Botanical insecticides in controlling Kelly's citrus thrips (Thysanoptera: Thripidae) on organic grapefruits

Kelly's citrus thrips, Pezothrips kellyanus (Bagnall) (Thysanoptera: Thripidae) was first recorded in Cyprus in 1996 and became an economic citrus pest. In Cyprus, Kelly's citrus thrips larvae cause feeding damage mainly on immature lemon and grapefruit fruits. Use of botanical insecticides is considered an alternative tool compared with synthetic chemicals, in offering solutions for healthy and sustainable citrus production. During 2008-2010, the efficacy of the botanical insecticides azadirachtin (Neemex 0.3%W/W and Oikos 10 EC), garlic extract (Alsa), and pyrethrins (Vioryl 5%SC) was evaluated in field trials against Kelly's citrus thrips larval stage I and II aiming at controlling the pest's population and damage to organic grapefruit fruits. In each of the trial years treatments with pyrethrins and azadirachtin (Neemex 0.3%W/W) were the most effective against Kelly's citrus thrips compared with the untreated control (for 2008: P < 0.018; for 2009: P < 0.000; for 2010: P < 0.008). In 2008, the mean number of damaged fruits in treatments with pyrethrins and Neemex was 9.6 (19.2%) and 9.7 (19.5%) respectively, compared with 12.2 (24.3%) in the untreated control. In 2009, the mean number of damaged fruits in treatment with pyrethrins was 3.7 (7.3%) and 3.9 (7.8%) in treatment with Neemex compared with 8.6 (17.3%) in the untreated control, while in 2010 the mean damaged fruits in these treatments was recorded at 18.7 (37.5%) and 19.6 (39.2), respectively, compared with 29.6 fruits (59.2%) in the control. Oikos 10 EC showed significant effect only in 2009 and 2010. In these years, the mean number of damaged fruits was recorded at 5.5 and 21.2 compared with 8.6 and 29.6 fruits in the untreated control, respectively. Garlic extract showed the lowest effect from all the botanicals used compared with the untreated control.     

Euseius addoensis addoensis,an effective predator of citrus thrips,Scirtothrips aurantii,in the eastern Cape Province of South Africa

The predacious mite,Euseius addoensis addoensis (McMurtry) (Acari: Phytoseiidae), significantly (P<0.05) reduced scarring on Palmer Navel oranges,Citrus sinensis (L.) Osbeck, caused by the citrus thrips,Scirtothrips aurantii Faure (Thysanoptera: Thripidae). A chemical exclusion method employing dicofol was used to determine the effect of the predator and numbers ofS. aurantii were monitored on sticky yellow traps. At two sites where numbers of citrus thrips and predacious mites were high, the percentage fruit culled from export quality due to damage byS. aurantii was reduced from 25.5 to 15.5% and from 32.7 to 18.7% by predacious mite activity.Euseius a. addoensis sometimes prevented damage from exceeding 1% cull and was most effective when mean numbers exceeded one mite per inside leaf at petal fall. The best correlation between numbers of mites and the percentage fruit culled at harvest was –0.96 at the location with the highest number of citrus thrips. Tartar emetic (0.398% a.i.) plus sugar (0.4% a.i.) treatments for citrus thrips were slightly detrimental toE. a. addoensis populations but their numbers soon recovered. An alternative treatment, methiocarb 0.016% a.i. plus sugar (0.4% a.i.), almost eliminated the predacious mites and resulted in more citrus thrips damage (10.7% cull) than in the adjacent untreated control (0.2% cull) which contained predacious mites.     

Distributions of western flower thrips (Thysanoptera: Thripidae) and its predatory bug Orius niger (Hemiptera: Anthocoridae) assessed by coloured sticky traps and plant samplings in cotton

The capturing efficiency of four coloured (yellow, green, white and blue) sticky traps, placed at the top, middle and bottom strata of cotton plants, was tested for the western flower thrips, Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae), and its predatory bug, Orius niger (Wolff) (Hemiptera: Anthocoridae), as well as spatial distributions of both insects on the plant in years 2006 and 2007. The white coloured trap was the most attractive to F. occidentalis and O. niger in the 2-year study. The blue coloured trap was the least attractive for Orius. The mean numbers of F. occidentalis and O. niger on the top plant parts (flowers and leaves) and in all coloured traps positioned on the upper parts of the plants were greater than those found in the lower two strata. Taylor's power law analysis showed that F. occidentalis and O. niger adults aggregated in the flowers or on the leaves. This study suggests that top flowers could be preferentially sampled to determine population densities of Frankliniella flower thrips and Orius species in cotton, and thus, sticky traps should be placed on the top level of plants. F. occidentalis: O. niger ratios in the flowers varied from 4 to 60 thrips per Orius adult in the three plant strata. This result may indicate that F. occidentalis experiences more predation from Orius.     

