Spatial variation in environmental conditions can lead to local adaptation of plant populations, particularly if gene flow among populations is low. Many studies have investigated adaptation to contrasting environmental conditions, but little is known about the spatial scale of adaptive evolution. We studied population differentiation and local adaptation at two spatial scales in the monocarpic grassland perennial Carlina vulgaris. We reciprocally transplanted seedlings among five European regions (northwestern Czech Republic, central Germany, Luxembourg, southern Sweden and northwestern Switzerland) and among populations of different sizes within three of the regions. We recorded survival, growth and reproduction over three growing periods. At the regional scale, several performance traits and the individual fitness of C. vulgaris were highest if the plants were grown in their home region and they decreased with increasing transplant distance. The effects are likely due to climatic differences that increased with the geographical distance between regions. At the local scale, there were significant interactions between the effects of the population of origin and the transplant site, but these were not due to an enhanced performance of plants at their home site and they were not related to the geographical or environmental distance between the site of origin and the transplant site. The size of the population of origin did not influence the strength of local adaptation. The results of our study suggest that C. vulgaris consists of regionally adapted genotypes, and that distance is a good predictor of the extent of adaptive differentiation at large scales ( > 200 km) but not at small scales. We conclude that patterns of local adaptation should be taken into account for the efficient preservation of genetic resources, when assessing the status of a plant species and during conservation planning.

No abstract

The drivers of plant richness at fine spatial scales in steppe ecosystems are still not sufficiently understood. Our main research questions were: (i) How rich in plant species are the natural steppes of Southern Siberia compared to natural and semi-natural grasslands in other regions of the Palaearctic? (ii) What are the main environmental drivers of the diversity patterns in these steppes? (iii) What are the diversity–environment relationships and do they vary between spatial scales and among different taxonomic groups? We sampled the steppe vegetation (vascular plants, bryophytes and lichens) in Khakassia (Russia) with 39 nested-plot series (0.0001–100-m2 plot size) and 54 additional 10-m2 quadrats across the regional range of steppe types and measured various environmental variables. We measured β-diversity using z-values of power-law species–area relationships. GLM analyses were performed to assess the importance of environmental variables as predictors of species richness and z-value. Khakassian steppes showed both high α- and β-diversity. We found significant scale dependence for the z-values, which had their highest values at small spatial scales and then decreased exponentially. Total species richness was controlled predominantly by heat load index, mean annual precipitation, humus content and soil skeleton content. The positive role of soil pH was evident only for vascular plant species richness. Similar to other studies, we found that the importance of environmental factors strongly differed among taxonomic groups and across spatial scales, thus highlighting the need to study more than one taxon and more than one plot size to get a reliable picture

Fig. 2. Spatial coverage of GrassPlot data from Morocco to Japan. Currently, the majority comes from sub-Mediterranean to hemiboreal Europe (black = multi-scale plots, grey = other plots). Current content v. 1.00 (January 2018) • 126 datasets • 198 data owners • 36 countries • 168,997 plots, among them 14,064 with data also for non-vascular plants • 66,000 0.01-m² plots, 17,206 1-m² plots, 5,520 10-(or 9-) m² plots, 2,545 100-m² plots • 2,797 nested-plot series (with at least 4 grain sizes)

Summary1. When range shifts or invasions of plant species are studied, it is important to know whether large-scale spatial variation in a species' demography can be ignored or approximated by variation observed over smaller spatial scales. 2. Here, we studied the population dynamics of three similar (as shown by elasticity analysis) shortlived perennial plant species in multiple sites in different European countries over 2 years. We constructed a total of 40 transition matrices and analysed the spatio-temporal variation in the projected population growth rate (k) with spatially nested life table response experiments (LTRE). 3. All species (Carlina vulgaris, Tragopogon pratensis and Hypochaeris radicata) showed considerable life-history variation among regions on top of variation among sites within regions. 4. Net variance contributions (NVC), a novel LTRE statistic, revealed that in each species, variation in one group of vital rates contributed most to variation in k among regions as well as among sites. However, that most important type of vital rates differed between species: plant growth in C. vulgaris, flower head production in T. pratensis and establishment probability of seedlings in H. radicata. The rankings of the NVCs of other vital rates varied between site and region effects, suggesting that buffering through negative vital rate correlations varies over different spatial scales, while the identity of the main contributor to k variation is more constant. 5. Temporal effects were smaller than spatial effects, but the LTREs showed strong interactions between time and space (region or site), suggesting that the effect of, e.g. climate fluctuations are not synchronized throughout the distribution of a species. 6. Synthesis. This study shows that the life histories of plant species are distinguishable even when mean elasticity values show only small differences, and that life histories vary over the distribution range of a species. Demographic differences over large spatial scales can therefore only be partly substituted by small scale spatial variation in modelling studies on the population dynamics of a species across its entire distribution.