Hide and seek: management and landscape factors affecting maize stemborers Busseola fusca (Fuller) infestation levels in Ethiopia

Worldwide land cover and land use change rapidly due to biophysical and complex socio-economic and political factors. These changes are directly affecting both local and global biodiversity and impeding the ability of agricultural landscapes to provide essential ecosystem services. These changes are particularly evident in sub-Saharan Africa, which is experiencing a rapid transformation in rural and urban areas as a consequence of urbanisation and population growth while simultaneously being the continent in most urgent need for increasing agricultural production and most threatened by climate change and pest outbreaks. There is growing need for alternative agricultural practices that conserve biodiversity and natural regulatory processes in order to meet the rising demands for food and dietary diversity, to mitigate climate change, and to restore degraded landscapes. To better understand the potential of current agricultural landscapes to provide essential ecosystem services, insight is required into the historic trajectories of farming systems, their drivers, and how these drivers shaped current landscapes in terms of composition and structure, as well as the impact they have on ecosystem services. These insights can contribute to informing the design of more sustainable cereal-based agroecosystems for the smallholder multifunctional subsistence agriculture.

Ethiopia is now the second most populated country in Africa with more than 100 million people and a population growth rate of 3% per year. In Ethiopia, cereals are the major staple crops with maize ranking second after teff (Eragrostis tef) in acreage and first in total production and productivity. In the Hawassa area, in the Rift Valley of Ethiopia, maize (Zea mays) productivity fluctuates widely, and that is caused in part by infestation by maize stemborers. Adults stemborer are nocturnal moths which disperse by flight, while the larval stages have a wide range of host plants. Current recommended pest management practices for stemborers only focus on the plot level and do not take into account the entire farming system or the composition of the surrounding landscape context. Yet farming systems and the associated management practices and landscape contexts are crucial for understanding the population dynamics of stemborers, their natural enemies, and the resulting pest pressure. In Hawassa, there is an ongoing transformation of farming systems, and these changes are influenced by institutional and socio-economic drivers, such as land tenure regulation, market access, and population growth. It is unclear how these drivers influence the dynamics of farming systems and ultimately shape agricultural landscapes and their potential for the provision of food, feed, and energy. The general objective of my work was to identify stemborer management strategies at the field, farm, and landscape levels for a more sustainable intensification of maize-based production systems that would (i) limit stemborer incidence, (ii) maintain or improve soil fertility, and (iii) improve fodder production for livestock. In particular, I studied how trajectories of farming systems have shaped current agricultural landscapes and assessed the implications for maize stemborer pest pressure and the potential for biocontrol. I quantified the relations between stemborer abundance, farming practices, and maize yields at the plot level and analysed the impact of landscape factors on stemborers and the abundance of their natural enemy. This brings new insight that can inform effective farming practices for increased maize production and stemborer suppression at multiple spatial scales.

In Chapter 2, I assessed how the ongoing expansion of arable land and urban areas is affecting the availability of common resources, such as forest and grazing land, and the availability of biomass for food, feed, and energy. By combining data from (i) farm household surveys, (ii) focus group discussions with farmers, (iii) statistical typology of trajectories of change in farming systems, (iv) remote sensing, and (v) secondary data analysis,  I showed that the current farming systems in Hawassa result from the combined effects of past policies, population growth, access to market, urbanisation, and biophysical characteristics. This work has revealed, among other things, that farm sizes declined rapidly over the past four decades and that farmers responded to this constraint by adopting three main livelihood strategies: consolidation, diversification, and specialisation. These trajectories combined with urbanisation led to more fragmented and more complex landscapes.

In Chapter 3, I assessed how the changes in land use and landscape composition and structure in the Hawassa region influenced the capacity of the landscape to support communities of the natural enemy of maize stemborers: Busseola fusca (Fuller). Natural enemies were sampled in maize fields adjacent to simple hedgerows, complex hedgerows, enset (Ensete ventricosum) fields, and khat (Catha edulis) fields at 1, 10, and 30 m using pitfalls and yellow pan traps in 2014 and 2015. The landscape analysis indicated that landscapes in the study area have changed over time from maize-dominated to more diverse small-scale and fragmented agroecosystems with a higher proportion of perennial crops. In maize fields adjacent to enset and complex

hedgerows, I found a higher abundance of predators (mostly ants and rove beetles) than in maize fields adjacent to khat and simple hedgerows, and the influence of border type decreases with distance from the border. The abundance of parasitoids and parasitic flies were not influenced by border type. I concluded that in terms of biocontrol service, the changes observed in landscape composition and structure may have influenced the capacity of the landscape to support populations of natural enemies of stemborers in different ways. On the one hand, smaller field sizes have resulted in more field borders that may support relatively high predator densities; on the other hand, the area of khat increased, and the area of enset decreased, which may have had a negative effect on predator densities.