The discovery of a new species of Ceranisus from southwestern Turkey

An explorative trip was made to southwesternTurkey in April 1995 in search of pear thrips,Taeniothrips inconsequens (Uzel)(Thysanoptera: Thripidae), and its naturalenemies. Forty-three females and 13 males of apreviously undescribed species of the genusCeranisus Walker (Hymenoptera:Eulophidae) were found, mainly in floweringArbutus andrachne L. Although theparasitoid was found in association withseveral thrips species, but especially T.inconsequens adults and larvae, its host orhosts were not determined. The parasitoid isdescribed and given the name Ceranisusantalyacus S. Triapitsyn.     

Population dynamics and integrated pest management of Thrips tabaci on leek under field conditions in northwest Italy

Thrips tabaci Lindeman (Thysanoptera: Thripidae) is a major pest of leek, Allium porrum L. (Alliaceae), in Piedmont, northwest Italy, and to control its infestation the leek crop is sprayed intensively with insecticides during the summer period. In order to find the most efficient and environment‐friendly method of thrips control, research was conducted on six commercial farms during 2005–2006 to assess thrips population composition and infestation levels, and in an experimental field during 2005–2007. Biological and chemical control were compared during 2005–2006, whereas integrated pest management was adopted during 2007. During the growing season, thrips and natural enemy populations were monitored at 14‐day intervals by beating plants; new leaves of plants were also visually inspected for thrips‐feeding symptoms. Furthermore, in the experimental field at harvest‐time, the level of thrips injury to plants was assigned to one of five classes, depending on the percentage of leaf area damaged. Over 99% of phytophagous adult thrips found were male and female T. tabaci. Infestations were very variable in the crops surveyed, partly due to broad‐spectrum chemical treatments against the leek pests, which often failed to control thrips. In general, populations peaked in September, when they reached the maximum mean values ranging between 1.7 and 33.1 thrips per plant. At harvest, none of the surveyed farms experienced quality losses due to thrips injuries. In the experimental field during 2005–2006, the mean number of thrips per plant was greater in the chemical than in the biological control treatment, even though damage indices showed no significant differences between the two treatments. Predatory thrips of the genus Aeolothrips (Thysanoptera: Thripidae) and predatory bugs of the genus Orius (Heteroptera: Anthocoridae), mostly Orius majusculus Reuter, were particularly abundant during 2007, supporting the importance of management with selective insecticides to encourage natural colonization by predators.     

Influences of fertilization on population abundance, distribution, and control of Frankliniella occidentalis on chrysanthemum

We examined the effects of fertilization on population abundance and within‐plant distribution of western flower thrips, Frankliniella occidentalis Pergande (Thysanoptera: Thripidae), on potted chrysanthemum, Dendranthema grandiflora (Tzvelev). We also investigated the effects of fertilization on the number of insecticide applications needed to control F. occidentalis on potted chrysanthemum. Under greenhouse conditions, rate of change in population abundance of F. occidentalis increased with fertilization levels from 0 to 100% of the standard fertilization level (375 ppm N) and was four times higher on plants fertilized with the standard level (rate of change = 0.14) than on plants fertilized with 0% during the first 4 weeks after thrips inoculation. Within‐plant distribution of F. occidentalis was influenced by the phenology of the plants rather than total nitrogen content of plant tissues. Prior to flower opening, more F. occidentalis were found in the middle region of the plants. When the flowers began to open, more thrips were found feeding inside the flowers than on the leaves. We further showed that production time, the time from transplantation to flower opening, shortened considerably with increased fertilization level. Production time was shortest, 12 weeks, for plants fertilized with 100% of the standard fertilization level. When the fertilization level was reduced to 20%, production time lengthened to 13 weeks. When fertilization level was reduced to 0%, production time lengthened to 14 weeks. Increased fertilization from 0 to 100% of the standard level did not result in higher numbers of insecticide applications. All three insecticides (acephate, bifenthrin, and spinosad) were effective in keeping the thrips infestation below a predetermined level, five thrips per plant, but bifenthrin required the most number of applications to do so. For chrysanthemum, a fast‐growing crop and heavy utilizer of fertilizer, fertilization influenced not only the population growth of pest insects but also plant production time. As a result, optimizing fertilization level to reduce pest population growth may be a useful tactic in an Integrated Pest Management program for managing F. occidentalis on potted chrysanthemum. However, the effect of fertilization on production time and plant quality should also be considered when implementing this tactic.     

