The western cabbage flea beetle, Phyllotreta pusilla Horn (Chrysomelidae: Coleoptera), is often the most important insect pest of brassica crops in the Rocky Mountain regions. Flea beetle feeding results in chewed pits on the leaves and cotyledons that, when severe, can cause seedling mortality. Less severe feeding can cause delayed plant development and reduce height, yield, and produce quality. Insecticides have long been the most widely utilized control method for flea beetles. In this study, alternative management tactics such as cultural (planting density, interplanting and host plant resistance), mechanical (trapping), chemical (repellents and insecticides) controls were investigated. In these studies action thresholds were also developed.

The western cabbage flea beetle showed feeding preferences within the Brassicaceae. Chinese cabbage, the cabbage cultivars ‘Golden Acre’ and ‘Copenhagen Market’ and the broccoli cultivar ‘Green Goliath’, were significantly most preferred. Cauliflower, Brussels sprout, collard, kale and the cabbage cultivars ‘Red Acre’, 'Earliana’ and 'Salad Delight’ and broccoli cultivars ‘Love Me Tender Hybrid’ and ‘Premium Crop’ were significantly least preferred by flea beetles.

Within an area, increasing plant density resulted in a significant decrease in flea beetle population infesting individual broccoli plants, but an increase in total number of flea beetles. A 40-cm in-row spacing plant density resulted in a significant decrease in flea beetle numbers in given area, but an increase in numbers infesting individual plants within that area. Interplanting of a more preferred host plant (radish) among a brassica crop can be an effective mechanism to reduce western cabbage flea beetle infestation and subsequent damage. Flea beetle density was highest in plots with no radishes interplanted between broccoli plants. Interplanting radish within broccoli plants did not significantly affect head weight and subsequent yield compared with a sole crop of broccoli.

Flea beetles responded significantly to several visual and chemical stimuli. The colors Saturn green, Saturn yellow and white were significantly the most attractive to flea beetles compared to the transparent control. Allyl isothiocyanate baited traps were significantly attractive to western cabbage flea beetle adults. Traps baited with canola oil were significantly less attractive compared to allyl isothiocyanate. Increase in allyl isothiocyanate concentration up to 2% significantly increased the attractiveness of the traps.

Some of the insecticides and repellents tested in the field showed efficacy in controlling the adults of western cabbage flea beetle. Thiodan (endosulfan) and Asana (esfenvalerate) were significantly the most effective among insecticides. The adult stage of western cabbage flea beetle may be best controlled by the application of Asana for rapid action or by Thiodan for prolonged effect up to two weeks. Margosan-O, SunSpray, Azatin, and diatomaceous earth were significantly the most effective tested repellents. Such repellents may be effective alternatives to conventional synthetic insecticides for the control of western cabbage flea beetle.

Western cabbage flea beetle feeding damage affected cabbage and broccoli plants, causing differences in size and weight of growing heads. Head diameter was significantly larger in plots maintained at 0-beetles per plant threshold compared to head diameter from plots maintained at 10-beetles per plant threshold and the control. Yield obtained from plots maintained at 0, 2 and 5-beetles per plant thresholds in 1997 and 0-beetles per plant threshold in 1998 was significantly higher compared to the 10-beetles per plant threshold and the control. Greatest head size reduction and yield loss occurred at the 10-beetles per plant threshold, suggesting an action threshold below that of 10-beetles per plant. The five flea beetles per plant threshold appears to be an appropriate action threshold for flea beetle control on seedling brassica crops.

Above all, I would like to thank Allah, most Gracious, Most Merciful, for his Compassion and Mercy. I would like to express my deepest thanks and appreciation to my father, Abdulaziz, mother, Meznah, brothers, especially Ahmed, sisters, and other family members for their encouragement, support and endless prayers during my study in the US. This endeavor would not have been feasible without the sacrifice, patience, understanding and encouragement of my dearest wife Haya and my children Abdulaziz and Shatha.

I wish to express my deep sense of gratitude and appreciation to my advisor, Dr. Whitney Cranshaw, for his advice and guidance in connection with the research and preparation and revisions of this manuscript. I am deeply indebted to Dr. Louis Bjostad for sharing his time to help in the trapping experiments and sharing with me his knowledge. I also like to thank my other committee members; Dr. Boris Kondratieff and Dr. Cecil Stushnoff, thank you for your helpful insight and suggestions in this study.

I am deeply indebted for the following for their help throughout this endeavor. Ahemd Al-Jaber, Daniel Gerace, Andrea Tupy, Karen Kramer, Nihat Demirel, Sally McElwey, Jason Bishop, Zana Jevremovic, and Loretta Mannix for providing help during the fieldwork and collection of the data. A special note of thanks goes to my friends, colleagues, and staff members of the college of Agriculture at Al-Qassim, Buriedah, in general and colleagues and faculty of the Department of Plant Protection in particular. I would also extend my thanks and gratitude to all friends, colleagues, and staff members of the Department of Bioagricultural Sciences and Pest Management here who helped me in one way or the other during the completion of this study.

I gratefully acknowledge King Saud University in Saudi Arabia for providing me with the financial support which enabled me to complete this degree. Thanks also extend to my advisor, Dr. Ali Babakir, from the Saudi Cultural Mission in Washington D.C. for his continuous help and support during my study.

