Untersuchungen zu Spinnengemeinschaften in Kakaoagroforstsystemen in Indonesien: Diversität, Netzdichte und räumlich-zeitlicher Artenaustausch

Primary rainforests and the associated biodiversity are disappearing from tropical landscapes at accelerating rates due to agricultural expansion. Cacao agroforestry systems are commonly established at the margin of natural forests in Indonesia and constitute a habitat with a considerable tree cover, which may benefit biodiversity conservation and ecosystem services in addition to providing income for the local population. However, current management is dominated by severe habitat simplification due to extensive weeding, removal natural forest trees, planting one or few shade-tree species or even a conversion into intensive full-sun monocultures. In two valleys at the margin of the Lore Lindu National Park in Central Sulawesi, Indonesia, we investigated the factors shaping spider communities in cacao agroforests. We collected spider data in a total of 42 differently managed cacao agroforestry sites. First, spider web densities of five web-building spider guilds on 10 cacao trees per plot were analysed with respect to tree and plot management (e.g. pruning, shade cover and weeding), landscape features (plot elevation, forest distance) and the presence of a numerically dominant Philidris ant species. Secondly, based on a continuous one year multi-strata sampling of spiders within a subset of 12 agroforestry sites, we investigated drivers of abundance and species richness of web-building and hunting spiders of litter, herb and canopy layer at the plot and landscape scale. Thirdly, we analysed how spatial and temporal changes in species richness are affected by different habitat structures, such as shade tree cover and weed diversity, or by landscape context such as distance to forest. On 420 cacao trees we found 5147 spider webs. Densities of webs of different types were differentially affected by habitat variables, while most frequently recorded webs (iO 42 %) were of the line web type, a simple construction of a few filaments. The orb web weaving spider guild, including Araneidae, Tetragnathidae and Uloboridae, was also abundant on cacao trees (iO 30 %). Our results suggest spider web density could be increased by pruning of cacao trees while keeping shade trees density in cacao plots. The presence of Philidris ants was positively associated with density of orb webs, while no pattern was found for other web types. Analysis of the one-year sampling within 12 agroforestry sites showed that canopy and litter communities were dominated by hunting spiders and the herb communities by web-building spiders. In general, species overlap between strata was low and herb layer communities exhibited highest variability in community composition. Spider communities of all habitat strata were positively affected by weed cover, while an increase in leaf litter density was accompanied by a decrease in spider abundance, which appeared to be due to previous pesticide treatments reducing litter decomposition and populations of potential prey. On a community level, we found no effect of forest edge proximity on spider abundance or species richness, but species-specific responses showed that with distance to forest there seems to be a shift of species dominance in favour of generalists. The analysis of temporal and spatial turnover revealed that the number of species occurring locally depends to a similar degree on spatial and temporal patterns, although interannual turnover differs between canopy and litter dwelling spiders. Both temporal and spatial turnover were affected mainly by habitat-scale variables, with no significant effect of forest distance. The largest effect was the positive relation of temporal species turnover to shade tree diversity. These results showed not only that snap-shot or aggregated species richness measures may not adequately represent diversity patterns through time, but that ignoring turnover may result in bias when evaluating the effect of management or landscape context. It is suggested that more attention should be given to the effect of spatial and temporal turnover on ecological effects of spider communities, such as predation. More studies including temporal aspects of species communities are required in conjunctions to biodiversity conservation but also to biological control within agroforestry systems.