Intensification of Arabica coffee agroforestry systems in southwestern Ethiopia: Effects on coffee pollinator communities and coffee quality

Agricultural expansion and intensification are the major causes of tropical deforestation and forest degradation. Continued increases in production seems imminent because of the local, ongoing human population growth and increasing global demand for food and other commodities. Arabica coffee (Coffea arabica) is an example of a commodity crop for which demand will increase. Wild populations of C. arabica are endemic to natural moist Afromontane forests in southwestern Ethiopia. Yet, only small proportions of these forests remain, mainly due to land-use change for agricultural activities, including forest transformation to coffee production systems. Given that further natural habitat loss should be avoided, it is becoming increasingly clear that the long-term Arabica coffee production in Ethiopia relies on a sustainable production intensification on existing farmland. During this PhD research, we evaluated if actions to increase the coffee pollinator diversity and the coffee organoleptic quality can contribute to a sustainable intensification of Arabica coffee production in southwestern Ethiopia. We combined experimental and observational studies on forest integrity, coffee pollinators and coffee yield quantity and quality in southwestern Ethiopian coffee production systems of varying management intensity. All data was collected within the Jimma region. First, we performed an analysis at two time points (2009 and 2017) to quantify effects of ongoing coffee production intensification on the forest integrity of small-scale coffee agroforestry systems over time (Chapter 2). We found reductions in woody vegetation species richness (-23%), diversity (-14%) and density (-31%), and an increase in overall height (+11%). Our findings show a rapid continuation in habitat structure simplification and loss of woody vegetation diversity over time, in already intensively managed coffee production systems. Therefore, we expect that less intensified coffee production systems may be prone to even faster degradation rates. Next, we estimated effects of current management practices on the community of potential coffee pollinators by means of a space-for-time substitution (Chapter 3). We found negative effects of increasing intensification of coffee production on the species richness and abundance of insects that were visiting coffee flowers. As smallholder coffee farmers continue to reduce the woody vegetation density and diversity of their production systems, we expect further reductions in habitat quality and associated declines in coffee pollinator communities. In the next chapter, our results suggest that the heterogeneous and extensive agricultural landscape surrounding intensively managed coffee production systems, provide feeding and nesting habitat to non-Apis bees present within the coffee production systems (Chapter 4). By use of a flower-exclusion experiment, we also showed that the initial coffee fruit set benefitted from insect pollination within these coffee production systems. It is possible that benefits on the final fruit set may have been masked by high proportions of fruit drop which were potentially caused by a poor physiological condition of the cultivated coffee shrubs. Furthermore, we found a positive relationship between the non-Apis bee diversity (collected in 2018) within these intensively managed coffee production systems and the three-year (2016 - 2018) coffee yield stability within the same systems. Overall, these results provide indications for the potential importance of pollination services to the short-term coffee yield and long-term yield stability, along with the potential of surrounding habitat to increase the pollinator diversity within intensified coffee production systems. In our final study, we used another space-for-time substitution to estimate potential effects of ongoing management intensification on the organoleptic quality of coffee beans (Chapter 5). We compared natural moist Afromontane forests to intensively managed coffee agroforestry systems of varying size, we standardized the post-harvest processing of the collected beans, and we controlled for variation in soil, elevation, and C. arabica genotype during statistical analysis. We found a significant decrease in organoleptic quality of coffee beans from natural forests over larger coffee agroforestry systems, to smaller coffee agroforestry systems. Only coffee beans collected from natural Afromontane forests qualified as specialty beans, according to the Specialty Coffee Association of America's standards. We suggest that the most important causes of deteriorating coffee organoleptic quality include decreased shade levels and associated changes in micro-climate and biotic interactions. Overall, the results of this PhD research contribute to a growing literature body highlighting the unsustainability of the current coffee production intensification in southwestern Ethiopia due to increased forest degradation and associated negative effects on biodiversity and ecosystem service provisioning. Our results suggest that increasing the pollinator diversity within the southwestern Ethiopian coffee production systems, through improvement of pollinator habitat quality either within or around coffee production systems, has potential to sustainably increase the coffee production. Improving the organoleptic quality of coffee beans produced on current coffee production systems can also contribute to obtain a sustainable coffee production. Partly because premium prices could increase the income of farmers, yet only if transmission along the value chain is improved. And partly because development of high quality coffee beans will most likely depend on the wild C. arabica gene pool and, thus, on the conservation of natural moist Afromontane forests. Additionally, we found indications for increased organoleptic quality of coffee beans with reduced management of the canopy structure. Further studies are, however, needed to elucidate on optimal density and diversity levels of the canopy and understory within coffee production systems, by balancing yield gains through increased pollination services and increased organoleptic quality on the one hand, and yield losses due to increased shade levels on the other hand. Moreover, further research is needed to identify which plant species and habitat structure provide space-efficient high-quality habitat to support targeted pollinator species in the landscape surrounding coffee production systems.status: publishe