DEDICATION

Page

CHAPTER I: 1

GENERAL INTRODUCTION 1

PEST STATUS OF THE WESTERN CABBAGE FLEA BEETLE 2

DESCRIPTION AND LIFE HISTORY 4

FLEA BEETLE MANAGEMENT 5

Cultural Control 6

Biological Control 8

Chemical Control 9

Mechanical Traps 9

CHAPTER II: HOST PLANT PREFERENCE OF THE WESTERN 15

Laboratory Experiments 23

CHAPTER III: EFFECTS OF HOST PLANT DENSITY AND 47

METHODS AND MATERIALS 53

Planting Density Evaluations 53

Interplanting Evaluations 54

RESULTS AND DISCUSSION 57

Planting Density Trials 57

Interplanting Trials 59

REFERENCES CITED 76

CHAPTER IV: ATTRACTIVENESS OF COLOR TRAPS AND ALLY 82

CHAPETR V: FIELD EVALUATION OF DIFFERENT INSECTICIDES 108

CHAPTER VI: DEVELOPMENT OF ACTION THRESHOLDS FOR 135

REFERENCES CITED 160

SUMMARY 162

Flea beetles, subfamily Alticinae (Chrysomelidae: Coleoptera) (derived from the Greek ‘haltikos’ meaning "good at jumping") are minute to medium-sized compact beetles whose enlarged hind femora and jumping habit have earned them the name flea beetles. Flea beetles, Phyllotreta species in particular, are occasional pests of natural and cultivated brassica crops in North America (Bonnemaison 1965, Chittenden 1927, Pimentel 1961a, 1961b, Root and Tahvanainen 1969). The beetles most often inflict substantial damage on early-planted crops.

The three most common flea beetles attacking brassica crops in North America include the crucifer flea beetle, Phyllotreta cruciferae (Goeze), the striped flea beetle, P. striolata (Fabricius) (Metcalf and Metcalf 1993, Jones and Jones 1964), and the western cabbage (black) flea beetle, P. pusilla Horn (Chittenden and Marsh 1920). In addition, a number of other species of flea beetles belonging to different genera attack a variety of important vegetable crops, including: the palestriped flea beetle, Systena blanda Melsheimer; the potato flea beetle, Epitrix cucumeris (Harris); the western potato flea beetle, E. subcrinita LeConte; the tobacco flea beetle, E. hirtipennis (Melsheimer); the tuber flea beetle, E. tuberis Gentner; the eggplant flea beetle, E. fuscula Crotch (Cranshaw 1998); the yellow-black flea beetle, Disonycha xanthomelas (Dalman); the larger striped flea beetle, D. crenicollis Say; the three-spotted flea beetle, D. triangularis Say (Forbes 1900); the corn flea beetle, Chaetocnema pulicaria Melsheimer; and the sweet-potato flea beetle, C. confinis Crotch (Metcalf and Metcalf 1993).

Since the western cabbage flea beetle is abundant and highly mobile, crop rotations are ineffective. However, other cultural practices may show more promise. It is the primary goal of this research to determine if practices such as planting density, interplanting and host plant resistance can assist in management of western cabbage flea beetle. Furthermore, as part of an overall pest management system, the use of other practices (trapping, repellents and insecticide applications, and chemical management using action thresholds) were investigated.

The western cabbage (black) flea beetle, Phyllotreta pusilla, is a serious pest of brassica plant species. It primarily attacks garden crops and can become a pest in large commercial plantings. The species, very common throughout Colorado, was first reported as being particularly injurious to turnips and cabbage (Cooley 1906). In eastern Colorado it has proved potentially limiting to production of several crops, including seeded brassica vegetables, midsummer mustard, and canola. It is particularly damaging to producers of Certified Organic produce for whom management options are limited.

The western cabbage flea beetle ranges from the Dakotas, into Texas and Mexico, westward to Arizona, southern California and Nevada (Horn 1889, Chittenden 1903, Chittenden and Marsh 1920). Phyllotreta pusilla is also widely distributed in the Rocky Mountain region of Colorado and New Mexico. It is known to occur in isolated areas in Arizona, Wyoming, Nebraska, Oklahoma, and Kansas (Chittenden and Marsh 1920). In addition, P. pusilla may have a wider distribution than reported above.

Although the western cabbage flea beetle is primarily a pest of plants of the Brassicaceae family, it also occasionally attacks sugarbeets and other nonbrassica vegetable crops. Historically, turnip, mustard, and radish are generally the most favored food plants. Horseradish, rape, cabbage, cauliflower, water cress, Chinese mustard, nasturtium, bee-plant, sweet alyssum, candyturf, wild peppergrass, hedge mustard, wild water cress, and tansy mustard are other common hosts (Chittenden and Marsh 1920).

As with most other flea beetles, damage is primarily produced by the adult stage. Although the larvae feed on the roots of brassica plants, they cause little appreciable damage. Chief damage is done by the overwintered adult beetles as they move into newly planted areas and by the subsequent early summer generation. Beetles of the first generation are the most destructive to plants, particularly during June and July. Beetles appear suddenly in huge numbers, and large areas can be devastated before the grower becomes aware of their presence (Chittenden and Marsh 1920).

Adult western cabbage flea beetles feed on the foliage of plants. Because of their small size and active habits, flea beetles eat little at each feeding site, producing damage consisting of small round holes often described as "shot holes". Such damage often does little harm to the plant or yield, unless the plant is very small or flea beetles very numerous. However, older leafy vegetables may be damaged by cosmetic injuries to leaves, and very susceptible plants may be defoliated (Chittenden and Marsh 1920).

