSM per leaflet /22/24 12/6/24 12/2/24 1/5/25 1/19/25 2/2/25 2/16/25 Date 3/1/25 3/16/25 3/3/ Number of P. persimilis per leaflet TSSMm TSSMe Ppm Ppe C Mites in N. californicus treated plots Number of TSSM per leaflet /22/24 12/6/24 12/2/24 1/5/25 1/19/25 2/2/25 2/16/25 Date 3/1/25 3/16/25 3/3/ Number of N. californicus per leaflet TSSMm TSSMe Ncm Nce D Figure 4-9. Continued.

57 CHAPTER 5 TREATMENT COMBINATION EFFECTS ON TWOSPOTTED SPIDER MITE AND PREDATORY MITE SPECIES Trumble and Morse (1993) found that the best economic returns were generated by using applications of abamectin in combination with releases of Phytoseiulus persimilis. Abamectin is a reduced-risk miticidal compound derived from the soil bacterium Streptomyces avermitilis Kim and Goodfellow, which interferes with the nervous system of mites. Similarly, Acramite is a reduced-risk miticide, which is much less toxic to predatory mites than abamectin; therefore, combination treatments with either P. persimilis or Neoseiulus californicus may be a highly effective control strategy. Also, Acramite may be useful as a single treatment to knock down a high population of TSSM before releasing predatory mites A combination of releases of P. persimilis and N. fallacis controls TSSM on strawberries in Taiwan (Lee and Lo, 1989). Therefore, it is possible that a combination of P. persimilis and N. californicus could be an effective treatment strategy to control TSSM. However, Walzer et al. (21) found that when reared together on detached bean leaf arenas, N. californicus eventually displaced P. persimilis. The purpose of these field experiments was to examine the effectiveness of three combination treatments including: P. persimilis/n. californicus, Acramite/N. californicus, and Acramite/P. persimilis and to compare them to single treatment applications. The effects the different combinations had on P. persimilis and N. californicus populations were also examined. 46

58 47 Methods Experiment 1 (23/24 Field Season) This experiment, which was conducted during 23/24 was similar to the ones outlined in chapter 4. An additional treatment was included that consisted of releasing a ½ rate of P. persimilis (half a bottle) and a ½ rate of N. californicus (half a bottle) into the same plots (Figure 5-1). The plots were of the same size (1.3 x 6.1 m) as in the previous experiment outlined in chapter 4. The experiment was a completely randomized block design with four replicates. Five treatments were evaluated including: 1) releases of P. persimilis (P), 2) releases of N. californicus (N), 3) application of a reduced-risk miticide Acramite 5WP (A), 4) releases of both P. persimilis and N. californicus at half the release rate (P/N) and 5) an untreated control (C) (Figure 5-1). Treatments were applied twice during the season. Predatory mites were released on December 11 and February 12. Acramite was applied on December 18 and February 14. Sampling Sampling, once per week, was initiated once the plants had established. Each week, six leaves per plot (24 leaves per treatment) were collected and brought back to the laboratory where the number of TSSM motiles and eggs on each leaf were counted under a microscope. After predators were released, the numbers of predators and their eggs were also counted. No yield data were taken. Data analysis Data were subjected to statistical analysis using the same system described in chapter 4. Briefly, TSSM motile and egg data were separated into five periods based on treatment application dates and times during both seasons. Periods included: 1)

59 48 pretreatment (3 weeks prior to first treatment), 2) early-season (weeks 4-7), 3) midseason (week 8 to when the second treatment was applied on week 12), 4) early-late season (week 13 to when the second application was applied in the 24/25 season at week 15), and 5) late season (weeks 15-19). The late season was split into two periods because of the difference in timing of the second application between seasons. In each period, data were log transformed and then the SAS program (SAS Institute, 22) was used to compare treatments using an ANOVA and separate means using a LSD test. Predatory mite data were compared using Satterthwaite t-tests for unequal variances. Experiment 2 (23/24 Field Season) This experiment was also conducted during the 23/24 field season. The plots were of the same size as in the previous experiments. Two treatments were evaluated: 1) a combination of a half rate of Acramite and a release of N. californicus at half rate (A/N) and 2) untreated control (C) (the same control plots as in experiment 1) (Figure 5-1). Treatments were applied on the same dates as in experiment 1. Sampling, once per week, was initiated once the plants had established. Each week, six leaves per plot (24 leaves per treatment) were collected and brought back to the laboratory where the number of TSSM motiles and eggs on each leaf were counted under a microscope. After predators were released, the numbers of predators and their eggs was also counted. No yield data were taken. Data were subjected to statistical analysis using the SAS program (SAS Institute, 22). A Satterthwaite t-test for unequal variances was used to analyze treatment differences.

