Plant Species Modulate Plant-Mediated Indirect Effects of Orius Sauteri on Pests Frankliniella Occidentalis and Bemisia Tabaci

Exploring the interactions between host plants, herbivores and natural enemies is an important experimental approach to enhance biological control and induced plant defense responses following infestation by herbivores could enable plants to minimize damage. Orius sauteri (Poppius), an important generalist predator, has been widely used as a biological control agent for suppressing many agricultural pests on agronomic and horticultural crops. It is also documented as ovipositing eggs in plant tissues and piercing and sucking as omnivorous predator. We therefore explored the �tness parameters of primary (western �ower thrips) and secondary pests (white�ies) on three different plant species pre-inoculated by O. sauteri. Pre-inoculation of O. sauteri signi�cantly reduced the survival of Frankliniella occidentalis (Pergande) on tomato and Bemisia tabaci (Gennadius) on cowpea. The reproduction of B. tabaci on tomato, cucumber and cowpea was affected by O. sauteri in a different manner. The presence of O. sauteri signi�cantly reduced the reproduction of F. occidentalis on tomato and cucumber plants, but there was no signi�cant effect on cowpea plants. This demonstrates that plants with the pre-inoculation of O. sauteri decrease the performance of herbivores on different plant species but to differing degrees. These results further enhance our knowledge of ecological relationships between natural enemies and herbivores, and provide the context for the early release of natural enemies to control pests.

reported that tomato plants exposed to the omnivorous predator M. pygmaeus reduced the performance of Tetranychus urticae Koch and Frankliniella occidentalis (Pergande) due to the activation of JA and ABA signaling pathways involved in plant defenses. The feeding and oviposition behavior of O. laevigatus were also shown to increase the resistance of plants against Bemisia tabaci (Gennadius) and F. occidentalis by activating JA or SA signaling pathways (De Puysseleyr et al. 2011). Given the importance of omnivorous predators in biological control, it is essential to study the ecological complex between these natural enemies and plants in the context of plant induced responses.
The omnivorous predator Orius sauteri (Poppius) is widely used in biological control in China for managing agricultural pests such as white ies, thrips, aphids, eggs and young larvae of lepidopteran pests, spider mites and leaf miners (Zhao et  Research has also revealed that O. sauteri has a strong predation ability on western ower thrips, consuming large numbers per day (Zhang et al. 2007). Wang et al. (2013) reported that O. sauteri could prey on all stages of B. tabaci and the number of prey consumed increased with prey density to a maximal daily predation on pseudopupa of 23.50. It has been used as the main predator to control thrips on strawberries, solanaceous and cucurbitaceous plants in greenhouses, and as supplementary predator for controlling white ies and aphids in China (Yin et al. 2013 Given the importance of examining this interaction, we used O. sauteri as the research subject which was pre-inoculated onto three plant species including tomato Solanum lycopersicum Mill., cucumber Cucumis sativus L. and cowpea Vigna unguiculata (L.) Walp. We subsequently observed the tness of the primary (F. occidentalis) and secondary (B. tabaci) pests. Examination of the survival and reproduction of pests allowed us to study three components in this system: (1) to quantify whether the omnivorous predator O. sauteri affects the tness of pests, (2) compare whether the tness of pests differs on different plants after pre-inoculation with the predator and (3) contrast differences between primary and secondary pests. Addressing such questions, this study aims to enhance the understanding of the ecological relationship between O. sauteri and pests in theory, but also provide new direction for the precise application of O. sauteri for biological control. Sanyo, Japan) set to 26 ± 1 ºC, 70 ± 5% RH, 16: 8 h L: D photoperiod, 3,000 Lux uorescent light. Adults and nymphs were reared with fresh rice moth Corcyra cephalonica (Stainton) eggs with 10% honey being supplied. The ovipositing substrate for female adults was hyacinth bean Lablab purpureus (L.) Swee and fresh new substrates were replaced every day. Beans with eggs were kept in a single box and adults of the same age (3-5 days) were used for the experiments.