In Chapter 4, I investigated how farmers’ management practices at the field scale and landscape context affect maize stemborer infestation levels and maize productivity. Maize infestation levels, yield, and biomass production were assessed in 33 farmer fields managed according to local practices. When considering field level factors only, plant density was positively related to stemborer infestation level. During high infestation events, the length of tunnelling, a proxy of infestation and plant damage, was positively associated with the date of  planting and negatively associated with the botanical diversity of hedges. However, the proportion of maize crops in the surrounding landscape was strongly positively associated with the length of tunnelling at 100, 500, 1000, and 1500 m radii. The findings reveal that the landscape context overrides farmers’ management practices in explaining maize infestation levels but also indicate that maize is tolerant to low and medium infestation levels of stemborers.

The push-pull system, a stimulo-deterrent cropping strategy consisting of intercropping cereals with legumes and surrounding by fodder grasses, is considered a promising crop diversification strategy for smallholder farmers in Africa as it may contribute to maize stemborer Busseola fusca (Fuller) suppression while improving soil fertility and providing feed for livestock. In Chapter 5, I investigated the performance of different push-pull systems in terms of stemborer suppression, predator abundance, and maize productivity in different landscape settings. Within each landscape (simple, intermediate, and complex), experimental plots were established on four representative smallholder farms. At each farm we used a split-plot factorial design with main plots surrounded or not by Napier grass and subplots consisting of sole maize, maize-bean, or maize-Desmodium. I assessed stemborer infestation levels and maize grain and stover yields for two years; I also assessed

natural enemy abundance and egg predation at two maize development stages in the second year. I demonstrated that the push-pull system was effective in reducing stemborer infestation only in the intermediate complexity landscape, where subplots with sole maize had higher stemborer infestation levels compared to maize-bean or maize-Desmodium. In the simple landscape, which was dominated by maize, all treatments had high stemborer infestation levels irrespective of within-field crop diversity; the presence of Napier grass was associated with higher predator abundance, while egg predation rates were the highest in the maize-bean intercrop. In the complex landscape, infestation levels were low in all treatments. I found no significant difference between the two push crops tested - Desmodium or bean - suggesting that beans can be used as push crop in push-pull systems with the additional advantage of increasing egg predation rate and being a common maize-bean farmers’ practice. However, there was no significant yield differences  between the sub-systems nor between the three landscapes. Thus, the benefits of the push-pull system mostly come from the companion crops (bean, Desmodium, and Napier) rather than from stemborer suppression per se.     

Agricultural landscapes in the Hawassa area went through important transformations over the last 40 years due to  the combined effects of national, regional, and local level drivers (agricultural policies, commodity prices, and infrastructures), regional/local level factors (population density and urbanisation),  farmers’ livelihood assets, and unexpected climate events. The area of maize monocultures declined and was progressively replaced by perennial crops, such as enset (food crop) and khat (cash crop). In addition, population growth and the expansion of urbanised areas have reduced the availability of land and led to more fragmentation of the croplands, with a potential positive benefit for the biocontrol of maize stemborers. Although the push-pull system is a promising crop diversification strategy for smallholder farmers in Africa, its adoption can be limited in land-constrained farming systems, most likely due the farmers’ reluctance to replace food crops, such as common bean, with fodder crops. In addition, the development of Napier grass can be hampered when nitrogen is a limiting factor. Therefore, there is a need for developing push-pull systems that use locally available plants. Napier grass could be planted as part of hedgerows to avoid hampering mechanical work of fields. In general, increasing plant diversity of hedgerows and their density can contribute to increasing natural enemy abundance and decreasing maize infestation levels. In addition, the intentional management of hedgerows could also consider adding plants and trees that are multifunctional (e.g. botanicals, feed for livestock, erosion control, carbon storage, beneficials for soil fauna). Above all, given the overriding influence of landscape context over field level management practices and the multifunctional nature of smallholder farming systems,  the design of sustainable agroecosystems requires a context-specific and strongly integrative (social, economic and, environmental objectives) approach. Taking maize stemborer pressure as an entry point, I showed that the infestation cannot be explained by field level factors only. Tackling maize infestation issues requires a landscape approach for sustainable pest management. Landscape composition, in particular, could either impact the pest abundance directly by affecting its dispersal, mortality, or reproduction or indirectly by affecting its natural enemies. Yet a landscape design which aims not only at the ecological control of maize stemborers but also addresses other farming constraints (i.e. soil fertility, fodder availability) and aim at maintaining moisture to avoid crop failure (by using cover crops, increasing rainfall infiltration) and diverse farming systems to increase nutrition and income diversity for smallholder farmers. Diversified farming systems which promote the conservation of natural enemies seem to be an ecologically sound solution, but more research is needed to understand and maximise the efficiency of existing mixed cropping systems. A systems approach to agriculture production such as agroecology could be the best option to respond to the requirements of the multifunctional small-scale subsistence agriculture in Africa.