60 49 Experiment 3 (24/25 Field Season) This experiment was conducted during the 24/25 field season at the same research unit. The plots were of the same size as in the previous experiments arranged in a 4 x 7 grid. The experiment was a completely randomized block design with four replicates. Seven treatments were evaluated including: 1) releases of P. persimilis (P), 2) releases of N. californicus (N), 3) application of a reduced-risk miticide Acramite 5WP (A), 4) releases of both P. persimilis and N. californicus at half the release rate (P/N), 5) application of a half rate of Acramite and a release of N. californicus at half rate (A/N), 6) application of a half rate of Acramite and a release of P. persimilis at half rate (A/P), and 7) an untreated control (C) (Figure 5-2). Treatments were applied on December 9 and March 1. Sampling Sampling was initiated once the plants had established. Sampling was done once a week. Each week, six leaves per plot (24 leaves per treatment) were collected and brought back to the laboratory where the number of TSSM motiles and eggs on each leaf were counted under a microscope. After predators were released, the numbers of predators and their eggs was also counted. Yield data were not taken for this experiment. Data analysis Data were subjected to statistical analysis using the SAS program (SAS Institute, 22). TSSM motile and egg data were separated into five periods: 1) pretreatment (3 weeks prior to first treatment), 2) early-season (weeks 4-8), 3) mid-season (week 9 to when the second treatment had been applied in the previous field season on week 13), 4) early-late season (week 14 to when the second treatment was applied on week 16), and 5) late season (weeks 17-19). The late season was spilt into two periods for reasons

61 5 discussed in the methods in chapter 4. In each period, data were log transformed and then treatments were compared using an ANOVA and means were separated using a LSD test. Predatory mite data over the entire season were compared using an ANOVA. Results Experiment 1 In the pretreatment period, there were statistically significant differences in TSSM motile numbers (F = 3.3, df = 4,12, p =.467) (Figure 5-3). However, there were no significant differences in egg numbers during this period (F =.9, df = 4,12, p =.4915). There were no significant differences in TSSM motile and egg numbers between treatments in the early-season (motiles: F =.35, df = 4,12, p =.838; eggs F =.27, df = 4,12, p =.4915). In the mid-season, the control and P. persimilis treatments had a trend of higher numbers of TSSM motiles and eggs when compared with the Acramite treatment (motiles: F = 2.4, df = 4,12, p =.182; eggs F = 2.5, df = 4,12, p =.975) (Figure 5-3). In the early-late season, the control had significantly higher numbers of TSSM motiles and eggs that the other four treatments, which did not differ significantly from each other (motiles: F = 3.1, df = 4,12, p =.553; eggs F = 3.4, df = 4,12, p =.43) (Figure 5-3). There were no significant differences in TSSM motile and egg numbers between treatments in the late season (motiles: F = 1., df = 4,12, p =.4449; eggs F = 1., df = 4,12, p =.4449) (Figure 5-3). Unlike the single treatment experiments, few trends can be seen from the weekly averages of the data (Figure 5-4). The control was the highest, peaking at 86 ± 73 motiles and 118 ± 88 eggs per leaflet on January 28, 24 (mid-season). The TSSM population in the P. persimilis/n. californicus plots peaked early in the season at an average of 23 ± 6 motiles and 8 ± 3 eggs per leaflet on December 1 (pretreatment period) and then

62 51 peaked again at 16 ± 2 motiles per leaflet on January 7 and 51 ± 33 eggs per leaflet on January 19 (mid-season). The TSSM population in the P. persimilis plots peaked at 37 ± 71 motiles per leaflet on Jan. 28 and 64 ± 57 eggs per leaflet on Jan. 19 (mid-season). In the Acramite plots, the TSSM population remained below 11 ± 2 motiles and 27 ± 1 eggs per leaflet except for the first week of the study and a spike of 23 ± 6 motiles and 42 ± 11 eggs per leaflet on Jan. 28 (mid-season). The TSSM population in the N. californicus plots never exceeded 17 ± 3 motiles and 3 ± 4 eggs per leaflet. There were significantly less P. persimilis motiles and eggs per leaflet in the combination treatment when compared with the P. persimilis treatment (for motiles t = , df = 469, p =.4; for eggs t = -3.78, df = 467, p =.1). In contrast, the number of N. californicus motiles and eggs in the combination treatment do not differ from those in the N. californicus treatment (for motiles, t =.11, df = 866, p =.4563; for eggs, t = -.57, df = 643, p =.2837). In both the P. persimilis and N. californicus treatments, the predatory motile and egg populations peaked when the TSSM motile population peaked (Figure 5-5). In the P. persimilis/n. californicus treatment, the N. californicus motile population peaked several weeks after the TSSM population while the P. persimilis population remained low throughout the season (Figure 5-6A). Both species were found in the control treatment (Figure 5-6B). Phytoseiulus persimilis motiles were found in the control plots sporadically after both the first and second releases into the treatment plots while N. californicus was found only after the second release. Phytoseiulus persimilis was found in the Acramite treatment once and N. californicus was not found there at all (Figure 5-6C).

