A series of recent studies have investigated the development, cold tolerance and winter field survival of glasshouse biological control agents that are non-native to the UK: Neoseiulus californicus, Macrolophus caliginosus, Delphastus catalinae, Eretmocerus eremicus and Typhlodromips montdorensis. The combined data sets for these species have been analysed to identify laboratory indices of cold tolerance that correlate with the duration of survival under winter field conditions. A strong correlation was found between the laboratory LTime 50 at 5°C and maximum field survival time. Both N. californicus and M. caliginosus survived substantially longer in the laboratory and in the field than the other species. It is suggested that the LTime 50 at 5°C may provide a reliable 'first stage' screen of establishment potential for candidate non-native biological control agents in the UK.

The recent unexpected local establishment of a non-native predatory mite, Neoseiulus californicus (McGregor) (Acari: Phytoseiidae), in the UK prompted us to undertake this study, which investigated the thermal biology of an alien species Typhlodromips montdorensis (Schicha) (Acari: Phytoseiidae). Laboratory and field experiments on its cold tolerance were used to assess its establishment potential outside of glasshouse environments in the UK. Currently, T. montdorensis is being tested as a glasshouse biological control agent against thrips and spider mites, but is not yet licensed for release in the UK. Typhlodromips montdorensis has a developmental threshold of between 10.3 and 10.7 ° C, and a thermal budget of between 108.7 and 105.3 degree-days when estimated by weighted and simple linear regression, respectively. Under outdoor conditions, T. montdorensis could theoretically complete up to six generations a year. The supercooling points of female and larval T. montdorensis were − 22 to − 24 ° C with 100% pre-freeze mortality apparent in both acute and chronic low temperature exposures. Typhlodromips montdorensis were unable to enter diapause under a selected laboratory regime. No reproduction occurred in the field from November to March, with 100% mortality within 7-14 days of release during this period. It is concluded that T. montdorensis would be a 'safe candidate' for introduction as a glasshouse biological control agent in the UK.

The ecological impact of introduced biological control agents on native species of arthropods is a matter of considerable debate. This study investigated the ability of the non-native predatory mite Neoseiulus californicus to feed on the native Typhlodromus pyri and vice versa, as both species now co-occur in UK orchards. Typhlodromips montdorensis is a candidate for introduction into the UK as a glasshouse biological control agent. The ability of T. montdorensis to feed on the widely used N. californicus was investigated to identify possible intraguild predation, which might influence the effectiveness of either or both species as predators of Tetranychus urticae. Both N. californicus and T. pyri consumed larval stages of each other, but in choice experiments both showed a preference for T. urticae. Both N. californicus and T. montdorensis also fed on each other, but whereas N. californicus again showed a preference for T. urticae, T. montdorensis fed equally on T. urticae and N. californicus. Interactions between N. californicus and T. pyri and N. californicus and T. montdorensis are discussed in relation to their effectiveness as biological control agents in the glasshouse and the natural control of spider mite in the field.

The predatory mirid Dicyphus hesperus Knight (Hemiptera: Miridae) is native to North America. The species has been used for the control of glasshouse whitefly on aubergine in the Netherlands, and is currently being evaluated for continued and wider release in Europe. Field and laboratory studies were conducted on a population collected from southern California, USA, to assess the cold tolerance and potential for outdoor establishment under prevailing northern European climates. The supercooling points (whole animal freezing temperatures) of nymphal and adult insects were around À208C. The lethal temperatures (LTemp 50 ) of non-diapausing nymphs and adults and diapausing adults were close to their respective freezing temperatures at À17.6, À17.6 and À19.28C. At 58C, the LTime 50 was 54, 101.7 and 117.5 days for fed nymphs, non-diapausing and diapausing adults respectively. When first instar nymphs were placed in the field in winter, starved samples died out after 70 days, but 5% of the fed nymphs survived until the end of winter (140 days) and developed to adult on return to the laboratory. After a similar 5-month field exposure, 50% of fed diapausing adults and 15% of fed non-diapausing adults were still alive at the end of winter, whereas starved diapausing adults died after 140 days. On return to the laboratory after 5 months in the field, both diapausing and non-diapausing adults mated and laid eggs, forming viable populations. Overall, the field and laboratory experiments indicate that this population of D. hesperus is able to enter diapause and that winter temperatures are not a barrier to establishment in northern Europe.

The outdoor establishment of non-native biocontrol agents released for inundative control of glasshouse pests is determined primarily by two factors: ecophysiological compatibility with local climate, particularly winter cold tolerance, and ability to locate and utilise wild prey. Observations on the number and diversity of acceptable wild prey as part of an assessment of establishment potential therefore overlap with more focused studies to determine host range. This study investigated two aspects of the interactions between biocontrol agents and non-target prey that are rarely considered in tests for establishment or host range: the role of different host plantprey associations in modifying the development and reproduction of biocontrol agents, and the longer term sustainability of such relationships beyond the single generation observed in most laboratory studies. Using the glasshouse whitefly (Trialeurodes vaporariorum) predator Macrolophus caliginosus Wagner (Hemiptera: Miridae) as a case study, the mirid was able to sustain viable populations over three generations on the related cabbage whitefly Aleyrodes proletella (Linnaeus) (Hemiptera: Aleyrodidae) and the aphid Myzus persicae (Sulzer) (Hemiptera: Aphididae), including when these prey were feeding on different host plants (Chinese cabbage, cabbage and Brussels sprout). However, the rate of development, fecundity and mortality of the predator varied between the different prey and host plant combinations, and in all cases differed than when feeding on its glasshouse prey T. vaporariorum reared on tobacco (Nicotiana tabacum). The results are discussed in the light of the current debate on methods for conducting host range testing as part of an environmental risk assessment.