Figure 1.
Roof construction and arrangement of the subplots in the field site.
Given are the different subplots and the size of subplots and the roof construction. For more details see main text.
Figure 2.
Soil moisture and daily precipitation patterns during the period of induced drought.
In summer of 2009 (left) and 2010 (right). Soil moisture data are shown for all three roof treatments (lines, average of N = 3 plots). Daily precipitation patterns (grey bars) were measured on the field site of the Jena Experiment.
Table 1.
Climatic parameters measured on the field site of the Jena Experiment during the study years 2009 and 2010 with the reference period 1961–1990 measured by the German Weather Service DWD in the city center of Jena.
Figure 3.
Effects of the presence of roofs on abiotic parameters: air temperature (a, b), soil temperature (c) during day (circles) and night (triangles) and photosynthetically active radiation (d).
Given are means and standard errors of the drought treatment (filled symbols, solid lines), unroofed (open symbols, short dashed line) and roofed controls (x symbols, long dashed line) for day (circles) and night (triangles). Data represent mean and standard error of all three treatments in four (respectively three in case of temperature) plots.
Figure 4.
Treatment effects on ecosystem properties.
Aboveground biomass production (a, measured in 2009 and b, measured in 2010) and litter decomposition (c). Data represent mean and standard error of all three treatments in 76 plots.
Table 2.
Summary of mixed effects models for aboveground biomass in August 2009 and 2010 as well as for decomposed wheat litter.
Figure 5.
Partial least square discriminant analysis (PLS-DA) of metabolite profiles.
(combined data of flowers, sink leaves and source leaves) with one individual (Medicago x varia) analyzed for each subplot out of seven plots. Metabolites correlating significantly with PLS-DA data (according to Monte Carlo-permutations, p<0.05) are represented as black arrows. Only identified metabolites are shown. These metabolites comprise: asparagine/flower (1), arabitol/flower (2), arabitol/sink leaf (3), citric acid/sink leaf (4), allantoin/flower (5), pinitol/sink leaf (6), asparagine/source leaf (7), arabitol/source leaf (8), 1,6-anhydro-glucose/source leaf (9), sorbitol/source leaf (10), 2,4-diamino-butanoic acid/source leaf (11), glucose/source leaf (12), erythritol/source leaf (13), beta-alanine/sink leaf (14), xylitol/source leaf (15), kestose/sink leaf (16), galactinol/flower (17), mannose/sink leaf (18), glucose-6-phosphate/sink leaf (19), phenylalanine/source leaf (20), phosphoric acid monomethyl ester/sink leaf (21), threonine/sink leaf (22). Retention time index and fragmentation pattern were not sufficient to differentiate between closely related isomers in the case of arabitol, xylitol, 1,6-anhydro-glucose, sorbitol, kestose, galactinol and mannose.