Flower model traps reduced thrips infestations on a pepper crop in field

A flower model trap developed by modifying an artificial yellow chrysanthemum flower was more attractive to flower thrips than commercial yellow sticky traps. Installation of these flower model traps (20 traps per 50 m² plot) was reported to reduce seasonal populations of Frankliniella intonsa (Trybom) (Thysanoptera: Thripidae) on strawberry flowers in greenhouses. In this study, we sought to determine if the installation of such flower model traps would reduce thrips populations in a pepper field. The traps were installed at the bottom of the plant canopy at varying densities (0, 5, 10, and 20 traps) in 20 plots (each 3 × 5 m²) using a completely randomized design. Thrips populations on pepper flowers were sampled from 1 to 29 July in 2009. All thrips sampled on the flowers were identified as F. intonsa. A significant effect of treatment and sampling date was found from repeated-measure analysis of variance. The highest density (20 traps per 15 m²) of traps significantly reduced the female and male F. intonsa population compared to the control by 61 and 49%, respectively. However, no difference in immature thrips numbers was found among the treatments. These results indicate that this flower model trap can be a useful tool for the management of flower thrips on field-grown peppers.     

Decreased preference and reproduction, and increased mortality of Frankliniella occidentalis on thrips-resistant pepper plants

The effect of thrips resistance in pepper (Capsicum annuum L.) – previously shown to result in impeded thrips population development (Maris PC, Joosten NN, Goldbach RW & Peters D (2003a) Restricted spread of Tomato spotted wilt virus in thrips‐resistant pepper. Phytopathology 93: 1223–1227. Maris PC, Joosten NN, Goldbach RW & Peters D (2003b) Spread of Tomato spotted wilt virus and population development of Frankliniella occidentalis in pepper resistant to thrips. Proceedings of the Section Experimental and Applied Entomology Netherlands Entomological Society (NEV) Amsterdam 14: 95–101.) – on thrips’[Frankliniella occidentalis Pergande (Thysanoptera: Thripidae)] reproduction, mortality, host preference, and behaviour was investigated. Reproduction, studied by oviposition and larval survival, was negatively affected by the thrips‐resistant (TR) phenotype, whereas the offspring's developmental rate did not differ on TR and the thrips‐susceptible (TS) phenotype. While thrips’ behaviour was hardly affected by thrips resistance, a significant preference for TS plants over the TR plants was found in different tests. When released on either a TR or a TS plant, thrips dispersed at significantly higher rates from the TR plants, demonstrating that not only an impeded reproduction, but also a reduced residence time adds to the reported lower thrips numbers on TR plants.     

Control of the broad mite (Polyphagotarsonemus latus (Banks)) on organic greenhouse sweet peppers (Capsicum annuum L.) with the predatory mite, Neoseiulus cucumeris (Oudemans)

The predatory mite Neoseiulus cucumeris (Oudemans) (Acarina: Phytoseiidae) successfully controlled the broad mite Polyphagotarsonemus latus (Banks) (Acarina: Tarsonemidae) on two varieties of greenhouse-grown sweet peppers (Capsicum annuum L.). A survey of pre-plant seedlings showed that nurseries were a source of infestation for the broad mite. The predatory mites were released twice (on day 1 and 5, or 15 days later) on each plant, every second plant or every fourth plant. Broad mite populations were evaluated by sampling young leaves from the top of the plant. The effect of the broad mite on plant height, dry mass and yield was evaluated. Additionally, since N. cucumeris is known to control thrips, Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae), blue sticky traps and flower sampling were used to evaluate changes in thrips populations. All three release rates of N. cucumeris significantly (P<0.05) controlled broad mite populations, but when the predatory mites were released only on every fourth plant, the overall height and yield of the plants were adversely affected by broad mites. Releasing N. cucumeris on each or every second plant was as efficacious in controlling broad mites as sulfur treatments in terms of plant height, dry mass and yield. Plants treated with sulfur, however, had significantly higher thrips populations and fruit damage.     