63 52 Experiment 2 The combination of Acramite and N. californicus was highly effective against TSSM. There were an average of 2.2 ±.5 motiles and 5.7 ± 1.4 eggs per leaflet in the Acramite/N. californicus treatment compared with an average of 16.9 ± 2.7 motiles and 36. ± 5.5 eggs per leaflet in the control. Differences between the two were highly significant: for motiles, t = -5.39, df = 492, p <.1; for eggs, t = -5.36, df = 512, p <.1. No TSSM motiles or eggs were found in the Acramite/N. californicus treatment after Jan. 19, 24 (Figure 5-7). Neoseiulus californicus motiles were found on only two dates and eggs were found on only one sample date. The second release and spray appear to have been unnecessary. Experiment 3 Twospotted spider mite populations peaked much later in the 24/25 season than in the 23/24 season. There were no significant differences in TSSM motile numbers between treatments in the pretreatment period (F = 1., df = 6,18, p =.4552) or in the early-season (F =.65, df = 6,18, p =.6923) (Figure 5-8). There were no significant differences in TSSM egg numbers between treatments in the pretreatment period (F =, df = 6,18, p = 1). In the early-season, however, there were significant differences in TSSM egg numbers between treatments (F = 2.7, df = 6,18, p =.451) (Figure 5-8B). In the mid-season, there was a trend toward more TSSM motiles and eggs in the control compared to the N. californicus, P. persimilis/n. californicus, Acramite/P. persimilis, and Acramite/N. californicus treatments (motiles: F = 2.1, df = 6,18, p =.19; eggs F = 2.3, df = 6,18, p =.785). In the early-late season and late season, the control had significantly higher numbers of TSSM motiles and eggs than all of the other

64 53 treatments except when compared with egg numbers in the Acramite treatment in the early-late season (early-late season: motiles: F = 7.3, df = 6,18, p =.4; eggs F = 7.8 = 6,18, p =.3; late season: motiles: F = 11.1, df = 6,18, p <.1; eggs F = 11.7, df = 6,18, p =.3) (Figure 5-8). In both periods, there were significantly more TSSM motiles and eggs in the Acramite treatment when compared with the N. californicus, P. persimilis/n. californicus, Acramite/P. persimilis, and Acramite/N. californicus treatments (Figure 5-8). During the early-late season, numbers in the P. persimilis treatment fell between those in the Acramite treatment and those in the other four treatments, but did not differ significantly from any of these treatments. In the late season, however, there were significantly fewer TSSM motiles in the N. californicus and Acramite/N. californicus treatments and significantly fewer TSSM eggs in the N. californicus, P. persimilis/n. californicus, Acramite/P. persimilis, and Acramite/N. californicus treatments compared with the P. persimilis treatment (Figure 5-8). Similar trends can be seen when examining the weekly averages over the season (Figure 5-9). The control had the highest numbers, peaking at 66 ± 32 motiles per leaflet and 225 ± 84 eggs per leaflet on March 9, 25 (early-late season). The next highest was the Acramite treatment, which peaked at 29 ± 26 motiles and 84 ± 72 eggs per leaflet on March 9 (early-late season). The P. persimilis treatment was slightly higher than the other treatments peaking at 19 ± 11 motiles and 54 ± 23 eggs per leaflet on March 16 (late season). The other four treatments were indistinguishable from each other. Twospotted spider mite populations did not exceed 3 ± 3 motiles per leaflet or 1 ± 11 eggs per leaflet in any of these four treatments.

