Abstract: The fields of plant water relations and plant biomechanics have traditionally been studied separately even though often the same tissues are responsible for water transport and mechanical support. There is now increasing evidence that hydraulic and mechanical adaptations may influence one another. We studied the changes in the hydraulic and mechanical properties of the wood along lateral roots of two species of buttressed trees. In these roots, the mechanical contstraints quantified by strain measurements are known to decrease distally. Further, we investigated the effect of mechanical loading on the vessel anatomy in these and four other species of tropical trees. We found that as the strain decreased, the wood became progressively less stiff and strong but the conductivity increased exponentially. This was reflected in that adaptations towards re-enforcing mechanically loaded areas resulted in xylem with fewer and smaller vessels. In addition a controlled growth experiment on three tree species showed that drought adaptation may results in plants with stronger and stiffer tissue. Our results indicate that hydraulic and mechanical stress adaptations may be interrelated, and so support recent studied suggesting that physiological responses are complex balances rather than pure optimisations. ©2008 Landes Bioscience.

Abstract: Sodium, an element which is needed by animals but often toxic in high concentrations to plants, may be deficient and limit animal abundance in inland continental regions, but may be overabundant and limit plant productivity in coastal regions. Here we present data from 50 independent plots (including leaf data from more than 2480 individual trees) showing that leaves in the Amazon basin uptake high amounts of sodium (Na) in a manner more similar to the essential cation potassium (K) than to the toxic cation aluminium (Al). Leaf Na increases linearly with soil Na concentrations, and there is no apparent mechanism for selective exclusion of Na in comparison to K, a key attribute of halophytes. This indicates that the Amazon basin is broadly non-halophytic and increased sodium concentrations in non-halophyte plants often decrease plant productivity. Total Na concentrations are ∼ 10 times higher in coastal regions than inland regions. Such concentration gradients in nutrients may have been reduced in the past because large animals that were abundant in the Pleistocene have been hypothesized to play a large role in reducing nutrient concentration gradients at continental scales. We use a diffusion model and a Na loss rate based on empirical data to estimate that large animals may have moved significant quantities of Na inland away from coastal regions in the Amazon Basin. Therefore, our simple model suggests that large animals may play an important, yet diminishing, role in maintaining the sodium balance of the planet. © 2016 Nordic Society Oikos.

Abstract: Resource control over abundance, structure and functional diversity of soil microbial communities is a key determinant of soil processes and related ecosystem functioning. Copiotrophic organisms tend to be found in environments which are rich in nutrients, particularly carbon, in contrast to oligotrophs, which survive in much lower carbon concentrations. We hypothesized that microbial biomass, activity and community structure in nutrient-poor soils of an Amazonian rain forest are limited by multiple elements in interaction. We tested this hypothesis with a fertilization experiment by adding C (as cellulose), N (as urea) and P (as phosphate) in all possible combinations to a total of 40 plots of an undisturbed tropical forest in French Guiana. After 2 years of fertilization, we measured a 47% higher biomass, a 21% increase in substrate-induced respiration rate and a 5-fold higher rate of decomposition of cellulose paper discs of soil microbial communities that grew in P-fertilized plots compared to plots without P fertilization. These responses were amplified with a simultaneous C fertilization suggesting P and C colimitation of soil micro-organisms at our study site. Moreover, P fertilization modified microbial community structure (PLFAs) to a more copiotrophic bacterial community indicated by a significant decrease in the Gram-positive : Gram-negative ratio. The Fungi : Bacteria ratio increased in N fertilized plots, suggesting that fungi are relatively more limited by N than bacteria. Changes in microbial community structure did not affect rates of general processes such as glucose mineralization and cellulose paper decomposition. In contrast, community level physiological profiles under P fertilization combined with either C or N fertilization or both differed strongly from all other treatments, indicating functionally different microbial communities. While P appears to be the most critical from the three major elements we manipulated, the strongest effects were observed in combination with either supplementary C or N addition in support of multiple element control on soil microbial functioning and community structure. We conclude that the soil microbial community in the studied tropical rain forest and the processes it drives is finely tuned by the relative availability in C, N and P. Any shifts in the relative abundance of these key elements may affect spatial and temporal heterogeneity in microbial community structure, their associated functions and the dynamics of C and nutrients in tropical ecosystems.

Abstract: The elucidation of relationships between biodiversity and ecosystem processes has been limited by the definition of metrics of biodiversity and their integration into experimental design. Functional trait screening can strengthen the performance of these designs. We suggest the use of Rao's quadratic entropy to measure both functional diversity and phylogenetic diversity of species mixtures proposed for an experimental design, and demonstrate how they can provide complementary information. We also present an index assessing the statistical performance of these independent variables in different experimental designs. Measurement of independent variables as continuous vs. discrete variables reduces statistical performance, but improves the model by quantifying species differences masked by group assignments. To illustrate these advances, we present an example from a tropical forest tree community in which we screened 38 species for nine functional traits. The proposed TropiDEP design is based on the relative orthogonality of two multivariate trait axes defined using principal component analysis. We propose that independent variables describing functional diversity might be grouped to calculate independent variables describing suites of different traits with potentially different effects on particular ecosystem processes. In other systems these axes may differ from those reported here, yet the methods of analysis integrating functional and phylogenetic diversity into experimental design could be universal.

Abstract: We investigated in the laboratory the initial behavior of propagules of the invasive ant Wasmannia auropunctata in Cameroon where it has been introduced. Both workers and queens at first feigned death (thanatosis), and then the workers slowly moved around the experimental arena; the queens did the same about 10 seconds later. Each queen antennated selected workers that then aggregated together by grasping the hind leg of another ant with their mandibles. When encountering the queen again, the lead worker climbed up the queen's hind leg and onto her back, followed by some other individuals. The remaining workers followed the queen to a location in the experimental arena. When brood was present, the workers transferred it to this location. Orphaned workers did not aggregate, but gathered the brood together and took care of it. By permitting propagules to survive, these behaviors likely contribute to the success of W. auropunctata as an invader. (C) 2011 Academie des sciences. Published by Elsevier Masson SAS. All rights reserved.