Persistence-related plant traits
During the spring and summer of 2019, we sampled three well-developed and healthy individual plants per species at each edaphic island (total sample size = 538; except in a few instances when only two individuals were available). Trait-data collection and measurement procedures followed standard protocols (Klimešová et al., 2019; Pérez-Harguindeguy et al., 2013). We collected seven persistence-related traits linked to non-acquisitive functions of plants (Table 1). In the field, we scored maximum plant height and lateral spread (maximum distance between offspring individuals linked to the parent plant through rhizomes and/or stolons; for clonal species only). For the other five traits, we collected plant material in the field to be processed in the laboratory afterwards. For belowground organ dry matter content (BDMC), a ~2 cm-long portion of either rhizome (for clonal species) or thick root (for non-clonal species) was cut and put in a paper bag inside a sealed plastic bag. Fresh weight was recorded, then the plant material was oven-dried for 72 hours at 60°C, and the dry weight measured. BDMC was calculated as the ratio between the oven-dry and fresh mass (similarly to leaf dry matter content [LDMC]; Pérez-Harguindeguy et al., 2013).
Plant anatomical traits – i.e. age, radial growth, storage tissue, vessel size (Table 1) – were measured for non-clonal species only (346 individuals analyzed overall). In the field, we cut a ~2 cm-long fragment located between the root and stem system from each sampled individual. This is indeed the oldest plant part present in the studied growth forms, therefore enables the best estimate of plant age (Klimešová et al., 2019). These fragments were preserved in 50% ethanol, and once in the laboratory, sectioned using a sledge lab–microtome (Gärtner et al., 2015), with thickness between 15 and 40 µm. Cross-sections were then double-stained using a mixture of Astrablue and Safranin dye, dehydrated (with a series of solutions having different ethanol concentrations), washed with xylene and fixed on slides with Canada balsam. These slides were examined at the microscope, and ImageJ software (Schneider et al., 2012) was used to evaluate the number of the annual rings (age), mean annual radial growth, percentage of storage (parenchyma) tissue type, and maximum vessel size in the cross-section.
TABLE 1 Name (indicating if the trait was collected for clonal, non-clonal species, or both), definitions (with variable type, units), relevant functions and key references of the non-acquisitive persistence-related traits included in this study.