Abstract: Insects, particularly ants, are good bioindicators of the state of ecosystems. Nevertheless, incorporating them into conservation surveys is expensive due to problems associated with their identification, which is exacerbated by the fact that there are fewer and fewer taxonomists working today. “Taxonomic sufficiency” (TS), which identifies organisms to a level of taxonomic resolution sufficient enough to satisfy the objectives of a study, has never been applied to Neotropical ant communities. We analysed five Neotropical datasets representing ant assemblages collected with different sampling methods in various habitats. We first treated them using two complementary and cumulative TS methods, higher-taxon and “indicator taxa” surrogacies, before testing a new approach called “mixed-level method” that combines the two previous approaches. For the higher-taxon surrogacy, we showed that, above species, genus is the most informative taxonomic level. Then, mixed-level method provided more information on ant assemblages than did the two others, even though the “indicator taxa” surrogacy was based on relevant indicator genera. Although habitat type has no effect on its efficiency, this new method is influenced by the dataset structure and the type of sampling method used to collect data. We have thus developed a new method for analyzing Neotropical ant faunas that enables the taxonomic work linked to the identification of problematic species to be significantly reduced, while conserving most of the information on the ant assemblage. This method should enhance the work of Neotropical entomologists not specialised in taxonomy, particularly those concerned with biological conservation and indication. (C) 2010 Elsevier Ltd. All rights reserved.

Abstract: Drier periods from the late Pleistocene and early Holocene have been hypothesized to have caused the disappearance of various rainforest species over large geographical areas in South America and restricted the extant populations to mesic sites. Subsequent improvement in climatic conditions has been associated with recolonization. Changes in population size associated with these extinction-recolonization events should have affected genetic diversity within species. However, these historical hypotheses and their genetic consequences have rarely been tested in South America. Here, we examine the diversity of the chloroplast and nuclear genomes in a Neotropical rainforest tree species, Vouacapoua americana (Leguminosae, Caesalpinioideae) in French Guiana. The chloroplast diversity was analyzed using a polymerase chain reaction-restriction fragment length polymorphism method (six pairs of primers) in 29 populations distributed over most of French Guiana, and a subset of 17 populations was also analyzed at nine polymorphic microsatellite loci. To determine whether this species has experienced extinction-recolonization, we sampled populations in areas supposedly not or only slightly affected by climatic changes, where the populations would not have experienced frequent extinction, and in areas that appear to have been recently recolonized. In the putatively recolonized areas, we found patches of several thousands of hectares homogeneous for chloroplast variation that can be interpreted as the effect of recolonization processes from several geographical origins. In addition, we observed that, for both chloroplast and nuclear genomes, the populations in newly recolonized areas exhibited a significantly smaller allelic richness than others. Controlling for geographic distance, we also detected a significant correlation between chloroplast and nuclear population differentiation. This result indicates a cytonuclear disequilibrium that can be interpreted as a historical signal of a genetic divergence between fragmented populations. In conclusion, the spatial genetic structure of contemporary V. americana populations shows evidence that this species has experienced large extinction-recolonization events, which were possibly caused by past climatic change.

Abstract: We analysed 1040 individual trees, located in 62 plots across the Amazon Basin for leaf mass per unit area (M-A), foliar carbon isotopic composition (delta C-13) and leaf level concentrations of C, N, P, Ca, Mg, K and Al. All trees were identified to the species level with the dataset containing 58 families, 236 genera and 508 species, distributed across a wide range of soil types and precipitation regimes. Some foliar characteristics such as M-A, [C], [N] and [Mg] emerge as highly constrained by the taxonomic affiliation of tree species, but with others such as [P], [K], [Ca] and delta C-13 also strongly influenced by site growing conditions. By removing the environmental contribution to trait variation, we find that intrinsic values of most trait pairs coordinate, although different species ( characterised by different trait suites) are found at discrete locations along a common axis of coordination. Species that tend to occupy higher fertility soils are characterised by a lower M-A and have a higher intrinsic [N], [P], [K], [Mg] and delta C-13 than their lower fertility counterparts. Despite this consistency, different scaling patterns were observed between low and high fertility sites. Inter-relationships are thus substantially modified by growth environment. Analysing the environmental component of trait variation, we found soil fertility to be the most important predictor, influencing all leaf nutrient concentrations and delta C-13 and reducing M-A. Mean annual temperature was negatively associated with leaf level [N], [P] and [K] concentrations. Total annual precipitation positively influences M-A, [C] and delta C-13, but with a negative impact on [Mg]. These results provide a first basis for understanding the relationship between the physiological functioning and distribution of tree species across Amazonia.

Abstract: Abstract Wood properties and especially wood density have been used as functional traits organized along major axes of species life history and strategy. Beyond statistical analyses, a better mechanistic understanding of relationships among wood traits is essential for ecologically relevant interpretation of wood trait variations. A set of theoretical relationships mechanistically linking wood basic density with some other wood traits is derived from cellular material physics. These theoretical models picture basic physical constraints and thus provide null hypotheses for further ecological studies. Analysis is applied to data from two original datasets and several datasets extracted from the literature. Results emphasize the strong physical constraint behind the link between basic density and maximal storable water on the one hand, and elastic modulus on the other hand. Beyond these basic physical constraints, the developed framework reveals physically less expected trends: the amount of free water available for physiological needs increases in less dense wood of fast-growing species, and the cell wall stiffness decreases with density in temperate hardwoods and is higher in sapling stages in the rainforest understorey where competition for light is associated with high mechanical risk. We emphasize the use of theoretically independent traits derived from models of cellular material physics to investigate the functional variation of wood traits together with their environmental and phylogenetic variations. Although the current study is limited to basic density, green wood lumen saturation and wood specific modulus, we further emphasize the identification of complementary independent wood traits representing other biomechanical functions, nutrient storage, hydraulic conductance and resistance to drought. A plain language summary is available for this article.

Abstract: In this paper we give a priori error estimates for finite element approximations of linear parabolicproblems with pointwise control and incomplete data. We discretise the optimal control problemby using piecewise linear and continuous finite elements for the space discretisation of the state,and we use the backward Euler scheme for time discretisation. We prove a priori error estimates forthe state, the adjoint-state as well as for the low-regret pointwise optimal control.