Impacts of weeding and fertilisation management on arthropod communities in oil palm plantations

Research Objectives
The proposed Master thesis will be conducted within subproject B09 (Tscharntke/Grass) “Aboveground patterns of biodiversity and associated ecosystem processes across tropical rainforest transformations” within CRC 990. Specifically, I will study the effects of different oil palm management practices (fertilization, weeding) on insect diversity. The studies will be conducted within the “Oil Palm Management Experiment”, established at PTPN VI. The overall aim of this study is to identify oil palm management that enables a high insect diversity without affecting oil palm yields.

Methodology and concept
The Oil Palm Management Experiment is a full-factorial management experiment with 4 sites comprising 4 plots each (16 plots in total). Plots are 50 × 50 m in size and are established with all combinations of the following fertilizer and weeding treatments: conventional fertilization (260 N, 50 P, 220 K kg/ha/yr) vs. reduced fertilization (136 N, 17 P, 187 K kg/ha/yr) and conventional weeding with herbicide application (Glyphosate) vs. mechanical weeding (without herbicide). First, the undergrowth vegetation on all plots will be described (plant species composition, flowering plant community). Second, insects will be caught on all plots, with different trapping methods that target different insect groups.
The proposed insect trapping methods are pan traps, sweep netting and vane traps. Pan traps attract flying insects such as bees, wasps (Hymenoptera) and flies (Diptera). In each plot, 6 pan traps will be placed for 48 h. Sweep netting of the undergrowth vegetation will be conducted to capture those insects that live within or feed on these plants (e.g., Hemiptera or Coleoptera). Sweep netting will be done 3–4 times in every plot during the study period, with about 30 min sampling effort per plot. Vane traps can be used to attract flying insects, but in contrast to pan traps they are not placed on the ground but can be placed within trees or palm at canopy height. I plan to expose 3 vane traps in each plot for a period of 4 days. By trapping in the canopy of oil palm, potential pollinators of oil palm can be identified. The trapping will not interfere with the natural pollination of the palm by the oil palm weevil; instead, it aims at identifying other potential pollinators that are so far poorly studied (e.g., bees visiting oil palm flowers).
All insects will be identified to the most specific taxonomic level as possible. Insects do not need to be exported; all identification can be done at the University of Jambi or Bogor Agricultural University. All insects will remain for permanent storage in Indonesia.

Research location
The research will be implemented in the Jambi Province in Sumatra, with the fastest and most complete transformation of tropical lowland rainforest. We focus on tropical lowland rainforests as these are experiencing the strongest losses worldwide (Achard et al. 2002). In particular, the massive transformation of lowland rainforest into oil palm plantations has been identified as a major threat for biodiversity and a potential driver of climate change (Danielsen et al. 2009, Giam et al. 2010, Wilcove and Koh 2010).
This study will be conducted within the framework of the Oil Palm Management Experiment located in PTPN VI.