65 54 There were no significant differences between P. persimilis motile and egg populations among the P. persimilis, P. persimilis/n. californicus, and Acramite/P. persimilis treatments (motiles, F = 1.8, df = 2,171, p =.1693; eggs F = 2.3, df = 2,171, p =.114). There were also no significant differences between N. californicus motile and egg populations among the N. californicus, P. persimilis/n. californicus, and Acramite/N. californicus treatments (motiles, F = 1., df = 2,171, p =.3547; eggs F =.9, df = 2,171, p =.489). Few predators were collected in any treatments and TSSM populations were much lower than in the previous year so there were few observable trends (Figure 5-1 and 5-11A). Predatory mites dispersed into the control plots like they did in the previous season (Figure 5-11B). Some also dispersed into the Acramite plots, but numbers declined after the second application (Figure 5-11C). Discussion Several trends seen in the single treatment experiments were also seen here. Overall, N. californicus performed better than P. persimilis in reducing the population of TSSM. Acramite did not adequately reduce TSSM numbers in the 24/25 season (experiment 3), although it did reduce numbers to significantly lower than those in the control. In the 23/24 season (experiment 1), the P. persimilis and P. persimilis/n. californicus combination treatments appeared to take longer to bring the TSSM numbers down than the N. californicus and Acramite treatments. However few of the differences were statistically significant due to high variability in the data. This high variability could be seen in the statistically significant differences in TSSM motile numbers during the pretreatment period.

66 55 The P. persimilis/n. californicus combination treatment differed greatly in its effectiveness between the two seasons. Although the combination treatment effectively reduced TSSM numbers in the 23/24 season, it took longer than the single treatment of N. californicus to do this. In the 24/25 season, however, TSSM numbers remained low throughout the season in both of these treatments. This implies that, like P. persimilis, a combination of the two species is more effective if released in seasons when initial TSSM numbers are relatively low. Data from the 23/24 season show that P. persimilis numbers are greatly reduced in the P. persimilis/n. californicus combination treatment while N. californicus numbers are unaffected. Walzer et al. (21) found that when he reared P. persimilis and N. calfiornicus together on bean leaves, N. californicus eventually displaced P. persimilis. This may have occurred in the P. persimilis/n. californicus combination treatment. It is also possible that N. californicus consumed P. persimilis as well as TSSM in this treatment. Very few predatory mites were found in any treatment in 24/25, so no conclusions could be drawn from this data. Both the Acramite/N. californicus and the Acramite/P. perisimilis treatments effectively controlled TSSM populations in 24/25. The Acramite/N. californicus treatment was also highly effective in 23/24. It appears from the data that the second application of treatments may have been unnecessary. This would be an important topic to research further, as the use of only one application would cut costs in half. Few predatory mites were found in treatments where they were released in combination with applications of Acramite. However, predatory mites were never found on leaves without TSSM or evidence of recent TSSM infestation. Therefore, the low

67 56 numbers of predatory mites most likely reflect the low numbers of available prey rather than an effect of the Acramite application. In both field seasons, Acramite was sprayed on the same day or a few days after predatory mite application. It would be interesting to look into the effects of when each treatment is applied on TSSM control and on predatory mite populations. Figure 5-1. Treatment layout for 23/24 field season. Figure 5-2. Treatment layout for 24/25 field season.

68 57 Average TSSM motiles in five periods during the 23/24 season TSSM motiles per leaflet a ab b b ab a a ab ab b a b b b b C P N A P/N pretrt early mid early-late late Period A Average TSSM eggs in five periods during the 23/24 season TSSM eggs per leaflet a a ab b ab a b b b b C P N A P/N pretrt early mid early-late late Period B Figure 5-3. Average TSSM per leaflet for 5 periods of the 23/24 season. A) TSSM motiles per leaflet and B) TSSM eggs per leaflet (C = control, P = P. persimilis, N = N. californicus, A = Acramite, and P/N = P. persimilis/n. californicus).

69 58 Weekly average TSSM motiles in each treatment Number TSSM motiles per leaflet C A P N P/N 11/26/23 12/1/23 12/24/23 1/7/24 1/21/24 2/4/24 2/18/24 3/3/24 3/17/24 3/31/24 Date A Weekly average TSSM eggs in each treatment Number TSSM eggs per leaflet /26/23 12/1/23 12/24/23 1/7/24 1/21/24 2/4/24 2/18/24 3/3/24 3/17/24 3/31/24 C A P N P/N Date B Figure 5-4. Weekly average number of TSSM per leaflet in each treatment during the 23/24 season. A) TSSM motile populations, B) TSSM egg populations. Arrows indicate dates when predators were released; triangles indicate dates when Acramite was sprayed. (C = control, P = P. persimilis, N = N. californicus, A = Acramite, and P/N = P. persimilis/n. californicus).

