Economic injury levels for western flower thrips (Thysanoptera: Thripidae) on greenhouse cucumber

Low, medium and high densities of western flower thrips, Frankliniella occidentalis (Pergande), were established in three greenhouses at the Greenhouse and Processing Crops Research Centre, Ontario, Canada, in 1996 and 1998 to develop economic injury levels for thrips on greenhouse cucumber. Thrips densities were monitored weekly using yellow sticky traps and flower counts. Fruit was harvested twice a week, graded for size, weighed, and rated for thrips damage using three damage categories. Significant yield reduction was detected 4 wk after severe fruit damage was observed in the high and medium thrips density treatments in 1996 and 7 wk in 1998. Percentage of severe damaged fruit (P(F3)) has significant linear relationships with the adult thrips density (x) that was sampled by sticky traps 1 wk before harvest (P(F3) = -0.2533 + 0.0828x) and that was sampled by flower counts 2 wk before harvest (P(F3) = -0.2025 + 0.5490x). Based on the regression equations, economic injury levels, expressed as adult thrips per sticky trap per day or adult thrips per flower, were calculated for various combinations of control costs, yield potential and fruit prices. The economic injury levels for F. occidentalis ranged from 20 to 50 adults per sticky trap per day or 3 to 7.5 per flower as determined under average greenhouse production conditions in Ontario, Canada.     

The Mirid Bug Dicyphus tamaninii as a Greenhouse Whitefly and Western Flower Thrips Predator on Cucumber

Dicyphus tamaninii Wagner is a polyphagous predator that also feeds on plants. Its effectiveness against several pests has been tested and good results have been obtained on tomato, but not on cucumber. This paper reports the results of three exclusion cage experiments. The first was carried out to evaluate the effectiveness of D. tamaninii for reducing greenhouse whitefly and western flower thrips populations on cucumber. The second experiment examined the safety of the mirid bug in terms of the potential for damaging cucumber fruits even at very low prey densities. Two predator (3 and 18 insects/plant) and variable prey release rates were used. The predator was able to reduce whitefly populations greatly, even at high initial pest infestation levels, especially at the higher release rate. Western flower thrips populations were maintained at very low densities a few weeks after the predator had been introduced into cages at the two release rates tested. The mirid bug was barely able to reproduce on cucumber in the absence of prey. Mirid feeding punctures on cucumber fruit or yield decrease due to the action of mirids were not observed. The potential of this mirid bug for inoculative and conservation biological control is discussed.     

Ontogenetic shifts in intraguild predation on thrips by phytoseiid mites: the relevance of body size and diet specialization

In greenhouse agroecosystems, a guild of spider mite predators may consist of the oligophagous predatory mite Phytoseiulus persimilis Athias-Henriot, the polyphagous predatory mite Neoseiulus californicus McGregor (both Acari: Phytoseiidae) and the primarily herbivorous but facultatively predatory western flower thrips Frankliniella occidentalis Pergande (Thysanoptera: Thripidae). Diet-specialization and the predator body size relative to prey are crucial factors in predation on F. occidentalis by P. persimilis and N. californicus. Here, it was tested whether the relevance of these factors changes during predator ontogeny. First, the predator (protonymphs and adult females of P. persimilis and N. californicus): prey (F. occidentalis first instars) body size ratios were measured. Second, the aggressiveness of P. persimilis and N. californicus towards F. occidentalis was assessed. Third, survival, development and oviposition of P. persimilis and N. californicus with F. occidentalis prey was determined. The body size ranking was P. persimilis females > N. californicus females > P. persimilis protonymphs > N. californicus protonymphs. Neoseiulus californicus females were the most aggressive predators, followed by highly aggressive N. californicus protonymphs and moderately aggressive P. persimilis protonymphs. Phytoseiulus persimilis females did not attack thrips. Frankliniella occidentalis larvae are an alternative prey for juvenile N. californicus and P. persimilis, enabling them to reach adulthood. Females of N. californicus but not P. persimilis sustained egg production with thrips prey. Within the guild studied here, N. californicus females are the most harmful predators for F. occidentalis larvae, followed by N. californicus and P. persimilis juveniles. Phytoseiulus persimilis females are harmless to F. occidentalis.     

