Ruud P. Bartholomeus

Plant strategies in relation to resource supply in mesic to wet environments: does theory mirror nature?

In ecology, strategy schemes based on propositions about the selection of plant attributes are common, but quantification of such schemes in relation to nutrient and water supply is lacking. Through structural equation modeling, we tested whether plant strategies related to nutrient and water/oxygen supply are reflected in a coordination of traits in natural communities. Structural equation models, based on accepted ecological concepts, were tested with measured plant traits of 105 different species across 50 sites in mesic to wet plant communities in the Netherlands. For each site, nutrient and water supply were measured and modeled. Hypothesized multivariate strategy models only partly reflected current theoretical schemes. Alternative models were consistent, showing that lack of consistency of the original models was because of (i) strong correlations among traits that supposedly belong to different strategy components; (ii) poor understanding of mechanisms determining the covariation of plant maximum height, leaf size, and stem density; and (iii) lack of integrative and long‐term measures of nutrient supply needed to predict coordinated plant trait responses. Our main conclusion is that a combination of trade‐offs (partly) across different plant organs and diverging effects of resource supply ultimately determines the coordination of plant traits needed to “make a living.”

Leaf habit and woodiness regulate different leaf economy traits at a given nutrient supply

The large variation in the relationships between environmental factors and plant traits observed in natural communities exemplifies the alternative solutions that plants have developed in response to the same environmental limitations. Qualitative attributes, such as growth form, woodiness, and leaf habit can be used to approximate these alternative solutions. Here, we quantified the extent to which these attributes affect leaf trait values at a given resource supply level, using measured plant traits from 105 different species (254 observations) distributed across 50 sites in mesic to wet plant communities in The Netherlands. For each site, soil total N, soil total P, and water supply estimates were obtained by field measurements and modeling. Effects of growth forms, woodiness, and leaf habit on relations between leaf traits (SLA, specific leaf area; LNC, leaf nitrogen concentration; and LPC, leaf phosphorus concentration) vs. nutrient and water supply were quantified using maximum-likelihood methods and Bonferroni post hoc tests. The qualitative attributes explained 8-23% of the variance within sites in leaf traits vs. soil fertility relationships, and therefore they can potentially be used to make better predictions of global patterns of leaf traits in relation to nutrient supply. However, at a given soil fertility, the strength of the effect of each qualitative attribute was not the same for all leaf traits. These differences may imply a differential regulation of the leaf economy traits at a given nutrient supply, in which SLA and LPC seem to be regulated in accordance to changes in plant size and architecture while LNC seems to be primarily regulated at the leaf level by factors related to leaf longevity.

The need of data harmonization to derive robust empirical relationships between soil conditions and vegetation

Abstract Question: Is it possible to improve the general applicability and significance of empirical relationships between abiotic conditions and vegetation by harmonization of temporal data? Location: The Netherlands. Methods: Three datasets of vegetation, recorded after periods with different meteorological conditions, were used to analyze relationships between soil moisture regime (expressed by the mean spring groundwater level – MSLt calculated for different periods) and vegetation (expressed by the mean indicator value for moisture regime Fm). For each relevé, measured groundwater levels were interpolated and extrapolated to daily values for the period 1970–2000 by means of an impulse-response model. Sigmoid regression lines between MSLt and Fm were determined for each of the three datasets and for the combined dataset. Results: A measurement period of three years resulted in significantly different relationships between Fm and MSLt for the three datasets (F-test, p < 0.05). The three regression lines only coincided for the mean spring groundwater level computed over the period 1970–2000 (MSLclimate) and thus provided a general applicable relationship. Precipitation surplus prior to vegetation recordings strongly affected the relationships. Conclusions: Harmonization of time series data (1) eliminates biased measurements, (2) results in generally applicable relationships between abiotic and vegetation characteristics and (3) increases the goodness of fit of these relationships. The presented harmonization procedure can be used to optimize many relationships between soil and vegetation characteristics.

Synthesis of ecosystem vulnerability to climate change in the Netherlands shows the need to consider environmental fluctuations in adaptation measures

Climate change impacts on individual species are various and range from shifts in phenology and functional properties to changes in productivity and dispersal. The combination of impacts determines future biodiversity and species composition, but is difficult to evaluate with a single method. Instead, a comparison of mutually independent approaches provides information and confidence in patterns observed beyond what may be achieved in individual approaches. Here, we carried out such comparison to assess which ecosystem types in the Netherlands appear most vulnerable to climate change impacts, as arising from changes in hydrology, nutrient conditions and dispersal limitations. We thus combined meta-analyses of species range shifts with species distribution modelling and ecohydrological modelling with expert knowledge in two respective impact studies. Both impact studies showed that nutrient-poor ecosystems and ecosystem types with fluctuating water tables—like hay meadows, moist heathlands and moorlands—seem to be most at risk upon climate change. A subsequent meta-analysis of species–environmental stress relations indicated that particularly endangered species are adversely affected by the combination of drought and oxygen stress, caused by fluctuating moisture conditions. This implies that adaptation measures should not only aim to optimise mean environmental conditions but should also buffer environmental extremes. Major uncertainties in the assessment included the quantitative impacts of vegetation-hydrology feedbacks, vegetation adaptation and interactions between dispersal capacity and traits linked to environmental selection. Once such quantifications become feasible, adaptation measures may be tailor-made and optimised to conserve vulnerable ecosystem types.