70 59 Mites in P. persimilis plots Number TSSM per leaflet /26/23 12/1/23 12/26/23 1/7/24 1/19/24 2/2/24 2/19/24 Date 3/2/24 3/19/24 4/2/ Number P. persimilis per leaflet TSSMm TSSMe Ppm Ppe A Mites in N. californicus plots Number TSSM per leaflet /26/23 12/1/23 12/26/23 1/7/24 1/19/24 2/2/24 2/19/24 Date 3/2/24 3/19/24 4/2/ Number N. californicus per leaflet TSSMm TSSMe Ncm Nce B Figure 5-5. Weekly average TSSM and predatory mite populations in each treatment during the 23/24 season. A) in the P. persimilis treatment and B) in the N. californicus treatment Arrows indicate dates predatory mites were released. (TSSMm = TSSM motiles per leaflet, TSSMe = TSSM eggs, Pm = P. persimilis motiles, Pe = P. persimilis eggs, Nm = N. californicus motiles, Ne = N. californicus eggs)

71 6 Mites in P. persimilis/n. californicus plots Number TSSM per leaflet /26/23 12/1/23 12/26/23 1/7/24 1/19/24 Sample Date 2/2/24 2/19/24 3/2/24 3/19/24 4/2/ Number predatory mites per Leaflet TSSMm TSSMe Ncm Ppm A Mites in control plots Number TSSM per leaflet /26/23 12/1/23 12/26/23 1/7/24 1/19/24 2/2/24 Date 2/19/24 3/2/24 3/19/24 4/2/ Number predatory mites per leaflet TSSMm TSSMe Ncm Ppm B Figure 5-6. Weekly average TSSM and predatory mite populations in each treatment during the 23/24 season. A) in the P. persimilis/n. californicus treatment, B) in the control, and C) in the Acramite treatment. Arrows indicate dates predatory mites were applied; triangles indicate dates Acramite was sprayed. (TSSMm = TSSM motiles per leaflet, TSSMe = TSSM eggs, Pm = P. persimilis motiles, Nm = N. californicus motiles).

72 61 Mites in Acramite treated plots Number TSSM per leaflet /26/23 12/1/23 12/26/23 1/7/24 1/19/24 2/2/24 Date 2/19/24 3/2/24 3/19/24 4/2/ Number predatory mites per leaflet TSSMm TSSMe Ncm Ppm C Figure 5-6. Continued. Mites in Acramite/N. californicus plots Number TSSM per leaflet Number N. californicus per leaflet TSSMm TSSMe Nm Ne 11/26/23 12/1/23 12/26/23 1/7/24 1/19/24 2/2/24 2/19/24 3/2/24 3/19/24 4/2/24 Date Figure 5-7. Weekly average TSSM and N. californicus populations each week in the Acramite/N. californicus treatment during the 23/24 season. Arrows indicate dates predators were released; triangles indicate dates Acramite was sprayed. (TSSMm indicates TSSM motiles, TSSMe indicates TSSM eggs, Nm indicates N. californicus motiles, and Ne indicates N. californicus eggs).

73 62 Average TSSM motiles in five periods during the 24/25 season TSSM motiles per leaflet a b a bc b a ab bab bcc cc bb b c d cd cd d pretrt early mid early-late late Period C P N A P/N A/N A/P A Average TSSM eggs in five periods during the 24/25 season TSSM eggs per leaflet a ab a b b bc a ab b a abab b b b ab b bb b c c c c c c c c pretrt early mid early-late late Period C P N A P/N A/N A/P B Figure 5-8. Average TSSM per leaflet for 5 periods of the 24/25 season. A) TSSM motiles per leaflet and B) TSSM eggs per leaflet (C = control, P = P. persimilis, N = N. californicus, A = Acramite, P/N = P. persimilis/n. californicus, A/N = Acramite/N. californicus, and A/P = Acramite/P. persimilis).

74 63 Weekly average TSSM motiles in each treatment Number of TSSm motiles per leaflet /13/24 11/27/24 12/11/24 12/25/24 1/8/25 1/22/25 2/5/25 2/19/25 3/5/25 3/19/25 4/2/25 4/16/25 C A P N P/N A/N A/P Date A Weekly average TSSM eggs in each treatment Number of TSSM eggs per leaflet /13/24 11/27/24 12/11/24 12/25/24 1/8/25 1/22/25 2/5/25 2/19/25 3/5/25 3/19/25 4/2/25 4/16/25 C A P N P/N A/N A/P Date B Figure 5-9. Weekly average number of TSSM per leaflet in each treatment during the 24/25 season. A) TSSM motile populations, B) TSSM egg populations. Arrows indicate dates when treatments were applied. (C = control, P = P. persimilis, N = N. californicus, A = Acramite, P/N = P. persimilis/n. californicus, A/N = Acramite/N. californicus, and A/P = Acramite/P. persimilis).