The adult western cabbage flea beetle is elongate oval, metallic copper in color, and approximately 1.5-2.0 mm long. Winter is spent as adults in protected places, such as under clods of earth, dead leaves, or other debris. They emerge in the spring and feed on wild vegetation, including tansy mustard, Descurainia (=Sophia) pinnata (Walter) and horseradish, Radicula armoracia L., until spring seeded plants are available (Chittenden and Marsh 1920).

Females lay light yellow oval-form eggs approximately 0.5 mm in length. Eggs are deposited in cracks in the soil about the roots of the host plant. The larvae of flea beetles are thread-like in appearance, uniformly white, except for the head sclerites, the legs, and a chitinized area on the caudal subabdominal segment, which are pale chestnut brown. The mature larva is approximately 5 mm in length and from 0.5 to 0.65 mm in width. The larvae spend their lives feeding on the roots of brassica plants. Pupation occurs in the soil in the vicinity of the roots on which larvae fed. The pupa is approximately the same size as the adult and is entirely white (Chittenden and Marsh 1920).

There are typically three generations of beetles produced per year. The complete life cycle from egg to adult varies among generations. The total duration for the first generation takes about 50 days. Chittenden and Marsh (1920) observed egg laying to start 14 April, eggs hatching 24 April, larvae pupating on 23 May, with first adults emerging on 3 June. In the second generation total duration was reported to take about 29 days with eggs deposited on 26 June and hatches on 1 July. Larvae pupated on 19 July and second generation adults first emerged on 25 July. The duration of the third generation is about 39 days. Deposition of eggs started on 12 August hatched on 19 August, and larvae pupated on 10 September. First adults were noticed to emerge on 20 September.

Western cabbage flea beetle management is usually directed at the adult stage early in the spring when plants are small and most susceptible to defoliation. Larval feeding damage to plant roots is minor and control is not generally productive. Later generations of adult flea beetles can cause some feeding damage, but brassica plants are better able to compensate for this injury at this stage through increased summer growth.

Although insecticides are considered the first line of defense especially during outbreaks, some alternative control methods have been given attention. These include the use of cultural practices, feeding deterrents and insect repellents, insect-resistant varieties, and the use of biorational insecticides.

During the last few decades, integrated control has been attempted to minimize the increasing problems with the use of insecticides. Although the approach does not eliminate the use of insecticides as a control method, it regulates their use with the integration of other control methods. The National Academy of Sciences (1969) has defined integrated control as "a concept of pest control in which all control techniques are evaluated and consolidated into a uniform program to manage pest populations so that economic injury is prevented and any detrimental effects to the environment are minimized".

It is the primary goal of this research to determine a control strategy which blends several control options consistent with the integrated control philosophy. This can be achieved through the use of cultural practices such as manipulation of plant density, interplanting, and host plant resistance as well as the use of other control options such as mass trapping. The application of insecticides, repellents and deterrents was also included as a mean to manage flea beetle. Chemical management using action threshold was also included as a part of this project.

A cultural control method that has been used against P. pusilla is clean cultivation. This approach eliminates the brassica weed species (e.g., mustards) that serve as alternative food for the beetles and as breeding sites for their larvae (Chittenden 1903). Also, when planting brassica crops, planting to a well-prepared seedbed can promote rapid seedling growth and help overcome insect injury (Cranshaw 1998). In the fall and after harvest, plowing weed and crop debris can limit overwintering sites for the adult flea beetles and, thus, allows fewer numbers of them to survive through the winter. However, flea beetles are very active and mobile and can migrate from one field to another; making it difficult for these cultural controls to effectively reduce their attack.

Chittenden and Marsh (1920) suggested as an alternative to chemical control the use of an early-season trap crop, where a very small planting (e.g., one percent of anticipated acreage) of preferred brassica crop (e.g., radish, mustard, or turnip) is planted. Trap crops are usually planted along a field edge with adult flea beetles attracted to the tallest, earliest crops available. Once beetles are actively feeding in these trap crops they can be swept or sprayed with an effective foliar insecticide minimizing the development of economically damaging infestations in the primary brassica crop.

Changes in plant density and intercropping or interplanting can be effective approaches in the cultural control of western cabbage flea beetle infesting brassica crops. Increasing plant density may reduce plant infestation, perhaps due to changes in the microenvironment or in the attractiveness of the crop to a particular pest (Coaker 1987). Dense plantings lower the contrast of plants against the soil that is used by flying insects to find their host plant.

Intercropping or interplanting a brassica crop with a susceptible, tolerant, more favorable host plant, which can be used as a trap crop, also may effectively divert flea beetle attack from the main crop (Hokkanen 1991). The more attractive and preferred host plant is utilized as a diversionary crop that can be sacrificed after establishment of the main crop. Interplanting collards with wild mustard resulted in reduced density of flea beetles, P. cruciferae (Altieri and Gliessman 1983). Flea beetle density was reduced on weedy collards compared to those weed-free because the beetles preferred wild mustards to collards.