Plant spatial distribution and predator–prey ratio affect biological control of the twospotted spider mite Tetranychus urticae (Acari: Tetranychidae) by the predatory mite Phytoseiulus persimilis (Acari: Phytoseiidae)

A greenhouse experiment was conducted to determine the effect of plant spacing and predator–prey ratio on dispersal and foraging efficiency of the predatory mite, Phytoseiulus persimilis, on the twospotted spider mite, Tetranychus urticae. When predators were released at the end of spider mite-infested arrays of lima bean plants that had either no spacing or two different patterns of spacing among plant rows, plant damage was uniformly low throughout the experiment at both predator–prey ratios (1:10 and 3:10) in the treatment with no spacing. In contrast, damage was higher in both treatments where plant rows were interrupted by spacing. At the 1:10 ratio, more plants closer to the predator release point experienced moderate damage than at the 3:10 ratio where only the plant rows farthest from the release point had unacceptable damage. Our findings suggest that point releases of P. persimilis at the standard 1:10 predator–prey ratio should be effective within a diameter of at least 65?cm on mite-infested patches of plants where pots are touching. However, if gaps in plant rows exist, even large numbers of predators may not be sufficient to protect parts of the crop unless predators are released at shorter fixed points in the greenhouse crop.     

The Effects of Timing and Rates of Release of Phytoseiulus persimilis against Two-spotted Spider Mite Tetranychus urticae on Dwarf Hops

Single inoculative releases of the phytoseiid mite Phytoseiulus persimilis were made against the two-spotted spider mite, Tetranychus urticae, on two varieties of dwarf hops in 1996 and 1997 at means of 20, 10, 5, 2.5 and nil per plant, and at up to three timings. The numbers of spider mites recorded on leaves after the predators were released were related inversely to the rates of release. The earliest releases of the predator maintained spider mites at lower population densities than did those made later in the year. In all treatments the numbers of spider mites decreased when the prey:predator ratio reached approximately 10:1.     

Host plant, temperature, and photoperiod effects on ovipositional preference of Frankliniella occidentalis and Frankliniella fusca (Thysanoptera: Thripidae)

Host plant effects of tomato, Lycopersicon esculentum Mill., and chickweed, Stellaria media (L.) Vill., foliage infected and uninfected with Tomato spotted wilt virus (family Bunyaviridae, genus Tospovirus, TSWV) on the ovipositional preferences of western flower thrips, Frankliniella occidentalis (Pergande), and tobacco thrips, Frankliniella fusca (Hinds), were investigated for whole plants in the greenhouse. In addition, the preference for leaf disks from the same host plants was investigated under a range of temperatures, 15-30 degrees C at a photoperiod of 12:12 (L:D) h, and at three photoperiods, 6:18, 12:12, and 18:6, at 20 degrees C in no-choice and choice studies conducted in growth chambers. In a choice test, F. fusca oviposited significantly more eggs per whole plant foliage over a 7-d period than F. occidentalis by an average ratio of 3:1 over both tomato and chickweed. The optimum temperature for oviposition of F. occidentalis and F. fusca was 24.5 and 24.9 degrees C, respectively. Both species laid significantly more eggs under the longest daylight hours tested, 18:6, in the choice study. Temperature and photoperiod did not significantly interact in terms of thrips ovipositional preference. Ovipositional preference for chickweed or tomato foliage was different for each thrips species in the choice and no-choice tests. However, both thrips species laid significantly more eggs per square centimeter of leaf area in chickweed than in tomato in the whole plant choice test.     

Non-consumptive effects of predatory mites on thrips and its host plant

Recent reviews on trait-mediated interactions in food webs suggest that trait-mediated effects are as important in triggering top–down trophic cascades as are density-mediated effects. Trait-mediated interactions between predator and prey result from non-consumptive predator effects changing behavioural and/or life history traits of prey. However, in biological control the occurrence of trait-mediated interactions between predators, prey and plants has been largely ignored. Here, we show that non-consumptive predator effects on prey cascade down to the plant in an agro-ecological food chain. The study system consisted of the predatory mites P. persimilis and N. californicus , the herbivorous non-target prey western flower thrips F. occidentalis and the host plant bean. Irrespective of predator species and risk posed to prey, the presence of predator eggs led to increased ambulation, increased mortality and decreased oviposition of thrips. Furthermore, the presence of predator eggs reduced leaf damage caused by thrips. To our knowledge this is the first experimental evidence suggesting a positive trophic cascade triggered by non-consumptive predator effects on non-target prey in an augmentative biological control system.     

