Landscape-scale mechanisms of biological pest control in a South Korean agricultural landscape

Biological pest control is a major ecosystem service provided by natural enemies to world crops. However, current understanding of how to manage this service in real-world landscapes is limited by our lack of knowledge of the factors determining landscape-scale pest control, enemy diversity and the relationship between the two. In order to clarify how biological control is provided at large spatial scales, and how it can be improved, this thesis explores the mechanisms determining large-scale pest control provision and natural enemy diversity in a South Korean agricultural landscape. Landscape complexity is known to benefit natural enemies, yet it is unclear whether and how this translates into increased pest control, damage avoidance and yields. In a field exclusion experiment replicated along a gradient in landscape complexity, characterized by increasing amounts of seminatural habitat around fields, pest control of aphids and Lepidoptera was shown to depend on interactions occurring between functional guilds of natural enemies. Control of Lepidopteran larvae by flying insect enemies increased with the proportion of seminatural habitat around fields. However, so did negative interactions between flying insects and birds. Thus, negative enemy interactions constrained Lepidopteran control in complex landscapes. This effect impacted all levels from pests, to herbivory, to yields. These results suggest that by altering the outcome of trophic interactions between natural enemies, landscape complexity may promote ecosystem services as well as disservices. In contrast, despite a range of positive and negative interactions between enemies, landscape complexity positively impacted total pest control of aphids. Similarly to Lepidoptera, aphids were best suppressed by the guild of flying insect enemies, whose effects were strongest under conditions of high landscape complexity. Interactions between flying insects and ground-dwellers were complementary in all landscapes, whereas birds had no overall impact on aphid control. Overall, these results suggest that flying insects including syrphids, parasitoid and predatory wasps show the highest potential for improvement of pest control particularly in complex landscapes. However, their effectiveness depends on limiting negative interactions with other enemies, thus on our ability to rigorously manage enemy diversity at all relevant spatial scales. Effects of landscape complexity on enemy diversity are most frequently investigated according to the amount of (semi)natural habitat around fields, i.e. landscape composition. However, this parameter is often highly correlated with landscape diversity and configuration. In order to disentangle the relative importance of these and local factors for enemy diversity across scales, seven enemy taxa were sampled along uncorrelated gradients in landscape composition, diversity and configuration. Instead of habitat amount, a complex configuration positively influenced most natural enemies at all scales. Interactions between local and landscape factors were found only for birds. Further, high enemy diversity was not reflected by low crop damage. This study shows that enemies respond to distinct landscape factors across scales, thus providing important windows of opportunity to manage potentially contrasting outcomes of enemy diversity and ecosystem service provision in agricultural landscapes. Theoretical exploration of landscape-scale pest control mechanisms yielded further insights. In a spatially-explicit model, the consequences for pests of intraguild predation (IGP) between enemies were found to depend on landscape complexity. In landscapes with high amounts of seminatural habitat and low spatial autocorrelation (high configurational complexity), IGP by vertebrate predators led to a release of shared herbivorous prey, as predicted by IGP theory. However, in landscapes with low amounts of habitat or high autocorrelation, IGP led to increased predation pressure on herbivorous prey. This result is explained by differences in the local stability of mesopredators across landscapes. In simple landscapes with locally unstable mesopredators, top predators stabilized predator-prey interactions and led to higher predation pressure than with mesopredators alone. These findings confirm empirical results of this thesis showing that landscape complexity may alter the outcome for pests of trophic interactions between natural enemies. Implications for empirical studies of predator-prey interactions include the importance of measuring turnover dynamics of predators and prey at large spatial scales, and show that effective implementation of landscape-wide biological pest control may ultimately depend on the relative population stability across landscapes of pests, arthropod and vertebrate natural enemies.

Level thesis:degree_level

thesis.doctoral

Grantor dc:publisher

Universität Bayreuth

Year

2014

Author dc:creator

• Martin, Emily

Contributors dc:contributor

• Reineking, Björn

Repository record source_url

https://epub.uni-bayreuth.de/id/eprint/55/

OAI identifier oai:identifier

oai:epub.uni-bayreuth.de:55