Plants
F. occidentalis adults were supplied by the Institute of Plant Protection (IPP), Chinese Academy of Agricultural Sciences (CAAS, Haidian District, Beijing, China) and a colony was established at the LAB, BIPP, BAAFS. These pests were reared in the same plastic boxes as described above. All the plastic boxes were put in a climate-controlled growth chamber (MH-351, Sanyo, Japan) set to 26 ± 1 ºC, 70 ± 5% RH, 16: 8 h L: D photoperiod, 3,000 Lux uorescent light. Hyacinth beans with 5% honey water brushed on the surface were provided and the plants were replaced every day to obtain the same aged F. occidentalis from each rearing box. Female adults aged 2-3 days after emergence were used in this experiment. To obtain the same aged white y adults, new four-leaf stage eggplant plants were placed in the cages for 48 h for oviposition. Adult white ies were removed and the eggs were allowed to develop on eggplant leaves. After 3 weeks, newly emerged adults (up to 3 days old) were collected for use.
Survival rate and F1 nymph number of Frankliniella occidentalis on three plant species inoculated by Orius sauteri Tomato plants at the ve-leaf stage, cucumber plants at the ve-leaf stage and cowpea plants at the fourleaf stage were used in the experiments. The third leaf from the bottom of the plants was xed with a leaf cage made of small plastic petri dishes (height = 4.5 cm, diameter = 3.5 cm) with a clip and a thin layer of sponge placed on the side contacting leaves to prevent damage. Plants were chosen randomly for each treatment. To assure the oviposition of O. sauteri on plants, the predators were provided su cient C. cephalonica eggs prior to transferring and a large number prey eggs were provided on an egg card (size: 2.0 × 1.0 cm, with approximately 2,000 eggs in the card) in the leaf cage. Three mated female adults (3-5 days after emergence) were carefully transferred to the leaf cage with a ne brush. Plants from the control groups were also xed with leaf cages but no predators were added. After 24 h, female adults were removed from the leaf cage and 20 female adults of F. occidentalis (aged 2-3 days) were put into the cage. The number of adults surviving were counted and recorded at 24 h and 48 h after which time all remaining F. occidentalis were removed. After a further 120 h, the number of F1 nymph of F. occidentalis was observed and recorded. Tomato, cucumber and cowpea expriments were repeated 12, 9 and 8 times, respectively.
Survival rate and egg number of Bemisia tabaci on three plant species inoculated by Orius sauteri The plants from the three species were treated as described above. After the female adults of O. sauteri were removed, 10 pairs of B. tabaci adults were transferred carefully into the leaf cage. The number of B. tabaci remaining was observed at 24 h and 48 h, after which surviving adults were removed and the number of eggs laid by B. tabaci was observed under a stereomicroscope (XTL-165-VT, Pheenix, China). Plants with a leaf cage but no predators were used as control groups and each plant species was repeated 8 times.
Oviposition, hatching rate and F1 nymph number of Orius sauteri on three plant species Plants and natural enemies were treated as described above. The number of eggs of O. sauteri on each plant was observed and female adults were removed 24 h after the inoculation. After removal, F. occidentalis or B. tabaci adults were added into the leaf cages and removed after 48 h. The F1 number of O. sauteri on the plants was quanti ed and recorded 72 h later after removing the pests. The hatching rate of the natural enemy was calculated and each plant species repeated 8-12 times.

Statistical analysis
The number of eggs, F1 nymph number and hatching rate of O. sauteri were analyzed using one-way ANOVA (Duncan's new multiple range test) at P < 0.05 level. Meanwhile, the data of the survival rate of F. occidentalis and B. tabaci, F1 nymph number of F. occidentalis and the egg number of B. tabaci were analyzed using Student's t-test (P < 0.05). The hatching rate of O. sauteri, the survival rate of F. occidentalis and B. tabaci were arcsine square root transformed prior to analysis. Pearson correlation test was used to analyze the correlation between the oviposition behavior of O. sauteri and the survival and reproduction of pests on three plant species. The oviposition behavior of O. sauteri included the egg number, F1 nymph number and hatching rate of O. sauteri ( Table 1). The survival and reproduction of pests included survival rate of F. occidentalis at 24 and 48 h, survival rate of B. tabaci at 24 and 48 h, F1 nymph number of F. occidentalis and egg number of B. tabaci (Table 1). All data analyses were conducted using the software SPSS 23.0 (IBM, Armonk, NY, USA).

Number of eggs of Bemisia tabaci reared on three plant species inoculated by Orius sauteri
The number of B. tabaci eggs was reduced signi cantly on tomato (t = 5.803, P < 0.001) and cowpea (t = 4.385, P = 0.001) plants inoculated with O. sauteri but no difference was detected from cucumber plants (t = -0.586, P = 0.567) (Fig. 2b).
Oviposition number, F1 nymph number and hatching rate of Orius sauteri on three plant species The number of eggs of O. sauteri showed no signi cant difference (F = 0.104, df = 5,54, P = 0.991) (Fig. 3a) but the number of F1 nymphs (F = 4.728, df = 5,54, P = 0.001) (Fig. 3b) and hatching rates (F = 6.747, df = 5,54, P < 0.001) (Fig. 3c)  activated Ca 2+ channel, CNGC19, plays a key mechanical role in the recognition of herbivore feeding and activation of defense signaling pathways. Bph9 and Bph6, resistance genes to the brown planthopper, Nilaparvata lugens (Stål), have the ability of repellency, resistance and inhibition of feeding by regulating the signaling pathway of SA and JA (Zhao et al. 2016;Guo et al. 2018). This provides a clear framework for future research to further decipher the role of plant defenses to herbivores following inoculation with omnivorous natural enemies.
In conclusion, this study provides a theoretical basis for the prevention and control of pest species through the interaction between O. sauteri and different host plants. This is the rst report demonstrate how inoculation of O. sauteri on plants can affect the tness of F. occidentalis and B. tabaci providing a new opportunity for the use of the omnivorous predator as an important generalist biological control agent.
Declarations Figure 1 Survival rate of Frankliniella occidentalis (a, b, and c) and Bemisia tabaci (d, e, and f) at 24 and 48 h on three plant species inoculated with Orius sauteri. Data in the gure were means ± SE. Signi cant differences based on Student's t-test are marked with asterisks (*, P < 0.05; **, P < 0.01; ns, not signi cant).