The employment of resistant cultivars in the cultural control strategy for the long-term management of flea beetles is also important. Flea beetles can discriminate among different oilseed rape cultivars, suggesting some kind of resistance in these cultivars that might be useful for controlling flea beetle damage in oilseed rape. Anderson et al. (1992) studied the feeding preference of the flea beetle, P. cruciferae, for two oilseed rapes, Brassica napus (cv. ‘Westar’) and B. campestris L. (cv. ‘Tobin’); white mustard, Sinapis alba L. (cv. ‘Tilney’); and crambe, Crambe abyssinica Hochs (cv. ‘Meyer’). Flea beetle feeding was significantly higher on oilseed rape than on crambe, and feeding damage was less on white mustard than on oilseed rape. They noticed that crambe did not inhibit flea beetle presence and feeding, and feeding pits were often shallow and smaller. They suggested that crambe tissues might have a gustatory deterrent that inhibits further feeding. Their results demonstrated flea beetle resistance in crambe, and confirmed the reports of Putnam (1977) and Lamb (1984) regarding resistance in white mustard.

Although the western cabbage flea beetle is largely free from natural enemies, several species of birds and three species of internal parasites have been reported to attack the beetle. Perilitus epitricis Viereck is a braconid wasp parasite of the adult flea beetle. Chittenden and Marsh (1920) observed that P. epitricis parasitized 16% of western cabbage flea beetles and were noted to have an even greater influence on the related striped cabbage flea beetle, Phyllotreta vittata Fab. Other parasites that attack the adult beetles are nematodes and gregarine worms. The species of birds reported to feed on P. pusilla are the common and Texas nighthawks, Chordeiles virginianus Gmelin and C. acutipennis Hermann, white-throated swift, Aeronautes saxatalis (=A. melanoleucus) (Woodhouse) horned lark, Otocorys alpestris L., starling, sturnus vulgaris L., song sparrow, Melospiza melodia Wilson, chipping sparrow, Spizella passerina Bechstein, tree swallow, Iridoprocne (=Tachyncincta) bicolor (Vicillot), and marsh wren, Cistothorus (=Telmatodytes) palustris (Wilson) (Chittenden and Marsh 1920).

Most flea beetle treatments are foliar sprays to protect against feeding of the adult beetle, particularly in early spring. In early experiments it was noted by Chittenden and Marsh (1920) that foliar applications did not kill the beetles but drove them away from the treated areas as the insects seemed to avoid treated plants and feed on untreated ones or untreated parts of a plant. It was noted that repellents and deterrents (e.g., tobacco dust and arsenicals) worked more effectively against this insect (Chittenden 1903, Chittenden and Marsh 1920). Application of a strong solution of soapsuds killed the beetle instantly, and sprinkling a mixture of wetted fresh cow manure on the plant drove the beetle away (Howard 1898). As adult flea beetles have the ability to migrate or be carried by wind from one field to the other, long term control with repeated applications may be required.

Traps, in general, can be either attractive, relying on physical or chemical stimulus to lure insects into them, or passive, collecting insect incidentally. Traps with chemical stimulus (i.e., semiochemicals) may be used in Integrated Pest Management (IPM) programs in surveys to monitor for the arrival and/or suppression of a certain pest species. Traps can be also used as a method for monitoring adult flea beetle populations in the field and to determine whether an insecticide application to fields of brassica crop plantings is warranted.

The use of sticky traps can be effective in controlling the active flea beetles especially when they occur in great numbers. The "Wisley turnip-fly trap" has been tested against two flea beetle species, Phyllotreta consobrina Curtis and P. undulata Kutsch in their occurrence on turnip. The trap gave good results in capturing adult flea beetles (Lefroy 1914).

Chemical stimuli can be used as baits to attract insects to traps. Boll weevil traps baited with allyl isothiocyanate were used in a study by Burgess (1984) to monitor for the abundance of P. striolata and P. cruciferae in parkland rapeseed and canola crops in Western Canada. Color traps baited with allyl isothiocyanate, a powerful attractant to flea beetles, may elicit a different response than when the color traps without the bait. Sometimes, the type or color of the traps used with the attractant does not restrain flea beetles from being strongly attracted to the traps. It was reported that the flea beetle, P. striolata, is strongly attracted to traps baited with allyl isothiocynate regardless of the type and color of the traps (Lamb 1983).

The purpose of this project was to evaluate different management tactics for the western cabbage flea beetle, P. pusilla. Therefore, studies were conducted to meet the following program objectives:

1. Objectives related to host plant preference of the western cabbage flea beetle:

1. To evaluate various brassica crop species and cultivars for relative susceptibility and preference of the western cabbage flea beetle.

2. Objectives related to the effect of host plant density and interplanting on the population density of the western cabbage flea beetle:

1. To investigate how differences in host plant density affect flea beetle populations;
2. To determine if interplanting with a flea beetle-susceptible, tolerant host would reduce infestation of the seedling main crop.

3. Objectives related to attractiveness of color traps and allyl isothiocyanate to western cabbage flea beetle adults:

1. To evaluate colors for attractiveness to the beetles;
2. To evaluate allyl isothiocyanates and related compounds for attractiveness.

4. Objectives related to evaluation of various insecticides and repellents for western cabbage flea beetle control:

1. To evaluate the effectiveness of various insecticides and repellents for western cabbage flea beetle control under field conditions.

5. Objectives related to development of action threshold:

1. To determine if an economic threshold could be delineated that would allow management decisions to be made well before feeding damage by the western cabbage flea beetle reached the economic injury level;
2. To establish action thresholds for western cabbage flea beetle population on different host plants.

