Can heterogeneous landscapes mitigate pesticide stress?

In this blogpost, Moritz Link reports how pesticide toxicity affects stream ecosystems in Transylvania, Romania, a region where traditional agriculture prevails. The researchers analyzed macroinvertebrate communities, leaf litter decomposition and parasite prevalence in relation to physico-chemical stressors in the stream and the landscape.

Pesticides are omnipresent

In regions such as Western Europe, agriculture was subject to intensification for decades. This led to the removal of non-agricultural structures, which can function as recolonization patches / refuges for organisms. In contrast, non-agricultural structures are preserved in parts of Eastern Europe, such as Transylvania, Romania, where traditional farming practices still prevail. Although the sales of agrochemicals there have been lower than in other parts of Europe, we found pesticide exposure levels similar to areas of intensive agriculture in Western Europe in a companion study. This raises the question – how landscape structures and agricultural stressors, such as pesticides, affect macroinvertebrate communities, leaf litter decomposition and parasite prevalence?

How sampling was conducted?

We conducted a field study on 19 agricultural streams in rural Transylvania, Romania (Figure 1).   We sampled macroinvertebrate communities and analyzed them via various biological metrics, that can provide information on how the community reacts to environmental stressors.Additionally, we quantified the leaf litter decomposition as well as parasite prevalence (Figure 2). Furthermore, we analyzed the landscape composition including the presence of upstream refuge areas and measured a wide range of water quality variables including pesticide toxicity.

Figure 1: Macroinvertebrate sampling in a Romanian stream (picture by Eduard Szöcs).
Figure 2: Preparation of leaf disks for the leaf litter decomposition analysis (picture by Moritz Link).

Our results

The macroinvertebrate community, expressed via different metrics, was best explained by landscape parameters such as upstream refuges or the proportion of agriculture in the catchment. In contrast to previous studies from Western Europe, we did not find a negative relationship between pesticide toxicity and the macroinvertebrate community. This was the case although the macroinvertebrate communities showed clear signs of exposure towards stress. The lack of relationship can probably be explained by the high levels of pesticide exposure in the streams, spanning over a short gradient (Figure 3). Neither the leaf litter decomposition nor the parasite prevalence showed any relationship to the analyzed environmental variables, which indicated that both were tolerant towards the stressors in the study region.

Figure 3: Boxplots of SPEARpesticides values (one of the macroinvertebrate metrics of this study) for sites without and with forested upstream sections (FUS). Horizontal black lines show median of the respective group. Mean SPEARpesticides values differ significantly (Welch’s t-test: t = -3.57, p = 0.003) (graphic by Moritz Link).

Our study suggests that macroinvertebrate communities benefit from recolonization from refuge areas as possible in heterogeneous landscapes. Based on this, landscape management that promotes and reestablishes landscape structures of low intensive land use may benefit stream ecosystems and diverse organism groups through increasing connectivity of suitable habitat patches.


 The paper was authored by Moritz Link, Verena C. Schreiner, Nadin Graf, Eduard Szöcs, MircoBundschuh, Karina P. Battes, Mirela Cîmpean, Bernd Sures, Daniel Grabner, Jörn Buse, Ralf B. Schäfer and published in Science of The Total Environment.

For further information about the project and additional results, see our previous blog posts:

Agriculture without paradises? Low intensity agriculture and pesticide use in Eastern Europe

Do agricultural pesticides in streams influence riparian spiders?

Field study in Transylvania: News from the Landscape Ecology Group