75 64 Mites in P. persimilis plots Number TSSM per leaflet /22/24 12/6/24 12/2/24 1/5/25 1/19/25 2/2/25 2/16/25 Date 3/1/25 3/16/25 3/3/ Number P. persimilis per Leaflet TSSMm TSSMe Pm Pe A Mites in Acramite/P. persimilis plots Number TSSM per leaflet /22/24 12/6/24 12/2/24 1/5/25 1/19/25 2/2/25 2/16/25 Date 3/1/25 3/16/25 3/3/ Number P. persimilis per leaflet TSSMm TSSMe Pm Pe B Figure 5-1. Weekly average TSSM and predatory mite populations in each treatment during the 24/25 season. A) in the P. persimilis treatment, B) in the Acramite/P. persimilis treatment, C) in the N. californicus treatment, and D) in the Acramite/N. californicus treatment. Arrows indicate dates treatments were applied. (TSSMm = TSSM motiles per leaflet, TSSMe = TSSM eggs, Pm = P. persimilis motiles, Pe = P. persimilis eggs, Nm = N. californicus motiles, Ne = N. californicus eggs).

76 65 Mites in N. californicus plots Number TSSM per leaflet /22/24 12/6/24 12/2/24 1/5/25 1/19/25 2/2/25 2/16/25 Date 3/1/25 3/16/25 3/3/ Number N. californicus per leaflet TSSMm TSSMe Nm Ne C Mites in Acramite/N. californicus plots Number TSSM per leaflet /22/24 12/6/24 12/2/24 1/5/25 1/19/25 2/2/25 2/16/25 Date 3/1/25 3/16/25 3/3/ Number N. californicus per leaflet TSSMm TSSMe Nm Ne D Figure 5-1. Continued.

77 66 Mites in control plots Number TSSM per leaflet /22/24 12/6/24 12/2/24 1/5/25 1/19/25 2/2/25 2/16/25 Date 3/1/25 3/16/25 3/3/ Number predatory mites per leaflet TSSMm TSSMe Pm Nm A Mites in acramite plots Number TSSM perlleaflet /22/24 12/6/24 12/2/24 1/5/25 1/19/25 2/2/25 2/16/25 Date 3/1/25 3/16/25 3/3/ Number predatory mites per leaflet TSSMm TSSMe Pm Nm B Figure Weekly average TSSM and predatory mite populations in each treatment during the 24/25 season. A) in the P. persimilis/n. californicus treatment, B) in the control, and C) in the Acramite treatment. Arrows indicate dates treatments were applied. (TSSMm = TSSM motiles per leaflet, TSSMe = TSSM eggs, Pm = P. persimilis motiles, Nm = N. californicus motiles).

78 67 Mites in P. persimilis/n. californicus plots Number TSSM per leaflet /22/24 12/6/24 12/2/24 1/5/25 1/19/25 2/2/25 2/16/25 Date 3/1/25 3/16/25 3/3/ Number predatory mites per leaflet TSSMm TSSMe Pm Nm C Figure Continued.

79 CHAPTER 6 CONCLUSIONS The Cost of Control When a cost analysis is done with respect to the treatments evaluated in the studies outlined in this thesis, Acramite is much more cost effective when compared to predatory mites. Acramite costs $ per kg and the recommended rate is kg/ha. The recommended rate for a light infestation of TSSM is 35 bottles/ha, a cost of $587.7 for P. persimilis and $657.3 for N. californicus. This still does not compare to Acramite. However, the recommended rate for a preventative application of predatory mites is only 1 bottles/ha, a cost of $167.9 for P. persimilis and $19. for N. californicus, a price more in line with an Acramite application. My combination treatments were also much more expensive than Acramite alone. At the recommended rate for a light infestation, the P. persimilis/n. californicus treatment would cost $622.19/ha. As a preventative measure, the cost would be $176.88/ha. Since there is no benefit from using this combination treatment, it is not suggested as an option for growers. When combination treatments of Acramite and P. persimilis or N. californicus are used at the recommended rate for a light infestation, it would cost $386.2/ha for Acramite (1/2 rate)/p. persimilis and $421.31/ha for Acramite (1/2 rate)/n. californicus. Using the preventative rate of predatory mites, the cost would be $175.64/ha for Acramite/P. persimilis and $186.56/ha for Acramite/N. californicus. 68

80 69 It would appear that the most cost effective treatment would be either a preventative release of N. californicus if initial TSSM populations are very low or a combination of an application of Acramite and a release of either predatory mite species at the preventative rate if initial TSSM populations are at a higher level. For north Florida, there may be potential to release N. californicus early in the season before TSSM populations build up. Since N. californicus can survive the winter season better than P. persimilis and it provides a better overall reduction, the cost of using N. californicus may be cheaper. Future Directions Several areas need to be researched further before any definite conclusions can be drawn about which treatment is the best option. It is clear from my experiments that two sprays of Acramite are needed to maintain season long control. There is also evidence that a second application of P. persimilis is necessary when initial TSSM populations are high. What is not clear is if two applications of N. californicus are necessary to maintain control. The Acramite/predatory mite combination treatments may give season long control with only one application. A very important point to consider is that I used the rate of release recommended for high levels of TSSM infestation. However, the number of predatory mites recovered was only a small fraction of the predators that were released. Would the rates for light infestation or preventative measures be as effective as the rate for heavy infestations at the same TSSM population levels? What rate of release is necessary to give season-long control with one application? Both are important questions to answer. I can conclude that most of my hypothesis was correct. Neoseiulus californicus did provide better control of TSSM than P. persimilis. Also, N. californicus reduced TSSM