INTRODUCTION

The western cabbage flea beetle, Phyllotreta pusilla Horn (Chrysomelidae: Coleoptera), is often the most important insect pest of brassica crops in the Rocky Mountain regions (Chittenden and Marsh 1920). These crops are most vulnerable in the spring when adult flea beetles attack newly emerged seedlings. Flea beetle feeding results in chewed pits on the leaves and cotyledons that, when severe, can cause seedling mortality. Less severe feeding can cause delayed plant development and reduced height, yield, and produce quality at harvest. The larvae of P. pusilla feed on the root tissues of brassica plants with no apparent damage to plants (Chittenden and Marsh 1920).

The preferred hosts of P. pusilla are in the family Brassicaceae (= Cruciferae). This pest also attacks other vegetable crops and sugar beets. Its preferred food plants, however, are turnip (Brassica rapa L.), mustards (Brassica. spp.), and radish (Raphanus spp.). Other food plants attacked by P. pusilla include, horseradish (Radicula armaracia L.), cabbage (B. oleracea L.), cauliflower (B. oleracea ssp. botrytis), rape (B. napus L. ssp. oleifera), Chinese mustard (B. juncea (L.)), watercress (Rorippa nasturtium-aquaticum (L.)), beeplant (Cleome serrulata Pursh.), sweet alyssum (Alyssum maritimum L.), candytuft (Iberis spp.), peppergrass (Lepidium spp.), hedge mustard (Sisymbrium spp.), and tansy mustard (Descurainia pinnata (Walter)) (Chittenden and Marsh 1920).

Recent detailed studies on the effects of P. pusilla on cabbage and other brassica crops have not been conducted. Thus, one objective of this study was to quantify the degree of flea beetle feeding on different brassica crops and different cultivars among Brassica species with emphasis on those crops most commonly cultivated. Feeny et al. (1970), sampling plants belonging to 23 different families, reported that adults of the cabbage flea beetles P. cruciferae (Goeze) and the striped flea beetle P. striolata (F.) have a narrow host range. This was restricted among the plants tested to the families Capparidaceae, Brassicaceae (=Crucifereae), and Tropaeolaceae. Similarly, Hicks (1972) tested the food plant preference of the P. cruciferae and P. striolata and reported that of 12 plant species of seven families in the field and 55 species of 31 families in the laboratory, only plants belonging to Capparaceae, Brassicaceae, Tropaeolaceae, and Limnanthaceae families were attacked by adults of both species.

It seems that feeding of these beetles is confined to plant species that contain glucosinolates as secondary plant compounds, which are thioglucosides found in all members of the family Brassicaceae (Feeny et al. 1970, Nielsen 1977, 1989, Rodman and Chew 1980, Louda and Mole 1991). In mustard, for example, mustard oil derived from glucosinolates impart a desirable pungent odor and taste, whereas in other crops (e.g., oilseed rape) glucosinolates and their breakdown products are deleterious to the extracted meal and have now been largely removed by plant breeding. Of more than 200 brassica species analyzed, almost all have thioglucosides (Kjaer 1963). In addition to the Brassicaceae, plants in 10 other families are known to have thioglucosides: Capparaceae, Resedaceae, Tovariaceae, Moringaceae, Limanthaceae, Tropaeolaceae, Caricaceae, Euphorbiaceae, Gyrostemonaceae, and Salvadoraceae (Ettlinger and Kjaer 1968, Larsen 1981). It has been documented that flea beetles, cabbage stem flea beetle, Psylliodes chrysocephala (L.) in particular, are attracted by isothiocyanates, hydrolysis products of glucosinolates released by brassica crops such as the oilseed rape (Bartlet et al. 1992).

The significance of mustard oil glucosides as attractants and of glucosinolates as feeding stimulants for specialized brassica feeding insect species has been acknowledged from the time of Verschaeffelt (1911) and since then has been well documented (Thorsteinson 1953, Wensler 1962, Nayar and Thorsteinson 1963, David and Gardiner 1966, Feeny et al. 1970, Hicks 1974, Nielsen 1978, Dilawari and Atwal 1987). However, host plant preference between flea beetles cannot be explained entirely in term of presence or absence of glucosinolates (Jonasson 1982, Larsen et al. 1985, Lamb, 1988). This is because flea beetles discriminate between different brassica species in a different manner. Phyllotreta cruciferae, for example, differentiate between brassica plants on the basis of the quality and quantity of glucosinolates present (Hicks 1974). Four other species of brassica feeding flea beetles, however, could not discriminate between different glucosinolates (Nielsen 1978). Explanation of this might be that flea beetles also respond to other groups of secondary plant compounds, which act either as additional feeding stimulants or as feeding deterrents. It has been noted that different classes of secondary plant compounds may have important antifeedant effects (Neilsen et al. 1977).

Glucosinolates in Brassicaceae contain a group of 80 thioglucosides that have side chains of varying structure (Larsen 1981, Rodman 1981) and, thus, varying effects upon organisms that come in contact with them or their breakdown products (Chew 1988). In this regard, glucosinolates in some brassica plants may not determine the feeding rate of the flea beetle, P. cruciferae (Bodnaryk and Palaniswamy 1990). Therefore, any interpretation for the putative effects of glucosinolates on insect feeding must be carefully assessed on a case-by-case basis (Chew 1988). In some cases (Bodnaryk 1991), the stage of host plant development, the concentration of a particular glucosinolate compound, and the insect species being considered, will effect the feeding behavior of the insect.