Biological Control of Thrips on Ornamental Crops: Interactions Between the Predatory Mite Neoseiulus cucumeris (Acari: Phytoseiidae) and Western Flower Thrips, Frankliniella occidentalis (Thysanoptera: Thripidae), on Cyclamen

Difficulties in controlling outbreaks of Western flower thrips, Frankliniella occidentalis, have obstructed the widespread adoption of biological control in many ornamental crops. The efficacy of the predatory mite, Neoseiulus cucumeris, in controlling F. occidentalis on two cultivars of cyclamen was tested in glasshouse experiments. The establishment and development of F. occidentalis populations was compared in three treatment introductions of N. cucumeris (50, 200 and 350 mites m -2 per week) and an untreated control. F. occidentalis were sampled in the flowers over eight weeks and counted into different life stages. No differences were observed between the two cultivars. All treatments with the predator resulted in a decline in numbers of F. occidentalis compared to the untreated control. Although the proportion of first instar F. occidentalis was similar in all treatments, the level of control varied with the number of N. cucumeris introduced. Lower populations of F. occidentalis, combined with a more rapid decline in their numbers, were observed at the 200 and 350 mites m -2 rates. Numbers of F. occidentalis remained low in the 350 N. cucumeris m -2 rate and the proportion of second instar F. occidentalis in the samples was consistently lower than in the other treatments. Trap counts of adult F. occidentalis were strongly correlated with the numbers of both adult and total F. occidentalis in flower samples. High inoculative releases of N. cucumeris early in the flowering cycle followed by frequent low introductions of predators should provide a strong basis for preventative control of F. occidentalis and other thrips species on cyclamen.     

东亚小花蝽对西方花蓟马和二斑叶螨的捕食选择性

多食性天敌对猎物的捕食选择性及猎物密度对天敌捕食的干扰作用直接影响到天敌对不同猎物的控制作用.在实验室条件下研究了东亚小花蝽对西方花蓟马和/或二斑叶螨不同虫态的捕食选择性,以及两猎物中一种猎物的密度变化对小花蝽取食另一种猎物的影响.结果如下:东亚小花蝽5龄若虫和成虫对西方花蓟马2龄若虫的捕食选择性均强于对其成虫,对二斑叶螨的选择性为雌成螨>若螨>幼螨.实验中西方花蓟马2龄若虫是东亚小花蝽最喜好的虫态.二斑叶螨雌成螨密度固定为60头/19.63cm2,西方花蓟马若虫密度从10增加到60时显著地减少了东亚小花蝽对二斑叶螨的取食.反之,固定西方花蓟马同样密度,增加二斑叶螨密度却没有显著改变小花蝽对西方花蓟马的取食.此结果进一步表明,西方花蓟马是东亚小花蝽更喜好的猎物.     

In vivo rearing of Thripinema nicklewoodi (Tylenchida: Allantonematidae) and prospects as a biological control agent of Frankliniella occidentalis (Thysanoptera: Thripidae)

Methods are described for the in vivo production of the nematode Thripinema nicklewoodi (Siddiqi), an obligate parasite and potential biological control agent of western flower thrips Frankliniella occidentalis (Pergande). Nematode infection is not lethal but causes sterilization of adult female hosts. Both fertilization and horizontal transmission of T. nicklewoodi is achieved in 1.5-ml microcentrifuge tubes (infection arenas), in the presence of 100% humidity, a temporary food source and preferably a damp substrate. Following exposure to infection arenas, F. occidentalis are reared on excised bean leaves Phaseolus vulgaris (L.) in polypropylene containers for 2 wk at 25 degrees C to allow the reproduction and development of a single generation of nematodes within infected hosts's abdominal cavity. To identify infected hosts after this incubation period, thrips are isolated in microcentrifuge tubes and monitored for free-living nematodes being released along with frass. Infected thrips are reintroduced back into infection arenas to inoculate further thrips to maintain the culture. We documented the output of the rearing procedure using a standard method and following simple manipulation of several individual parameters of the infection technique. The standard method was the most efficient, and resulted in an increased (output/input) ratio of infected thrips of approximately 2; i.e., the number of infected thrips approximately doubles each generation. Monitoring infected thrips revealed that nematodes were first released between 12-14 d postinfection and for an average of 7.9 d at 25 degrees C; highlighting the potential to reuse infective thrips between infection arenas. The possibility of using T. nicklewoodi as an inoculative agent against F. occidentalis infesting floricultural crops is discussed.     