Can Mass Trapping Reduce Thrips Damage and Is It Economically Viable? Management of the Western Flower Thrips in Strawberry

The western flower thrips Frankliniella occidentalis (Pergande) (Thysanoptera: Thripidae) is a cosmopolitan, polyphagous insect pest that causes bronzing to fruit of strawberry (Fragaria x ananassa). The main aim of this study was to test whether mass trapping could reduce damage and to predict whether this approach would be economically viable. In semi-protected strawberry crops, mass trapping of F. occidentalis using blue sticky roller traps reduced adult thrips numbers per flower by 61% and fruit bronzing by 55%. The addition of the F. occidentalis aggregation pheromone, neryl (S)-2-methylbutanoate, to the traps doubled the trap catch, reduced adult thrips numbers per flower by 73% and fruit bronzing by 68%. The factors affecting trapping efficiency through the season are discussed. Damage that would result in downgrading of fruit to a cheaper price occurred when bronzing affected about 10% of the red fruit surface. Cost-benefit analysis using this threshold showed that mass trapping of thrips using blue sticky roller traps can be cost-effective in high-value crops. The addition of blue sticky roller traps to an integrated pest management programme maintained thrips numbers below the damage threshold and increased grower returns by a conservative estimate of £2.2k per hectare. Further work is required to develop the F. occidentalis aggregation pheromone for mass trapping and to determine the best timing for trap deployment. Mass trapping of thrips is likely to be cost-effective in other countries and other high-value crops affected by F. occidentalis damage, such as cucumber and cut flowers.     

Spatiotemporal population dynamics of the banana rind thrips,Elixothrips brevisetis (Bagnall) (Thysanoptera: Thripidae)

The understanding of how environmental factors and agricultural practices affect population dynamics of insect pests is necessary for pest management. Here, we provide insight into the ecology of the banana rind thrips Elixothrips brevisetis (Bagnall) (Thysanoptera: Thripidae) by collecting and analysing a spatiotemporal database of population estimates in Martinique (West French Indies). We assessed the influence of climatic variables (which were rainfall and temperature) and biotic variables (which were banana and three weed species) on the adult thrips abundance for different components of the banana plant (sucker, mother plant and bunch) and evaluated the effect of thrips abundance and standard bunch covers on damages. The abundance of thrips on the sucker, the mother plant, and the bunch was significantly related to the abundance on neighbouring banana plants, and spatial autocorrelation indicated that E. brevisetis dispersed for only short distances. The number of thrips on the mother plant and on the bunch was positively related to the number of thrips on the sucker, suggesting that the thrips may disperse from the sucker to the mother plant and then to the bunch. The abundance of thrips on the sucker increased with sucker height and was positively correlated with the mean daily rainfall during the 17 days before sampling; the length of that period might correspond with the time required for an individual to complete its life cycle. Covered bunches had 98% fewer thrips than non‐covered bunches, and the damage caused by thrips was linearly related to the number of thrips present between the 2nd and 4th week after flowering. Finally, we found that the presence of Alocasia cucullata, Dieffenbachia seguine and Peperomia pellucida is significantly related with a decrease in thrips abundance on banana plants, suggesting the use of these weeds as potential trap plants.     

Distribution ofGeotrichum candidum citrus race in citrus groves and non-citrus fields in Japan

The presence ofGeotrichum candidum citrus race, the citrus sour rot pathogen, was examined in soils of citrus groves and non-citrus fields of Japan. Soil samples were collected from 223 sites (118 sites in citrus groves, and 105 sites in fields cultivated with 33 species of non-citrus plants and in evergreen broad-leaved forest) in 11 main citrus-growing prefectures, and Hokkaido, a non-citrus-growing area. Of 236 soil samples from citrus groves, 95.76% containedG. candidum citrus race and 0.42% contained the non-citrus race; while of 210 samples from non-citrus fields, 62.85% and 4.76% contained the citrus race and the non-citrus race respectively. All of the citrus race isolates obtained either from citrus groves or non-citrus fields were pathogenic on lemon (Citrus limon) and satsuma mandarin (Citrus unshiu), but some of these isolates failed to infect orange (Citrus sinensis). The non-citrus races were pathogenic on ripe tomato fruit (Lycopersicon esculentum) and ripe muskmelon fruit (Cucumis melo var.reticulatus). Results indicated that citrus sour rot pathogen is widely distributed in citrus groves and non-citrus fields of diverse plant species in Japan.     