Higher levels of feeding by brassica feeding insects might be expected on plant tissue with high concentrations of glucosinolates. In laboratory tests, Bartlet et al. (1994) reported that glucosinolates stimulated the feeding of P. chrysocephala when added to agar and the amount of feeding that occurred increased with increasing glucosinolate concentrations. It has been documented that glucosinolates vary enormously in their occurrence, amount, concentration, and distribution within the plant (Sang et al. 1984, McGregor 1988). Thus, it is important to relate insect feeding with the quality and quantity of glucosinolates actually present in the plant tissues being consumed. It has been found that feeding damage of flea beetles of the genus Phyllotreta to rape cultivars with high or low glucosinolates has no correlation with the glucosinolate designation (Jonasson 1982, Larsen et al. 1985, Lamb 1988).

Bartlet and Williams (1991) testing the feeding acceptability of 40 different plants to P. chrysocephala, found that only glucosinolate-containing plants were accepted as food. When they added the glucosinolate sinigrin (allyl glucosinolate) to leaf discs of the rejected food plants, flea beetles did not accept them as food. Bartlet and Williams have also found that when solvent extracts of the rejected plants are applied to oilseed rape, feeding by adults is inhibited. They concluded that the feeding of P. chrysocephala is influenced by the presence or absence of glucosinolates, which may act as feeding stimulants, and other, unknown chemicals which act as feeding inhibitors.

Several studies were conducted to identify the feeding response of different brassica feeding flea beetles. Tokunaga and Kadowski (1949) studied the feeding response of P. striolata on 23 species of Brassicaceae and 19 species belonging to other families. They found that the beetles would attack only plant species belonging to Brassicaceae and Cucurbitaceae. The feeding preference of flea beetles, P. cruciferae and P. striolata in particular, with respect to various brassica species have been investigated in the United States and Europe (Newton 1928, Haddock 1945, Dobson 1956, Brett and Rudder 1966, Tahvanainen 1971).

Plant species of Brassicaceae and plant cultivars may differ in their susceptibility and attack by flea beetles. Flea beetles can discriminate between different oilseed rape cultivar (Brett and Rudder 1966, Putnam 1977, Lamb 1980, 1984, 1988, Lamb and Palaniswamy 1990, Bodnaryk and Lamb 1991) suggesting some kind of resistance in these cultivars that might be useful for controlling flea beetle damage in oilseed rape. Anderson et al. (1992) have conducted field and laboratory choice tests to determine the feeding preference of the flea beetle, P. cruciferae, for two oilseed rapes, Brassica napus (cv. ‘Westar’) and B. campestris L. (cv. ‘Tobin’); white mustard, Sinapis alba L. (cv. ‘Tilney’); and crambe, Crambe abyssinica Hochs (cv. ‘Meyer’). They found that flea beetle feeding was significantly higher on oilseed rape than on the crambe and feeding damage was less on white mustard than on oilseed rape. They noticed that crambe did not inhibit flea beetle presence and feeding, and feeding pits were often shallow and smaller. They have suggested that crambe tissues may have a gustatory deterrent that inhibits further feeding. Their results demonstrated flea beetle resistance in crambe, and confirmed the reports of Putnam (1977) and Lamb (1984) on resistance in white mustard. Anderson and his group (1992) have also reported that P. cruciferae had the tendency to aggregate once feeding was initiated resulting in seedlings with extensive feeding on one cotyledon and little on the opposite cotyledon. This suggests that upon initiating feeding a flea beetle would influence the feedings of another flea beetles feeding.

Flea beetles may be selective in their choice of brassica hosts, suggesting that host plant selection can be modified by factors other than glucosinolate presence (Feeny et al. 1970). Morphological and physical characteristics (i.e., trichomes, surface waxes, silication, toughness, sclerotization, color, and shape) of brassica plants may effect herbivory by providing hostile surfaces or by evoking altered host behavior (Norris and Kogan 1980). However, such discrimination has been attributed to the chemical composition and concentration of glucosinolates in the plant (Hicks 1974, Tahvanainen 1983). The glucosinolate composition of brassica plants changes with plant growth stage (Milford et al. 1989), with level of damage by flea beetles (Koristas et al. 1991), as well as with plant lines or cultivars (Bartlet et al. 1996). When testing eight oilseed rape lines in the field, Bartlet and his colleagues (1996) found that the concentration of isothiocyanate-releasing glucosinolates varies greatly, exhibiting a twenty-fold variation in concentration of these glucosinolates, whereas those tested in the laboratory exhibited little variation.

Chemically-based resistance in different brassica crops and their components (antixenosis, antibiosis or tolerance) to the western cabbage flea beetle is unknown. In a related species, P. cruciferae, it has been documented that the presence of high concentration of p-hydroxy-benzyl glucosinolate (sinalbin) in the cotyledons of young seedlings and in young leaves of the mustard, Sinapis alba, will deter the feeding of P. cruciferae (Bodnaryk 1991). Prior wounding as a result of insect feeding or mechanical damage also enhances the resistance to insect herbivores in some plants (Kogan and Paxton 1983, Karban and Myers 1989, Lin et al. 1990, Coleman and Jones 1991, Karabn 1991, Kogan and Fischer 1991, Tallamy and Raupp 1991). This can be attributed to chemical alterations as a result of wounding, leading to a decrease in the susceptibility or nutritional quality of wounded plants to the herbivores (Moran and Hamilton 1980, Kogan and Paxton 1983, Edwards et al. 1985, Haukioja and Neuvonen 1987, Lin et al. 1990).