Predator-prey relationships on Apiaceae at an organic farm

Orius insidiosus (Say) and O. pumilio (Champion) were confirmed to be sympatric in north central Florida as the major predators of the Florida flower thrips, Frankliniella bispinosa (Morgan), on flowers of Queen Anne's lace, Daucus carota L. and false Queen Anne's lace, Ammi majus L. F. bispinosa was the predominant thrips observed on both flowers but colonized D. carota to a greater extent and earlier in the season than A. majus. Despite differences in the abundance of F. bispinosa on the two plants, neither Orius species showed host plant affinities. Population profiles for the thrips and Orius spp. followed a density dependent response of prey to predator with a large initial prey population followed by a rapid decline as the predator populations increased. The temporal increases in Orius spp. populations during the flowering season suggest that they were based on reproductive activity. As observed in a previous study, O. insidiosus had a larger population than O. pumilio and also had a predominantly male population on the flowers. By examining carcasses of the prey, there appeared to be no sexual preference of the thrips as prey by the Orius spp. as the prey pattern followed the demographics of the thrips sex ratio. Few immatures of either thrips or Orius spp. were observed on D. carota or A. majus, which suggests that oviposition and nymphal development occurred elsewhere. Based on these findings, D. carota and A. majus could serve as a banker plant system for Orius spp.     

Rapid estimation of numbers of whiteflies (Hemiptera: Aleurodidae) and thrips (Thysanoptera: Thripidae) on sticky traps

The presence or absence of greenhouse whiteflies, Trialeurodes vaporariorum Westwood, and thrips, primarily western flower thrips, Frankliniella occidentalis (Pergande), in cells of a grid laid over 7.6 cm by 12.7 cm sticky traps was used to estimate the population density of these pests on the trap. The method accurately predicted trap population densities of between 15 and 192 individuals per side for thrips on blue and yellow traps and between 15 and 168 whiteflies per side on yellow traps. The distribution of both whiteflies and thrips tended to be clustered on the sides and upper edge of the traps. The method is useful in giving a far more rapid estimate than counting individuals, particularly at high population densities.     

Overwintering Biology of the Predatory Bug Orius albidipennis (Hemiptera: Anthocoridae) in Israel

The predatory bug Orius albidipennis (Hemiptera: Anthocoridae) overwinters in Israel as adults. Oviposition continues irrespective of day length, but stops almost completely when the mean temperature drops below 15 C (beginning in December). During winter months, the outdoor females enter reproductive diapause, and resume oviposition in during March and April. All first instar nymphs die at low temperatures, whereas approximately 20% of the nymphs of later instars may survive and mature. Dense populations of the western flower thrips, Frankliniella occidentalis, develop outdoors in Israel about a month before the peak in population of O. albidipennis, and have the opportunity to cause much damage. Since the Israeli O. albidipennis does not enter diapause under short day photophase, it might be an efficient predator of F. occidentalis in greenhouses in countries of northern latitude.     

Management of western flower thrips, Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae), in hydroponic strawberry crops: using yellow sticky traps to determine action thresholds

Abstract  The relationship between western flower thrips (WFT) populations in flowers and catches on yellow sticky traps was investigated in a hydroponic strawberry crop in the Sydney area in 1999–2000. The thrips population was initially dominated by males, but by mid-summer it was composed primarily of females. At a point when approximately 60–65% of WFT on traps were female (approximately 20–30 females per trap), the overall density of thrips on traps and in the crop increased rapidly, resulting in severe damage to immature fruit. The sex ratio of 0.60–0.65 females corresponded very closely to a mean of five adult WFT per flower at the start of the trapping week, which was an infestation level previously calculated as the action threshold to prevent damage to young green fruit.     

