|
|
|
 |
|
|
Presentation
< back to presentation overview
| Details of the presentation |
| Presentation |
Oral presentation |
| Title |
From field findings on biodiversity to landscape-scale distributions and change over time – Ants and birds in Kakamega Forest
|
| PDF Download |
--- |
| Short title |
Spatial extrapolation of biological field findings
|
| Author(s) |
Lung, T.(1); Peters, M.K.(2); Farwig, N.(3); Sajita, N.(3); Böhning-Gaese, K.(3); Schaab, G.(1)
|
| Presenting author |
Lung, T.
|
| Institution(s) |
(1) Faculty of Geomatics, Karlsruhe University of Applied Sciences; (2) Zoologisches Forschungsmuseum Alexander Koenig, Bonn; (3) Institute of Zoology, University of Mainz
|
| Keywords |
Forest cover change; spatial extrapolation; army ants; bird communities; GIS and remote sensing
|
| Abstract |
Geospatial data is not only useful for complementing biological field observations recorded at georeferenced points, transects or plots, but it is required for extrapolating field findings in space and time. For Kakamega Forest, a spatially explicit time series on land cover change over the past 90 years has been derived from satellite imagery, aerial photography and old topographic maps enabling investigations of the anthropogenic influence on the forest’s biodiversity. It forms the basis for extrapolations of so far three different sets of biological field inventories: of army ants, ant-following birds and the overall bird community. Regarding the first two sets of observations, spatial information on the land cover and deduced small-scale forest fragmentation/disturbance for the close surroundings of the field observations’ locations is used to test for correlations in a regression model. Based on the determined relationships and applying a moving window, army ant abundance is modelled considering the proportion of forest cover in the close vicinity of each pixel of the modelling domain. Point counts of ant-following birds are extrapolated based on a spatially explicit forest fragmentation index and the simulated army ant distribution. As the overall bird community of forest specialists, forest generalists and forest visitors was monitored in five different forest types, here, analyses are directly linked to the forest classes as distinguished in the land cover time series. The final extrapolation results are landscape-scale distributions for 5 time steps between 1913 and 2003, each covering the entire study area. As such, the extrapolations allow for a forest-wide quantitative assessment regarding the influence of forest decline and large-scale fragmentation on army ants and birds. A scenario linked to the army ants reveals the need for the strictly protected forest areas to remain connected with near natural forest in order to sustain viable populations. Another scenario helps to understand how the reforestation of currently non-forest areas inside the official forest boundaries with mixed indigenous plantations would affect the overall bird community. Our findings thus contribute to recommendations for forest conservation and management sustaining the high biodiversity of Kakamega Forest.
|
| Congress Topic |
Process Analysis
|
| Topic No. |
2.6 |
| Notes |
---
|
| Ref. No. |
545 |
|
|
|