81 7 numbers to well below those found in the control plots in greenhouse studies and in both field seasons. Acramite was highly effective in the greenhouse and in the 23/24 field season. However, it did not perform as well in the 24/25 field season because of improper timing of the second application during this season. Both Acramite/predatory mite combinations were highly effective. These findings give growers several options for TSSM control that are effective and have no significant impact on human health, beneficial insects, or the environment.

82 APPENDIX BEHAVIORAL STUDIES Although both Phytoseiulus persimilis and Neoseiulus californicus are known to be highly effective predators, few studies have been done on the rate at which they consume prey. Gilstrap and Friese (1985) found that adult female P. persimilis consume 1.4 T. cinnabarinus (Boisduval) eggs per hour. This is almost 3x the rate of N. californicus adult females, which consume.41 T. cinnabarinus eggs per hour. Phytoseiulus persimilis adult females prefer tetranychid eggs to larvae while N. californicus adult females have no prey stage preferences (Schausberger and Croft, 1999). This may explain why P. persimilis consumes T. cinnabarinus eggs at a higher rate than N. californicus. This trend may not hold for consumption of larval or adult tetranychids. The purpose of this study was to examine the rate at which P. persimilis and N. californicus adult females consume TSSM eggs and adults and to see whether the rates differ between the two species. Protocol for Laboratory Bioassay The experimental arena consisted of a single, mite-free strawberry leaflet placed in a Petri dish. The leaflet was taped to a piece of filter paper using Scotch double-sided tape (Figure A-1). The leaflet was kept moist by wetting the filter paper with DI water as needed. These Petri dishes were placed upside down and kept at room temperature with a 14:1 photoperiod. A twospotted spider mite colony was maintained on strawberries in the laboratory. It was also kept at room temperature with a 14:1 photoperiod. 71

83 72 Preliminary Experiments Methods Several preliminary tests were conducted to examine the effectiveness of the arenas. In the first preliminary experiment, 1 arenas were constructed. Two TSSM adult females were released into each of five arenas labeled e1 to e5 to determine how many eggs they would lay in a 24 h period. After this 24 h period, the adults were killed and the eggs were counted at 6, 24, and 48 hours to determine if any eggs had hatched. The other five arenas labeled m1 to m5 were left empty for the first 24 h period. Ten to fifteen adult female TSSM were released into each of these five arenas. These were also checked every at 6, 24, and 48 hours to see how many were still alive and how many eggs had been laid. In a second preliminary experiment, 15 arenas were constructed. Five adult female TSSM were released into five of the arenas, 1 into five other arenas, and 15 into the remaining five arenas. The arenas were checked after 24 h to determine how many eggs had been laid. Results After a 24 h period, the TSSM on the e arenas laid an average of 4.4 ± 1.3 eggs. None of the eggs hatched during the next 48 h period. In the m arenas, there was an average percent mortality after 6 h of ± 3.67%, after 24 h of 21. ± 5.81%, and after 48 h of 44. ± 12.52%. Because of this high mortality, 24 h was chosen as the ovipositional period. Eggs were found on these arenas after only 6 h (Figure A-3). In the second preliminary experiment, 5 spider mites had laid an average of 11.4 ± 3. eggs after 24 h, 1 spider mites had laid an average of 12. ± 2. eggs after 24 h, and

84 73 15 mites had laid an average of 3.6 ± 1.2 eggs after 24 h. Based on these results, 5 mites were chosen for the egg consumption experiments. Egg Consumption Over Time by Predatory Mites Methods Fifteen arenas were used (Figure A-2). Five TSSM adult females were released into each arena and allowed to lay eggs for 24 h. The adults were then removed and the initial number of eggs present was counted. A female P. persimilis mite was released into each of five arenas. A female N. californicus mite was released into each of five other arenas. The last five arenas were untreated controls. Experimental design was a completely randomized block design with five replicates of three treatments. The arenas were checked after 1, 3, 6, 12 and 24 hours. The number of remaining TSSM eggs and the number of predatory mite eggs were both counted. Data were acrsine transformed and percent reduction of TSSM eggs was then calculated. Data were subjected to statistical analysis using the SAS program (SAS Institute, 22). Treatments were compared using an ANOVA and means were separated using a LSD test. This experiment was repeated twice. Results In the first of the two trials of this experiment, only nine arenas had eggs laid in them. Five were controls, four were N. californicus, and only one was P. perisimilis. These treatments were not applied but containers were labeled. In the second of the two trials, only 6 arenas had eggs laid in them, two from each treatment. This did not produce enough data to analyze. Predatory mites were not released into arenas where no TSSM eggs had been laid. Observations indicated that both species of predatory mite consume TSSM eggs.