Efficacy of entomopathogenic nematode Steinernema feltiae (Rhabditida: Steinernematidae) as influenced by Frankliniella occidentalis (Thysanoptera: Thripidae) developmental stage and host plant stage

Entomopathogenic nematodes were investigated as an alternative biological control strategy for western flower thrips, Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae), in ornamental greenhouse crops, by using potted chrysanthemum as a model crop. The susceptibility of various life stages of F. occidentalis to different concentrations of the nematode Steinernema feltiae (Filipjev) (Rhabditida: Steinernematidae) was investigated in petri dish bioassays. This was followed with trials using potted chrysanthemums comparing the efficacy of nematode application to plants in vegetative (exposed habitat) versus flowering (cryptic habitat) stages. In both trials, the effect of the wetting agent Agral 90 (nonylphenoxy polyethoxyethanol), which is used in combination with the nematode spray, on F. occidentalis mortality was assessed. In petri dish trials, the prepupae and pupae were the most susceptible developmental stages of F. occidentalis to infection by S. feltiae. First and second instars were killed by very high rates of nematodes (> or =20,000 infective juveniles per milliliter), but corrected mortality was only approximately 28-37%. No significant mortality was observed for adult thrips. Results from the petri dish trials were confirmed on chrysanthemum plants. Foliar application of S. feltiae did not result in significant mortality in larvae or adults. No significant differences in efficacy were detected by application of nematodes on vegetative versus flowering chrysanthemum. Agral 90 had a significant impact on mortality on the first stage larvae and prepupae in the petri dish trials but not in the plant trials. Thrips control by S. feltiae in greenhouses may be partly or completely due to prepupal and pupal mortality.     

Effects of feeding frequency and sugar concentration on behavior and longevity of the adult aphid parasitoid: Aphidius ervi (Haliday) (Hymenoptera: Braconidae)

The effectiveness of the predatory mite, Phytoseiulus persimilis Athias-Henriot (Acari: Phytoseidae), as a suppressive agent of the twospotted spider mite, Tetranychus urticae Koch (Acari: Tetranychidae), was evaluated on greenhouse ivy geraniums at predator:prey release ratios of 1:60, 1:20, and 1:4. Releases at each predator:prey ratio were made at moderate and high T. urticae densities to determine if initial pest population size influenced the suppressive ability of the predator. At ratios of 1:4 and 1:20, P. persimilis significantly reduced T. urticae populations 1 week after release and kept them at low levels thereafter. Plant damage also was significantly reduced at these densities. After 4 weeks, the P. persimilis that were released at a ratio of 1:4 consistently reduced T. urticae populations from densities as high as 30 T. urticae per leaf to fewer than 0.6 per leaf. We found no interaction between release ratio and T. urticae density, indicating that predator effectiveness remains constant, at least within the range of T. urticae densities used. Our work demonstrates the potential of P. persimilis to provide effective control of T. urticae on a greenhouse-grown floricultural crop at a moderately low predator:prey ratio (1:20) and over a range of initial pest densities. However, we recommend that P. persimilis be released at a ratio of 1:4 for greatest reliability and successful control of T. urticae on ivy geraniums.     

Evaluation of the nematodes Steinernema feltiae and Thripinema nicklewoodi as biological control agents of western flower thrips Frankliniella occidentalis infesting chrysanthemum

In greenhouse studies, we evaluated a commercial formulation of the entomopathogenic nematode Steinernema feltiae and the inoculative release of the thrips-parasitic nematode Thripinema nicklewoodi against western flower thrips (WFT), Frankliniella occidentalis Pergande infesting potted chrysanthemums. Foliar sprays of S. feltiae applied at 1.25-2.5×10³ IJ mL^-1 and 1000 - 2000 L ha^-1 at 3-day intervals alone (targeting feeding stages) or in combination with soil applications (simultaneously treating non-feeding stages in the soil at the same rates) decreased but did not provide adequate control of thrips in flowering plants artificially infested with a dense population. Similar nematode treatments applied for four to five applications at 6-day intervals in two batches of initially clean chrysanthemums failed to prevent unacceptable damage to flowers and leaves from a dense natural infestation within the greenhouse. Although some IJ survived up to 48 h within flowers and flower buds, few nematode-infected thrips (larvae and adults) were recovered. In studies with T. nicklewoodi (which is not amenable for mass production), the inoculative releases of two parasitized hosts per plant enabled the nematode to become established within existing WFT populations under greenhouse conditions. However, relatively poor transmission and slow speed of kill (nematode primarily suppresses populations through host sterilization) prevented low level inoculations being effective over a single crop cycle. Further studies showed that transmission of T. nicklewoodi persisted for nine host generations, infected up to 83% of adult thrips and provided long-term suppression of discrete caged populations, but only after uneconomically high thrips densities had been reached.     