Studies in this section of the research were conducted to investigate, by field and laboratory experiments, the host plants preference of the western cabbage flea beetle when offered various brassica hosts. Such investigation may allow a tentative distinction between a mechanism of host plant selection and preference based on host plant physical and/or habitat characteristics, as well as a mechanism based on secondary plant compounds that may attract or repel the insect.

METHODS AND MATERIALS

Host plant preference trials in the field were conducted at the Horticulture Field Research Center in Fort Collins, Colorado. Experiments were performed during the summer of 1996 and 1997 on naturally occurring populations. Individual plots consisted of single rows, 77-cm row spacing, each 7.5-m in length in 1996 and 6-m in length in 1997, arranged in a randomized complete block design with different brassica crops and different cabbage cultivars as treatments. Each study involved four replications. Brassica crops and cabbage cultivars were planted in the greenhouse for about 4 weeks prior to transplanting into the field. After transplanting into the field, plants were grown under standard husbandry practices and plots were kept weed-free. Because flea beetle adults are very active, especially in the middle of the day when it is warm and sunny, and may quickly hop or fly away before they are counted, a special method called it the "sneak-up" was adopted. This involved counting flea beetles in early mornings when beetles are less active and avoid making noises and shadowy figures around the counted plants to avoid disturbing flea beetles.

In 1996, two separate trials, each in a separate row, were conducted. The first trial included four different brassica crops, cabbage Brassica oleracea L., cv. ‘Golden Acre’; Chinese cabbage B. rapa L ssp. pekinensis., cv. ‘Michihili’ (Lake Valley Seed Inc.); cauliflower B. oleracea ssp. botrytis, cv. ‘Snow Ball’ (The Rocky Mountain Seed Co.); and Brussels sprout B. oleracea spp. Gemmifera, cv. ‘Long Island Improved’ (Lake Valley Seed Inc.). The second trial included four different cabbage cultivars, ‘Golden Acre’, ‘Copenhagen Market’ (Lake Valley Seed Inc.), ‘Chieftain Savoy’, and ‘Red Acre’ (The Rocky Mountain Seed Co.). Plants for each trial were transplanted on 13 May in 41-cm in-row spacing. Insect counts were made on 17, 21, and 30 May; 3, 6, 12, 18, and 26 June; and 3 and 11 July, by whole plant examination of 6 plants in the center of each plot.

In 1997, two trials were conducted. The first trial included eleven brassica crops, Chinese cabbage, cv. ‘Michihili’; Brussels sprout, cv. ‘Long Island Improved’; mustard greens B. juncea (L.), cvs. ‘Green Wave’ and ‘Florida Broadleaf’; collard B. oleracea ssp. acephala, cv. ‘Champion’; turnip B. campestris ssp. rapifera, cvs. ‘Purple Top White Globe’ and ‘Seven Top’ (The Rocky Mountain Seed Co.); kale B. oleracea ssp. acephala, cv. ‘Dwarf Blue Curled Vates’; and broccoli raab Brassica rapa, cv. ‘Spring’ (Lake Valley Seed Inc.). The second trial tested six different cabbage cultivars, ‘Golden Acre’, ‘Copenhagen Market’, ‘Chieftain Savoy’, ‘Red Acre’, ‘Dynamo’ (The Rocky Mountain Seed Co.), and ‘Early Flat Dutch’ (Lake Valley Seed Inc.). Plants for each trial were transplanted on 20 May in a 41-cm in-row spacing. Insect counts were made on 2, 6, 12, 16, and 23 June; and 2 July, by whole plant examination of 5 plants in the center of each plot.

Host plant preferences in the laboratory were conducted in the summer of 1997, using 12 different brassica crops, 11 cabbage cultivars, and 4 broccoli cultivars. Brassica crops included Chinese cabbage, cv. ‘Michihili’; Brussels sprouts, cv. ‘Long Island Improved’; mustard, cvs. ‘Green Wave’ and ‘Florida Broadleaf’; collard, cv. ‘Champion’; turnip, cvs. ‘Purple Top White Globe’ and ‘Seven Top’; kale, cv. ‘Dwarf Blue Curled Vates’; broccoli raab, cv. ‘Spring’); broccoli, B. oleracea ssp. italica, cv. ‘Premium Crop’; cauliflower, cv. ‘Snow Ball’; and Bok Choy, B. juncea L. ssp. rugosa, cv. ‘Oriental Mustard Cabbage’ (The Rocky Mountain Seed Co.). Cabbage cultivars included ‘Golden Acre’, ‘Copenhagen Market’, ‘Chieftain Savoy’, ‘Red Acre’, ‘Dynamo’, ‘Early Flat Dutch’, ‘Hybrid Stonehead’, ‘Earliana’, ‘Fast Ball’ (W. Atlee Burpee & Co.), ‘Salad Delight’, and ‘Early Jersey Wakefield’ (Week Seed Co.). Broccoli cultivars included ‘Premium Crop’, ‘Green Goliath’ (Henry Field Seed & Nursery Co.), ‘Love Me Tender Hybrid’ (W. Atlee Burpee & Co.), and ‘Green Comet Hybrid’ (Thompson & Morgan Inc.).