Decision-making tools for Frankliniella occidentalis management in strawberry: consideration of target markets

The western flower thrips (WFT), Frankliniella occidentalis Pergande (Thysanoptera: Thripidae), a cosmopolitan pest of many crops, is considered a major pest of low tunnel and greenhouse strawberries. The extent of damage to strawberry is unclear because different studies have produced contradictory results. Also, economic thresholds published for WFT in strawberry vary greatly, and most fail to incorporate economic factors. This study was aimed at developing a decision‐making tool for WFT management in strawberries in Israel. Toward this end, economic injury levels (EIL) and economic thresholds were calculated, based on target markets (export vs. domestic). Results indicate that serious infestation of ripe berries may cause a dull, rough appearance, and the fruit may be soft and have a reduced shelf life, rendering it unsuitable for export. Most fruit damage occurred at green and turning‐red stages of development. Two decision‐making tools were developed, one for winter, when WFT populations increase slowly but crop value is high (export market); and the second for spring, when the pest increases rapidly but crop value is low (local markets). Economic thresholds of 10 and 24 WFT/flower were calculated for winter and spring strawberries, respectively, based on direct thrips damage to fruit. This calculation does not take into account the recorded WFT damage to flowers, or its role in facilitating Botrytis cinerea fruit infection. Western flower thrips has proved only an occasional economic pest in Israeli strawberries, and no routine control measures are warranted. Furthermore, augmentative releases of Orius laevigatus or Neoseilus cucumeris against WFT are not justified in this system, because Orius colonizes strawberry fields spontaneously in high numbers when no broad spectrum insecticides are used.     

Spatial dependence, dispersion, and sequential sampling of Anaphothrips obscurus (Thysanoptera: Thripidae) in timothy

The spatial distribution and dispersion of Anaphothrips obscurus (Müller) (Thysanoptera: Thripidae) was examined with the goal of establishing a sequential sampling plan for this pest in timothy, Phleum pratense L. (Poaceae). Approximately 16 different California timothy fields were sampled twice yearly from 2006 to 2008 using direct observation and the beat cup method. For direct observation, the number of thrips on each leaf of the plant was counted. For the beat cup method, tillers were tapped into a cup and dislodged thrips were counted. Samples were separated by ≈3 m in 2006 and 2007 and exactly 3 m in 2008. Spatial autocorrelation of intrafield population distribution was tested for significance in 2008 using Moran's I, but autocorrelation was not detected. The population dispersion was assessed by Taylor's power law and was determined to be aggregated and density-dependent. Intraplant population dispersion and distribution for each year were also evaluated for adults, larvae, and total thrips. All lifestages were highly spatially dependent and more thrips were found near the top of the plant than the bottom. Direct observation proved to be a more accurate and precise method than the beat cup method, especially when thrips abundances were greater than one. However, the number of samples required to provide an accurate level of precision was unrealistic for both methods. A sequential sampling plan was evaluated, but was not practical for the beat cup method because few thrips were found using this method. Because there was no spatial autocorrelation at sampling distances of 3 m, samples can be taken at intervals at 3 m to obtain spatially independent population abundance estimates.     

Thrips Cross-Pollination of Popowia pisocarpa (Annonaceae) in a Lowland Dipterocarp Forest in Sarawak1

This paper reports thrips (Thysanoptera) cross-pollination in the primitive angiosperm, Popowia pisocarpa (Annonaceae), in lowland rain forest of Sarawak, Malaysia. Flowers of P. pisocarpa have a tiny pollination chamber (3–4 mm in depth) with the entrance almost closed by the disk-shaped stigmatic heads, allowing only small insects to enter. Experiments showed that thrips were effective pollinators and flowers ofP. pisocarpa were self-incompatible. Seed set was limited by pollen. Clumped adult trees had higher seed-set than more distant individuals (>5 m from nearest flowering neighbor), indicating limited inter-tree movement by thrips.     

