Figure captions

Figure 1 . Changes in pest abundance, crop damage, yield, and natural enemy abundance when biocontrol interventions are implemented compared to untreated crops (untreated/monocropping). The values are expressed in percentage with 95% bias-corrected confidence intervals. Results that cross zero indicate no significant difference between control and treatment groups. K = number of articles, n = number of effect sizes
Figure 2 Changes in (a) pest abundance, (b) crop damage, (c) yield, and (d) natural enemy abundance when biocontrol interventions are implemented compared to untreated crops (untreated/monocropping). The values are expressed in percentage with 95% bias-corrected confidence intervals categorised as Botanical Pesticides (BP), Field margins (FM), Intercropping (Int), Push-Pull (PP). Results that cross zero indicate no significant difference between control and treatment groups, n = number of effect sizes
Figure 3 Changes in (a) pest abundance, (b) crop damage, (c) yield, and (d) natural enemy abundance when biocontrol interventions are implemented compared to untreated crops (untreated/monocropping). The values are expressed in percentage with 95% bias-corrected confidence intervals categorised as Cereal, Fibre, Fruit, Pulses and Vegetable (Veg) where available. Results that cross zero indicate no significant difference between control and treatment groups; n = number of effect sizes
Figure 4 Changes in (a) pest abundance, (b) crop damage, (c) yield, and (d) natural enemy abundance when biocontrol interventions are implemented compared to untreated crops (untreated/monocropping). The values are expressed in percentage with 95% bias-corrected confidence intervals categorised as Coleoptera, Hemiptera, Lepidoptera and Blattodea where available. Results that cross zero indicate no significant difference between control and treatment groups; n = number of effect sizes
Figure 5 Changes in (a) pest abundance, (b) crop damage, (c) yield, and (d) natural enemy abundance when biocontrol interventions are implemented compared to untreated crops (untreated/monocropping). The values are expressed in percentage with 95% bias-corrected confidence intervals categorised as small farms and research farms. Results that cross zero indicate no significant difference between control and treatment groups; n = number of effect sizes.
Figure 6 Changes in pest abundance, crop damage, yield, and natural enemy abundance when biocontrol interventions are implemented compared to crops treated with synthetic pesticides. The values are expressed in percentage with 95% bias-corrected confidence intervals. Results that cross zero indicate no significant difference between control and treatment groups. K = number of articles, n = number of effect sizes.