Tests were conducted using screened cages (35-cm wide by 68-cm high by 61-cm deep) with clear polystyrene tops. Cages were placed under wide-spectrum artificial lighting. Plants used in all trials were grown in flats in the greenhouse, watered as needed, and fertilized. Ten-day seedlings were used for each of the experiments. Before the transfer to the lab, seedlings with the media was transferred to 0.03 liter plastic cups (Prairie Packaging Inc.). Four seedlings of each crop/cultivar were placed randomly inside the cage. About 100 adult flea beetles were introduced into the cage and left to feed for 24 hours. Seedlings were then removed from the cage for feeding-preference evaluation. Insect feeding to seedlings was estimated by counting the flea beetle feeding holes produced in leaves. In each trial seedlings belonging to a crop/cultivar were replicated four times (4 seedlings per crop/cultivar) and repeated three times (three different test dates).

To examine the effects of host plant on the colonization and infestation of western cabbage flea beetle, data were analyzed with two-way ANOVA (SAS institute 1985), testing for effects of host plant and block. Analyses were carried out on the mean number of flea beetles per plant for each treatment for each block. Comparisons of host plants were carried out using Student-Newman-Keuls (SNK) (SAS institute 1985).

RESULTS AND DISCUSSION

Flea beetles in field and laboratory tests demonstrated variation in feeding preference among different brassica crops, cabbage and broccoli cultivars.

In the 1996 evaluations (Table 2.1), the number of flea beetles observed was significantly different among brassica crops (P<0.05). Chinese cabbage was significantly the most attractive to flea beetles on all, except on the 30-May, sampling dates (F=5.40, P=0.0211; F=8.48, P=0.0055; F=2.34, P=0.1421; F=30.70, P=0.0001; F=26.83, P=0.0001; F=11.03, P=0.0023; F=61.48, P=0.0001; F=9.04, P=0.0044; F=62.86, P=0.0001; F=7.60, P=0.0077, respectively). Cauliflower, Brussels sprouts and cabbage were significantly less attractive to flea beetles compared to Chinese cabbage. In the 1997 evaluations (Table 2.2), the number of flea beetles observed was significantly different among brassica crops (P<0.05). Again, Chinese cabbage, as well as mustards, and broccoli raab were significantly most attractive to flea beetles on most sampling dates (F=6.75, P=0.0001; F=13.67, P=0.0001; F=4.67, P=0.0019; F=4.45, P=0.0021; F=4.79, P=0.0013; F=45.57, P=0.0001, respectively). Collards, kale and Brussels sprout were significantly less attractive to flea beetles.

In the 1996 evaluations of cabbage cultivars (Table 2.3), the number of flea beetles observed was significantly different on the 3, 6, and 18 June sampling dates (F=5.99, P=0.0158; F=37.76, P=0.0001; F=9.32, P=0.004, respectively). The cultivars 'Red Acre' and 'Chieftain Savoy' were significantly less attractive to flea beetles, while the cultivars 'Golden Acre' and 'Copenhagen Market' were significantly the most attractive to flea beetles. On sampling dates 17, 21 and 30 May, 12 and 26 June, and 3 and 12 July, there were no significant differences in the number of flea beetles colonizing different cabbage cultivars (F=0.58, P=0.6417; F=1.72, P=0.2323; F=2.39, P=0.1368; F=3.08, P=0.0827; F=1.90, P=0.1995; F=2.88, P=0.0956; F=0.39, P=0.7632, respectively). In the 1997 cabbage cultivars preference evaluations (Table 2.4), numbers of flea beetles again were significantly different among the cabbage cultivars studied (P<0.05). 'Copenhagen Market', ‘Dynamo’ and 'Golden Acre' were significantly more attractive to flea beetles on 12 and 23 June (F=4.05, P=0.0158; F=3.14, P=0.0296, respectively) than cultivars 'Red Acre' and 'Early Flat Dutch'. On the 2 July sampling date, only the cultivar ‘Copenhagen Market’ was significantly more attractive to flea beetles (F=3.55, P=0.0256) than other cabbage cultivars. There were no significant differences in the number of flea beetles colonizing cabbage cultivars on the 2, 6, and 16 June sampling dates (F=0.65, P=0.6674; F=1.27, P=0.3261; F=3.40, P=0.0298, respectively).

In the laboratory evaluation of brassica crops (Table 2.5), the number of flea beetle feeding holes was significantly different among cruciferous crops (P<0.05). The number of feeding holes produced was significantly higher in Chinese cabbage on 25 and 26 June (F=6.49, P=0.0001; F=6.08, P=0.0001) compared to Brussels sprout, cauliflower, broccoli, collard, kale, and mustard (cv. ‘Green Wave’). On the 27 June sampling date, Chinese cabbage and turnip (cv. ‘Seven Top’) had a significantly higher number of feeding holes (F=3.38, P=0.0032) than Brussels sprout and broccoli.

In the laboratory evaluation of cabbage cultivars (Table 2.6), the number of flea beetle feeding holes was not significantly different among cabbage cultivars on all sampling dates (F=0.85, P=0.5860; F=1.59, P=0.1571; F=0.37, P=0.9495, respectively). In the broccoli cultivars evaluation (Table 2.7), the number of flea beetles holes was not significantly different among broccoli cultivars (P>0.05) except on the 16 July sampling date (F=7.25, P=0.009). 'Green Goliath' cultivar had a significantly higher number of flea beetle feeding holes than the 'Love Me Tender Hybrid' cultivar.

Table 2.1. Mean number of western cabbage flea beetles per plant observed on different brassica crops, Colorado State University Horticulture Field Research Center, Fort Collins, Colorado, 1996.