Arthropod predators and parasitoids support the health and functioning of the world's ecosystems, most notably by supplying biological control services to agricultural landscapes. Quantifying the impact that these organisms have on their prey can be challenging, as direct observation and measurement of arthropod predation is difficult. The use of sentinel prey is one method to measure predator impact; however, despite widespread use, few studies have compared predation on different prey types within a single experiment. This study evaluated the predation rates on four sentinel prey items in grass and wheat fields in south‐east Queensland, Australia. Attack rates on live and dead Helicoverpa armigera eggs, and dead H. armigera larvae and artificial plasticine larvae, were compared and the predators that were attracted to each prey type were documented with the use of field cameras. There was no significant difference in predation rates between sentinel eggs, while dead larvae were significantly more attacked than artificial larvae. Prey were attacked by a diverse range of predators, including ants, beetles, various nymph and juvenile insects and small mammals. Different predators were active in grass and crop fields, with predator activity peaking around dawn and dusk. The same trends were observed within and between the two habitats studied, providing a measure of confidence in the sentinel prey method. A range of different sentinel prey types could be suitable for use in most comparative studies; however, each prey type has its own benefits and limitations, and these should be carefully evaluated to determine which is most suitable to address the research questions.

Grasslands are valuable non-crop habitats in the world's agricultural regions, providing more than simply forage for grazing domestic and wild animals. They provide refuge and resources that support high levels of arthropod biodiversity, most importantly for natural enemies that provide vital biological control services to the surrounding cropped landscape. Ideally, farmers could manage and manipulate grasslands to boost their biocontrol services, although to achieve this, knowledge of the ecological function of these habitats is essential. Unfortunately, grasslands are often bundled together with other habitat types, such as scrubland and forest, under the label 'non-crop habitats', and little is known about the contribution that these specific habitats make towards landscape pest suppression. While recent research has been investigating the importance of other non-crop habitat such as native vegetation remnants, the contribution that grassland habitats may make towards landscape pest suppression remains a significant knowledge gap in biocontrol research. Here, the current understanding of grassland habitats as biological control service providers in the world's mixed farming systems is reviewed. Limited research into whether grassland habitats support natural enemies, and thus contribute biological control services to the adjacent surrounding cropped landscape, has returned inconclusive, even conflicting, results. Potential explanations for this inconsistency are explored, including the lack of studies, inadequate estimates of predator impact and the variety of different grassland habitat types studied, including their diversity in composition and management practices. Conclusions drawn from these studies are discussed, and suggestions for management are recommended, including increasing grassland floristic diversity, limiting intensive management practices and implementing weed control. Future research directions are proposed, along with the need to develop a universal grassland classification system for research, management and conservation purposes, using grassland type, floristic diversity and management practices as key axes for classification.

Grasslands provide valuable resources and refuges that support arthropod biodiversity in everchanging agricultural landscapes. Potential exists to manipulate and manage grasslands to enhance the efficiency of natural biocontrol services provided to surrounding croplands, whilst discouraging the build-up of agricultural pests. In order to successfully achieve this, a deeper understanding of the ecological function of grassland habitats is needed. Grasslands are diverse habitats that differ significantly in their composition, function and management, despite consistently being bundled together in research. Exploring these differences and how they influence arthropod communities, and thus landscape pest suppression, remains a significant gap in biocontrol research. Current knowledge of the arthropod pests and natural enemies that utilise different grassland habitats is scarce, and even less is understood about the potential for different grasslands to be biological control service providers, making progress in this area challenging. Three distinct grassland types were distinguished in the mixed farming regions of southeast Queensland, Australia. Native grasslands comprised a mix of predominantly native grass species, Mixed Improved grasslands were a sown mix of both introduced and native grass species, and Pure grasslands comprised a single dominant grass species. To compare the above-ground arthropod communities that inhabit these grasslands, field sampling was carried out over a two-year period. Arthropods were collected from grassy fields by sweep netting, while vegetation assessments monitored grassland composition, floristic diversity, quality and management practices. Only 6% of all arthropods collected were agriculturally important pests, while 25% were beneficial predators and parasitoids. Arthropods were almost 50% more abundant in Native grasslands than in the other types, although Pure grasslands supported the largest communities of natural enemies. Pest and other herbivorous insects favoured grasslands with higher proportions of weeds and other undesirable vegetation (i.e. Native and Mixed), and were less abundant in grasslands with high proportions of desirable introduced grass species (i.e. Pure). Natural enemies used all grasslands as a refuge habitat when adjacent crops were absent, while pest and other herbivorous arthropods did not. This study highlights the importance of grasslands as valuable arthropod refuges, particularly for beneficial natural enemies, and draws attention to the differences between different grassland types that can influence these communities. The observation that grasslands harbour significant communities of predatory and parasitic arthropods raises the question of whether these natural enemies are providing biological control benefits to the greater landscape. An experiment using sentinel prey was conducted to test the hypothesis that grasslands do provide pest suppressive services to surrounding crops, although iii services provided by different grassland types will differ...