Effects of local and landscape scale factors on ant diversity and biocontrol of the coffee berry borer in Colombia

The coffee berry borer –CBB- (Hypothenemus hampei) is the economically most important pest of coffee worldwide. Within the Integrated Pest Management –IPM- strategy for this borer, the biological control component has received great attention, particularly under a classical biological control approach. However, despite the large numbers of relevant papers on this topic it is still unknown i) what are the most effective biocontrol agents and which crop management practices are most successful. Furthermore, despite the growing evidence on the importance of the landscape for natural enemies diversity and pest control services in agroecosystems, knowledge gap of CBB biocontrol is still big. We need to answer questions such as i) does spatial heterogeneity in coffee landscapes affect diversity patterns of CBB biocontrols? and ii) how do local and landscape-scale factors (and their interactions) affect fruit infestation by the CBB? This dissertation aims to fill the above-mentioned research gaps in three main chapters. In chapter 1 we reviewed the most relevant peer-reviewed literature on CBB biocontrol published between 1990 and 2017 in order to compare control efficacy among taxonomic groups and to identify how crop management practices at the farm level and landscape-scale factors affect biocontrol success. We found that different taxonomic groups, mainly fungi, ants, microhymenoptera and nematodes, provide successful control. Ants were the enemy group with the highest number of published papers on effective CBB control under field conditions being up to 6 times more effective to reduce pest impact than experimental treatments without ants. Over 40% of the studies showing effective CBB control do not disclose information about the use of agrochemicals or shade management, which makes evaluations of potential biocontrol difficult. Only one out of 22 the studies showing successful biocontrol explicitly included landscape-scale factors in their evaluations for CBB success. Based on the knowledge gaps found in the literature review, we assessed in chapter 2 the effects of local and landscape-scale management practices on the diversity of ants, which are known to be successful CBB biocontrol agents. Using tuna baits, we sampled coffee-foraging ants in three land-use types along an agricultural intensification gradient (forest, shaded coffee and unshaded coffee) in a Colombian coffee landscape. Results showed that ant species turnover among plots was reduced with management intensity (i.e. loss of shade cover), from a mean species richness of 14.5 (± 1.32) in forest to 10.0 (± 1.23) species in unshaded coffee plantations. We also found the highest species habitat specificity in forest and the lowest habitat specificity by dominant ants in unshaded coffee when around a 40% of forest was bordering the plots. Our results suggest that preserving the forest in the coffee agroecosystem could enhance ant diversity at the landscape scale through increasing species turnover among coffee plantations, and by decreasing habitat specificity of highly dominant ants in unshaded coffee plantations. We evaluated in chapter 3 the importance of ant presence, together with management practices at local and landscape-scale, on CBB attack rates. Results confirmed the importance of ants to control and reduce CBB infestation rates as we found a 16.9 % relative increase in fruit infestation rates when ants were excluded from coffee branches. Additionally, this chapter showed that CBB biocontrol management should be implemented in coffee agroecosystems and landscapes at different spatial scales. Specifically, we found that local conditions such a higher number of twigs and the absence of shade, were related with lower CBB infestation rates. At the landscape-scale, larger plot perimeter sections bordering other crops, increased infestation rates in the studied region. In conclusion, we demonstrate the importance of natural enemies to provide autonomous pest control in a Colombian coffee landscape. We showed that i) ant presence and diversity, and ii) the interaction of local (i.e., number of twigs in the ground, tree presence) and landscape factors (i.e., non-cropped area surrounding the coffee plantations) should be considered and explicitly managed in areas where coffee is the dominant crop. We propose that coffee management in this region should be oriented to i) promote coffee-foraging ants and forested habitat to maintain high ant species turnover in coffee landscapes, and additionally, ii) combine management practices at local and landscape scales to better control CBB infestation. In this dissertation we also highlight the importance of implementing a Conservation Biological Control approach in the IPM strategy for CBB control, which seeks to enhance abundance and efficacy of natural enemies at local and landscape scale.