85 74 Adult Consumption Over Time by Predatory Mites Methods Fifteen arenas were used in this experiment (Figure A-2). They were left empty for 24 h. After 24 h, 1 adult TSSM were released into each arena. A P. persimilis female was released into each of five arenas. A N. californicus female was released into each of five other arenas. The last five arenas were untreated controls. Experimental design was a completely randomized block design with five replicates of three treatments. The arenas were checked after 1, 3, 6, 12, and 24 hours. The number of remaining TSSM adults, the number of TSSM eggs, and the number of predatory mite eggs were all counted. Data were acrsine transformed and percent reduction of TSSM eggs was then calculated. Data were subjected to statistical analysis using the SAS program (SAS Institute, 22). Treatments were compared using an ANOVA and means were separated using a LSD test. An analysis was made of the cumulative average (average of the average percent reduction for each time period) using an ANOVA and a LSD test. This experiment was repeated twice. Results There were no significant differences in percent reduction between any of the three treatments (1 h: F =.86, df = 2,12, p =.4486; 3 h: F = 1.8, df = 2,12, p =.277; 6 h: F = 2.6, df = 2,12, and p =.111; 12 h: F = 1.5, df = 2,12, p =.2595; 24 h: F =.34, df = 2,12, and p =.7178) (Figure A-4). Most of the reduction in the control arenas was due to TSSM dying for unknown reasons. Those that died could be observed as intact mites. In the N. californicus and P. persimilis treatments, in contrast, shells of mites that had been fed on could be seen and both species of predatory mite were

86 75 observed feeding upon TSSM. Some of the predatory mites also walked off of the arenas, lowering the average percent reduction. However, when the cumulative average over the 24 h period was analyzed, the percent reduction was significantly lower in the control treatment (F = 25., df = 2,8, p =.4) (Figure A-5). There was no significant difference in the average number of TSSM eggs laid between treatments (F = 1. df = 2,22 p =.386). Very few predatory mite eggs were laid. One P. persimilis laid an egg after 1 h and one N. californicus laid an egg after 24 h. Discussion In the egg consumption experiment, whether the TSSM females would lay eggs was very unpredictable. This is primarily due to the fact that it is not possible to know how many eggs they have already laid or if they have mated or not. To deal with this, one would have to raise a small colony from eggs and use only newly eclosed, freshly mated females. Based on the results, it was confirmed that both species of predatory mites consume TSSM eggs (Blackwood et al., 21; Palevsky et al., 1999). The main problem in the adult consumption experiment was that both TSSM and the predatory mites could walk off the arenas and some did. Clearly, double-sided tape is not sticky enough to confine them. Gilstrap and Friese (1985) used a narrow band of Tree Tanglefoot TM, which proved effective. With a better barrier, differences between the control and the two predatory mite treatments would probably be more conclusive. However, it does not appear that differences between the two predatory mite species would differ significantly, implying that both consume similar amounts of TSSM adults. This differs from Gilstrap and Friese (1985) findings that P. persimilis consumes 3x more T. cinnabarinus eggs than N. californicus. This was probably because one or more of

87 76 several key differences between our experiments. Only adult predatory mite species were followed in these experiments whereas they followed the predatory mites from eggs to adult females. Their prey was T. cinnabarinus eggs whereas in these experiments TSSM adults were used. Lastly, their predatory mites had abundant prey whereas the predatory mites in these experiments had a limited number of prey. Several useful recommendations came out of these experiments. One is that the arenas are effective, but a better barrier is needed to keep mites from walking off of the leaflets. Two, the predatory mites consumed both TSSM eggs and adults and will lay their own eggs under these conditions. Repeating these experiments with better arenas and a more strictly monitored TSSM colony could produce some interesting and valuable results. These techniques will be used in future studies. Figure A-1. Experimental arena.

88 77 Figure A-2. Experimental setup. Number of eggs present on each arena after 6, 24, and 48 h Number of eggs Eggs after 6 h Eggs after 24 h Eggs after 48 h Arena number Figure A-3. Number of eggs laid in each m arena.