Forest Fragmentation in Space and Time - New perspectives from remote sensing and forest modelling

Bitte benutzen Sie diese Kennung, um auf die Ressource zu verweisen:
Open Access logo originally created by the Public Library of Science (PLoS)
Titel: Forest Fragmentation in Space and Time - New perspectives from remote sensing and forest modelling
Autor(en): Dantas de Paula, Mateus
Erstgutachter: Prof. Dr. Andreas Huth
Zweitgutachter: Prof. Dr. Horst Malchow
Zusammenfassung: Empirical studies on severely fragmented regions suggest that decades after fragmentation, forest edges located near human-modified areas exhibit the structure of early successional states, with lower biomass per area and higher mortality compared to non-edge areas. These habitat changes (edge effects) can also have a considerable impact on ecosystem processes such as carbon and water balance, which in turn have a major impact on human activities. Also, the disruption of ecological interactions caused by the loss of animals (defaunation) has the potential of impacting human influenced fragmented landscapes much deeper than only through the microclimate-induced increased tree mortality caused by edge effects. Since large animals are most vulnerable in these landscapes, large tree seeds which are dispersed by them become more vulnerable to pre-dispersal seed predation, reducing tree recruitment in latter stages. Also, the loss large animals which predate on smaller animals can cause relaxation of top-down controls of this small seed-eating animal group, further impacting tree recruitment. Even though detailed and long term studies were developed on the topic of edge effects at local scale, understanding edge effect characteristics in fragmented forests on larger scales and finding indicators for its impact is crucial for predicting habitat loss and developing management options. Using field data from a long-term fragmented landscape in the Brazilian Northeastern Atlantic Forest, and the Forest Model FORMIND, we were able to visualize the time scale in which edge effects influence tropical forests by performing 500-year simulations. We observed changes in community composition, aboveground biomass, total evapotranspiration and total runoff, and evaluate the consequences of defaunation on biomass retention of a Brazilian Northeastern Atlantic Forest tree community by varying pre- and post-dispersal seed predation pressures in fragmented and intact scenarios. Finally, we evaluate the spatial and temporal dimensions of edge effects in large areas using remote sensing by using tree cover as an indicator of habitat quality and in relation to edge distance. FORMIND simulations show forest biomass degradation lasting around 100 years. If edge effects cease, recovery of biomass lasts around 150 years. Carbon loss is especially intense during the first five years after fragmentation, resulting in a decline of over 5 Mg C ha−1 y−1. Finally, edges of large fragments face an evapotranspiration loss of 43% and total runoff gains of 57% in relation to core areas of large fragments. The effects of large seed loss are only notable after 80% seed reduction or 10 times higher predation rates, but can cause the extirpation of this species group and up to 29% less biomass retention for the area. Our remote sensing results show that for all 11 LANDSAT scenes pixel neighborhood variation of tree cover is much higher in the vicinity of forest edges in relation to forest interior. Our studies suggest that fragmented landscapes can be of significantly lower value in terms of ecosystem services, and that defaunation has the potential to reduce biomass retention and species richness through dispersal collapse. Satellite based estimations of tree cover at edges suggest a maximum distance for edge effects and can indicate the location of unaffected core areas. However, tree cover patterns in relation to fragment edge distance vary according to the analyzed region, and maximum edge distance may differ according to local conditions.
Schlagworte: Forest Fragmentation; Waldfragmentierung; Ecological Modelling; Ökologisch Modellierung; Remote Sensing; Fernerkundung
Erscheinungsdatum: 17-Jul-2017
Lizenzbezeichnung: Namensnennung-Keine Bearbeitung 3.0 Unported
URL der Lizenz:
Publikationstyp: Dissertation oder Habilitation [doctoralThesis]
Enthalten in den Sammlungen:FB06 - E-Dissertationen

Dateien zu dieser Ressource:
Datei Beschreibung GrößeFormat 
thesis_dantas_de_paula.pdfPräsentationsformat8,99 MBAdobe PDF

Diese Ressource wurde unter folgender Copyright-Bestimmung veröffentlicht: Lizenz von Creative Commons Creative Commons