Combined releases of entomopathogenic nematodes and the predatory mite Hypoaspis aculeifer to control soil-dwelling stages of western flower thrips Frankliniella occidentalis

The effect of the predatory miteHypoaspis aculeifer Canestrini (Acarina:Laelapidae) on soil-dwelling stages of thewestern flower thrips (WFT) Frankliniellaoccidentalis Pergande (Thysanoptera: Thripidae)and the influence of combined releases of H.aculeifer and two entomopathogenic nematodes(EPNs) Heterorhabditis bacteriophora Poinar(Rhabditida: Heterorhabditidae) (strain HK3,HK3) and Steinernema feltiae Filipjev(Rhabditida: Steinernematidae) (Nemaplus®,SFN) were investigated in pot trials usingseedlings of green beans (Phaseolus vulgarisL.). Ten H. aculeifer adults per pot and 400infective juveniles (IJs) cm^–2 soil, of the twoEPN strains were used. In comparison withuntreated control, H. aculeifer reduced theproportion of adult F. occidentalis emergenceby 46%, while SFN and HK3 led to a reductionin adult thrips emergence by 46% and 61%,respectively. Significant differences in adultWFT emergence were found between combinedtreatments of EPNs and H. aculeifer, andindividual applications of EPNs and/or H.aculeifer, with significantly lower adultthrips emergence in the combined treatments.These findings highlight the potential for acombined use of EPNs with H. aculeifer for thecontrol of soil-dwelling stages of thrips.     

Modern thrips families Thripidae and Phlaeothripidae in Early Cretaceous amber (Insecta: Thysanoptera)

Two specimens of Thysanoptera with forked sensilla on third and fourth antennal segments were described from the Lebanese Neocomian and the Spanish Albian ambers, and attributed to the new genus Tethysthrips n. gen. in the family Thripidae Stevens 1829. One specimen with a tubular tenth abdominal segment was also discovered in the Lebanese Neocomian amber, and attributed to the new genus Rohrthrips n. gen. belonging to the family Phlaeothripidae Uzel 1895. Thripidae and Phlaeothripidae are nowadays the most species-rich families of Thysanoptera. The present discoveries of Early Cretaceous fossils show how diversified these families and thrips already were at that time. Moreover, this tubuliferan Rohrthrips specimen has plesiomorphies no longer present in the recent genera, in particular on the wings. Therefore it brings new insight in the evolution of Tubulifera.     

Independent cytoplasmic incompatibility induced by Cardinium and Wolbachia maintains endosymbiont coinfections in haplodiploid thrips populations

Cardinium and Wolbachia are common maternally inherited reproductive parasites that can coinfect arthropods, yet interactions between both bacterial endosymbionts are rarely studied. For the first time, we report their independent expression of complete cytoplasmic incompatibility (CI) in a coinfected host, and CI in a species of the haplodiploid insect order Thysanoptera. In Pezothrips kellyanus, Cardinium‐induced CI resulted in a combination of male development (MD) and embryonic female mortality (FM) of fertilized eggs. In contrast, Wolbachia‐induced CI resulted in FM together with postembryonic mortality not previously reported as a CI outcome. Both endosymbionts appeared to not influence fecundity but virgins produced more offspring than mated females. In coinfected individuals, Wolbachia density was higher than Cardinium. Wolbachia removal did not impact Cardinium density, suggesting a lack of competition within hosts. Maternal transmission was complete for Wolbachia and high for Cardinium. Our data support theoretical predictions and empirical detection of high endosymbiont prevalence in field populations of the native range of this pest thrips. However, previous findings of more frequent loss of Wolbachia than Cardinium, particularly in field populations of the host's invasive range, suggest that genetic diversity or varying environmental factors between field populations also play a role in shaping host‐endosymbiont dynamics.     

Effect of maize plants on colonisation of cowpea plants by bean flower thrips,Megalurothrips sjostedti

The effect of non-host maize plants on colonisation of cowpea byMegalurothrips sjostedti (Trybom) (Thysanoptera; Thripidae) was investigated. There were no differences in population density and activity ofM. sjostedti on sole cowpea crop and mixed cowpea/maize crop during the colonisation phase (i.e. 10–30 days after emergence of the plants. However, subsequentlyM. sjostedti numbers were lower in the mixed than in the sole crop, suggesting that maize did not interfere with colonisation of cowpea crop by thrips. In a choice situation, higher numbers ofM. sjostedti oriented towards, and settled on, sole cowpea plants than on those mixed with maize. Olfactory tests indicated that fewer thrips oriented towards a cowpea/maize mixed odour source. When equal numbers of thrips were introduced into the centre of sole- and mixed-cropped cowpea plots, the thrips became randomly distributed in each plot. Fewer thrips were recovered from the mixture than from the sole crop. It is concluded that, although non-host plant odours do not reduce thrips colonisation they interfere with host plant utilisation.