Introduction

One of the greatest global challenges of the twenty-first century is meeting the increasing demands for human food production while minimising adverse impacts on biodiversity and ecosystem health (Godfray et al. 2010). This challenge is particularly critical in sub-Saharan Africa (SSA) where the population is predicted to double over the coming decades (Rosegrant et al. 2009) and food production is hampered both by climate change impacts (Lobell et al. 2011) and significant yield losses caused by crop pests (Lenné 2000; Oerke & Dehne 2004). For example, the invasion of the fall armyworm (Spodoptera frugiperda ), which has caused crop losses of about $3 billion a year in SSA, has become one of the most important threats to maize production (Stokstad 2017). The fall armyworm is also a cause of major damage to other crops including rice, sorghum, millet, cabbage, and tomatoes (CABI, 2018), demonstrating the vulnerability of smallholder farming to crop pests.
Conventional synthetic pesticides have severe limitations as a means of pest control in SSA because they are economically inaccessible for a large portion of smallholder farmers in the region (Ahissou et al. 2021). Pesticide residues also put human and livestock populations at risk from contaminated food and forage (Nesser et al. 2016; Jepson et al. 2020). Furthermore, synthetic pesticides may lead to resistance within pest populations (Sawadogo et al. 2020), and have negative impacts on non-target organisms, such as pollinators and natural enemies, and the ecosystem services that biodiversity provides in the production of food (Losey et al. 2006; Chaplin-Kramen et al. 2011; Kennedy et al. 2013). If the reduction of natural enemy populations is greater than that of the pest, this may lead to the resurgence of pests following pesticide applications (Janssen and van Rijn 2021), which is a widely reported problem associated with synthetic pesticides (Guedes et al. 2016).
Biological control methods (hereafter biocontrol), which employ natural enemies of crop pests, have been promoted globally as an alternative approach to pest control, and are often used as part of an integrated pest management strategy (Giles et al. 2017; Baker et al.2020). Extensive evidence is available on the responses of natural enemies to the landscape composition surrounding crop fields (Karp et al. 2018), which reveals that landscape effects are a key driver of pest regulation by natural enemies. Recent syntheses show consistent positive responses of natural enemies to landscape complexity (Chaplin-Kramen et al. 2011), a reduction of natural pest control in simplified landscapes (Rusch et al. 2016), and higher natural enemy populations in complex versus simple landscapes (Bianchi et al. 2006).
In SSA, smallholder farmers have employed conservation biocontrol methods and botanical pesticides for the control of crop pests (Sporleder & Lacey 2013). Conservation biocontrol methods include intercropping, push-pull and field margins (Table 1). Growing evidence highlights the potential of biocontrol interventions to reduce pest incidence and increase yield (Amoabeng et al. 2020; Farsia Djidjonri et al. 2021). For example, push-pull technology has been shown to be effective against a range of crop pests, particularly maize stemborers (Midega et al. 2018), and plant-based botanical pesticides can reduce pest incidence and enhance yield in vegetables crops (Mpumi et al. 2020; Odewole et al. 2020).
Although biocontrol interventions may provide sustainable and accessible alternatives to synthetic pesticides, their adoption by smallholder farmers has not been widespread (Ratto et al . 2022). This may be due to knowledge gaps relating to their effectiveness and the factors that lead to their success or failure, particularly in comparison to synthetic pesticides. Biocontrol techniques have been applied to numerous crops and targeted a wide variety of pests in the region, yet there is a lack of understanding of how the effectiveness of biocontrol varies across different crop types and pest taxa (Ratto et al.2022). Recent research in Tanzania found greater natural enemy diversity in fields surrounded by intercropped fields, suggesting spatial flow of potential biocontrol services across landscapes (Tripathi et al. 2022), but the established relationship between landscape composition, natural enemies and pest regulation is almost entirely based on studies carried out in the global north, and very seldom in sub-Saharan regions where farmers are most exposed to food insecurity caused by crop pests (Steward et al. 2014). This represents a key knowledge gap; more clarity is needed about the environmental factors affecting biocontrol performance in sub-Saharan Africa to better assist in smallholder farmer decision making, and to determine the broader indirect impact of pest management options on biodiversity compared to synthetic pesticides, both on a farm and at a landscape scale.
Quantitative analyses have been conducted on the performance of biocontrol agents (Stiling & Cornelissen 2005), on the impact of landscape context on augmentative biocontrol (Perez-alvarez et al. 2019) and pest and natural enemy responses (Chaplin-Kramen et al. 2011). However, none of these approaches have focussed specifically on the sub-Saharan region, nor have they evaluated the efficacy of different biocontrol interventions on crop pest populations and their damage to crops.
Here, we aim to better understand the key factors driving the success or failure of biocontrol interventions using quantitative meta-analysis. Specifically, we posed the following questions:
(1) What are the effects of biocontrol interventions on the management of insect crop pests in sub-Saharan Africa? (2) Are these effects consistent across biocontrol techniques, crop types, target pests and farming systems? (3) How does the effectiveness and impact of biocontrol interventions on crop pests and non-target insects compare to synthetic pesticides? (4) Does the surrounding landscape composition affect the efficacy of biocontrol interventions?
We hypothesised that pest abundance and crop damage would decrease, and yield would increase in crops subject to biocontrol interventions, that the impact on natural enemy abundance would be less than that of synthetic pesticides, and that these effects would be enhanced in fields surrounded by greater landscape complexity.