Abstract: High-throughput sequencing of the 16S rRNA gene has considerably helped revealing the essential role of bacteria living on insect cuticles in the ecophysiology and behaviour of their hosts. However, our understanding of host-cuticular microbiota feedbacks remains hampered by the difficulties of working with low bacterial DNA quantities as with individual insect cuticle samples, which are more prone to molecular biases and contaminations. Herein, we conducted a methodological benchmark on the cuticular bacterial loads retrieved from two Neotropical ant species of different body size and ecology: Atta cephalotes (~15 mm) and Pseudomyrmex penetrator (~5 mm). We evaluated the richness and composition of the cuticular microbiota, as well as the amount of biases and contamination produced by four DNA extraction protocols. We also addressed how bacterial community characteristics would be affected by the number of individuals or individual body size used for DNA extraction. Most extraction methods yielded similar results in terms of bacterial diversity and composition for A. cephalotes (~15 mm). In contrast, greater amounts of artefactual sequences and contaminations, as well as noticeable differences in bacterial community characteristics were observed between extraction methods for P. penetrator (~5 mm). We also found that large (~15 mm) and small (~5 mm) A. cephalotes individuals harbour different bacterial communities. Our benchmark suggests that cuticular microbiota of single individual insects can be reliably retrieved provided that blank controls, appropriate data cleaning, and individual body size and functional role within insect society are considered in the experiment.

Abstract: A revision of Resinicium collections (Basidiomycota, Hymenochaetales) from French Guiana and French West Indies is provided, and a new species, Resinicium grandisporum sp. nov., supported by morphological as well as phylogenetic analyses based on ITS DNA sequences, is described and illustrated. An updated key of the genus Resinicium is also provided, which includes species previously described from outside of the studied area.

Abstract: Context: Many aquatic communities are linked by the aerial dispersal of multiple, interacting species and are thus structured by processes occurring in both the aquatic and terrestrial compartments of the ecosystem. Objectives: To evaluate the environmental factors shaping the aquatic macroinvertebrate communities associated with tank bromeliads in an urban landscape. Methods: Thirty-two bromeliads were georeferenced to assess the spatial distribution of the aquatic meta-habitat in one city. The relative influence of the aquatic and terrestrial habitats on the structure of macroinvertebrate communities was analyzed at four spatial scales (radius = 10, 30, 50, and 70 m) using redundancy analyses. Results: We sorted 18,352 aquatic macroinvertebrates into 29 taxa. Water volume and the amount of organic matter explained a significant part of the taxa variance, regardless of spatial scale. The remaining variance was explained by the meta-habitat size (i.e., the water volume for all of the bromeliads within a given surface area), the distance to the nearest building at small scales, and the surface area of buildings plus ground cover at larger scales. At small scales, the meta-habitat size influenced the two most frequent mosquito species in opposite ways, suggesting spatial competition and coexistence. Greater vegetation cover favored the presence of a top predator. Conclusions: The size of the meta-habitat and urban landscape characteristics influence the structure of aquatic communities in tank bromeliads, including mosquito larval abundance. Modifications to this landscape will affect both the terrestrial and aquatic compartments of the urban ecosystem, offering prospects for mosquito management during urban planning. © 2017, Springer Science+Business Media B.V.

Abstract: Decomposition experiments that control leaf litter species across environments help to disentangle the roles of litter traits and consumer diversity, but once we account for leaf litter effects, they tell us little about the variance in decomposition explained by shifts in environmental conditions versus food-web structure. We evaluated how habitat, food-web structure, leaf litter species, and the interactions between these factors affect litter mass loss in a neotropical ecosystem. We used water-filled bromeliads to conduct a reciprocal transplant experiment of two litter species between an open and a forested habitat in French Guiana, and coarse- and fine-mesh enclosures embedded within bromeliads to exclude invertebrates or allow them to colonize leaf litter disks. Soft Melastomataceae leaves decomposed faster in their home habitat, whereas tough Eperua leaves decomposed equally in both habitats. Bacterial densities did not differ significantly between the two habitats. Significant shifts in the identity and biomass of invertebrate detritivores across habitats did not generate differences in leaf litter decomposition, which was essentially microbial. Despite the obvious effects of habitats on food-web structure, ecosystem processes are not necessarily affected. Our results pose the question of when does environmental determinism matter for ecosystem functions, and when does it not. © 2017, Springer International Publishing Switzerland.

Abstract: The Neotropical understory plant Tachia guianensis (Gentianaceae)- known to shelter the colonies of several ant species in its hollow trunks and branches-does not provide them with food rewards (e.g., extrafloral nectar). We tested whether these ants are opportunistic nesters or whether mutualistic relationships exist as for myrmecophytes or plants sheltering ant colonies in specialized hollow structures in exchange for protection from enemies and/or nutrient provisioning (myrmecotrophy). We noted 37 ant species sheltering inside T. guianensis internodes, three of them accounting for 43.5% of the cases. They protect their host plants from leaf-cutting ant defoliation and termite damage because individuals devoid of associated ants suffered significantly more attacks. Using the stable isotope 15N, we experimentally showed that the tested ant species furnish their host plants with nutrients. Therefore, a mutualism exists. However, because it is associated with numerous ant species, T. guianensis can be considered a nonspecialized myrmecophyte. © 2017 by The University of Chicago.