Do functional responses of predatory arthropods reach a plateau? A case study of Orius insidiosus with western flower thrips as prey

Parameters determining the functional and numerical response of Orius insidiosus to the density of second instar larvae of the western flower thrips (WFT), Frankliniella occidentalis, were estimated with the aim to be used in models calculating adequate release ratios of predators vs prey. Especially when the prey population must decrease immediately after predator release and the time to suppression must be short, it is of crucial importance to know whether the functional response reaches a plateau or keeps on increasing with prey density within the range of densities that are realistic for the prey. Such plateaus may arise from constraints on the time budget or from constraints on gut-fullness-associated motivation to attack the prey. Estimates of the plateau for O. insidiosus based on prey handling times by far exceed realistic values. In addition, motivation constraints did not apply, because the level of gut fullness above which attack ceases coincides with gut capacity. This implies that the predation rate keeps on increasing linearly with the square root of the density of WFT-larvae and does not reach a plateau within the realistic range of thrips densities (0–20 WFT per cm2). Such plateaus are likely to occur for smaller-sized predators and smaller-sized stages of the same predator and they may also occur when the prey stage offered has better capacities to escape or resist attack. We argue that the presence and level of plateaus in functional responses are of importance for determining initial predator-to-prey ratios.     

Variation in preference and performance of Frankliniella occidentalis (Thysanoptera: Thripidae) on three strawberry cultivars

Western flower thrips, Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae), is a major pest of strawberry, causing substantial yield loss through direct feeding on the flowers and fruit. Insecticides are the main method used for its control; however, F. occidentalis has developed resistance to insecticides from all major chemical classes. Chemical control is not a long-term strategy and integrated pest management is required. This study determined whether F. occidentalis damage could be reduced by host plant resistance or tolerance in three commercial strawberry cultivars (Fragaria X ananassa [Rosaceae]: 'Albion', 'Camarosa', and 'Camino Real'). Determination of resistance or tolerance to F. occidentalis was based on olfactory response, feeding damage, ovipositional preference, and host suitability for reproduction on leaves. F. occidentalis adults preferred to feed on Camarosa; however, if leaves had been fed on previously by conspecifics, there was no difference in feeding preference. Camarosa was the most preferred cultivar for oviposition, and more eggs were laid by F. occidentalis on Camarosa than either Albion or Camino Real. More larvae hatched and adults were reared from Camarosa than either Albion or Camino Real. The percentage of unhatched eggs, larvae, and pupae that died was highest on Camino Real. Survival rate was highest on Camarosa. Egg incubation, prepupation, pupation, and total developmental periods were shortest on Camarosa, but the larval period was longest on Camarosa. Camarosa was the most favorable cultivar for F. occidentalis population growth on leaves.     

Temperature-mediated relationship between western flower thrips (Thysanoptera: Thripidae) and chrysanthemum

Modeling the effect of temperature on the sustainability of insect-plant interactions requires assessment of both insect and plant performance. We examined the effect of temperature on western flower thrips, Frankliniella occidentalis (Pergande), a generalist herbivore with a high reproductive rate, and chrysanthemum inflorescences, a high quality but relatively fixed, ephemeral resource for thrips population growth. We hypothesized that different thrips versus plant responses to temperature result in significant statistical interaction of temperature with thrips abundance and flower damage attributes over time. Experiments were conducted at five temperatures between 20.7 and 35.3 degrees C, with thrips infestation and time after infestation as main effects. Only minor, uncontrolled variations in relative humidity and light intensity may otherwise have influenced the results. High temperatures lead to an initially rapid increase in density of thrips followed by abrupt declines in abundance. The rate of floral senescence increased with temperature and thrips infestation, as indicated by a reduced fresh biomass and greater leaching of yellow pigments. Multiple regression indicated that indices of plant damage responded more directly to thrips density at low than high temperature, supporting the conclusion that temperature affected the outcome beyond what was predictable simply from differential plant and insect optima. The relative intensity of damage caused by individual thrips decreased with increasing temperature, likely caused by thrips competition and reduced survival, growth, and fecundity on depleted inflorescences. Reduced per capita damage at high temperature may be common in insects exploiting fixed plant resources that exhibit an accelerated rate of deterioration at high temperatures.