Do short-lived ruderal and arable weed communities reflect regional climate differences? A case study from SE Styria

The Feldbach district is situated near the town of Feldbach in the southeastern province of Styria, Austria. Since 2007, a climate observation network of 151 climate stations within an area of approximately 20 x 15 km (grid cell 1.4 x 1.4 km) has provided a unique collection of spatial and temporal meteorological data. Examining short-lived ruderal and arable weed communities, we search for correlations and spatial patterns between the community composition and climate data. Do such plant communities respond to temperature differences within an agricultural landscape and on a regional scale? Data of 277 releves from short-lived weed communities were collected across the investigated area during the summer of 2011. Releves were assigned to the corresponding climate stations and classified. Average Ellenberg indicator values for temperature were calculated for each releve and community cluster. Measured temperature data were assigned and correlated with community data by applying linear regression and redundancy analyses (RDA). The classification resulted in six associations; the two most frequently observed associations were divided into subtypes resulting in 13 vegetation clusters that could be analyzed further. A significant relationship could be found only between the clusters of arable weed communities and the average winter temperatures. Site variables explain twice as much variance as measured climate variables; this ratio changes to 50 : 50 when we analyzed only arable field community data. No clear spatial patterns concerning mean annual temperature were visible. However, the Setaria faberi subtype of the Echinochloo-Setarietum and the Sorghum halepense subtype of the Convolvulo-Agropyretum show a tendency toward a temperature-induced spatial pattern, such that both were sensitive to winter temperature. On a regional scale, the occurrence and composition of short-lived ruderal plant communities correlated weakly with climate variables. However, the studied arable weed communities showed a certain tendency to follow small-scale temperature differences, especially those of average winter temperature. We conclude that short-lived weed communities have the potential to be indicators for global warming, but the spatial temperature gradients are not clear enough in our approach to allow the production of better regression models and elucidation of distinct spatial patterns.