dc.description.abstracteng | Global biodiversity is increasingly found in fragmented landscapes and populations due to altered human land use. Expansion of agricultural land and changed agricultural practices to the detriment of natural and semi-natural habitats are the most important drivers of this process, caused by the rocketing human demand for agricultural products. Ensuring the persistence of biodiversity and the services it provides to humanity (management of biodiversity and ecosystem services) will require a sound understanding about how biological communities and ecological processes in fragmented landscapes are affected by potentially interacting factors at different spatial scales. However, it remains unclear how properties of fragmented landscapes (landscape composition, landscape configuration, local habitat quality) interact across spatial scales in shaping community structure and ecological processes.
In this thesis I assess the relative role of landscape composition, landscape configuration and local habitat quality at different spatial scales for the conservation of biodiversity and ecosystem processes in fragmented human modified landscapes. My research had two geographical foci: In the first part, I capitalized on the well-known fauna and flora of a traditional, highly fragmented central European agro-ecosystem, to investigate general and trait-mediated patterns across a wide taxonomic range. In the second and third part, I studied bird communities and animal-mediated pollination in the tropics of Central America, which are highly biodiverse but currently suffer an intensive period of human modification.
In the first paper of this study (chapter 2), I used small calcareous grasslands in Germany as our model system, a highly biodiverse, but threatened habitat. Using a large dataset with more than 600 species of nine taxa (bees, butterflies, grasshoppers, hoverflies, leafhoppers, rove beetles, spiders, true bugs, and plants), I tested the separate effects of habitat connectivity, landscape complexity and local management across taxa. In particular, we assessed species richness, community composition and universal, trait-mediated responses. While a high proportion of arable land resulted in a 29% loss of species richness, increasing connectivity generally enhanced species richness across taxa. Only the large species per taxon, assumed to be more dispersive, profited from increased connectivity. While all three management types led to distinct communities, prolonged grazing reduced species richness and abundance, in particular so for red-listed species. In conclusion, we suggest a strategy of alternating mowing and short-term abandonment, focusing on connected sites in diverse landscapes for the conservation of small calcareous grasslands.
The second paper (chapter 3) focuses on the effect of habitat configuration on animal-mediated pollination in tropical fragmented forest landscapes. In Costa Rica, I combined manipulative experiments and field observations to test if narrow woody strips (living fencerows and narrow riparian strips) enhanced pollinator movement, pollen transfer, pollinator availability and pollination success. Using experimental flowers and live plants, I show that such corridors consistently enhance functional connectivity for forest-associated hummingbirds, and in turn pollen transfer between neighboring forest fragments. Corridors drastically increased forest-associated pollinator availability in small fragments, which approached zero in equally sized patches lacking such connections. In parallel, corridors not only substantially increased pollination success of an ornithophilous keystone herb, but averted complete pollination breakdown in small forest fragments. Overall, these results suggest that simple corridors elements can maintain pollination mutualisms and plant gene in tropical forest fragments through increased functional connectivity and pollinator availability.
In the third paper (chapter 4), I evaluated the effect of landscape composition (landscape wide amount and type of forest) and configuration (fragment size and edge proximity) on bird communities, disentangling the effect of old growth forest from secondary forest in a human-modified landscape of southern Costa Rica. I characterized the entire bird community in 49 forest fragments, representing independent variation of patch size and landscape wide forest amount, and found that α-diversity and abundance at the plot scale varied little between fragments. In contrast, α - diversity and abundance of forest birds at the plot scale, in particular insectivores, was nearly halved in edges and secondary forest compared to core areas of primary forest. The same was found in small fragments, but only below a critical threshold of old growth forest within the landscape (22.6%). Similarly, β - diversity of the entire bird community was strongly reduced among small fragments, but only at low landscape-level percentage of oldgrowth forest. Finally, bird communities were similar to primary forest only in fragments surrounded by a high proportion of old growth forest. Overall, the minimum local area (fragment size) required to support substantial levels of α - and β - diversity is lowered in landscapes with a high proportion of old growth forest. Thus, the benefits of old growth not only manifest locally but extend into the landscape scale - not the local fragment area, but the percentage of oldgrowth forest within a landscape drives biodiversity conservation success. Given that human land use is driving landscape structure, there is a need to keep a minimum amount of primary forest - thus, the conservation of old growth forest should remain a key conservation priority.
In conclusion, the results of my dissertation show that biodiversity patterns are shaped by local habitat characteristics, and habitat composition and configuration at the landscape scale. Landscape configuration (corridors) strongly influenced movement between fragments, which in turn affected an ecosystem process (pollination). While small fragments can contribute significantly to conservation in the tropics, there is a need to keep old growth forest above a critical landscape-level threshold. My results emphasize that future conservation of biodiversity and ecosystem services in fragmented landscapes should not only aim to improve local habitat quality, but additionally improve habitat quality and configuration at an appropriate landscape scale. Conservation schemes lacking such a landscape perspective will likely fail to achieve an optimum conservation outcome and thus waste the limited resources available. | de |