Abstract: It was recently shown that Pheidole megacephala colonies (an invasive species originating from Africa) counterattack when raided by the army ant, Eciton burchellii. The subsequent contact permits Pheidole cuticular compounds (that constitute the “colony odour”) to be transferred onto the raiding Eciton, which are then not recognised by their colony-mates and killed. Using a simple method for transferring cuticular compounds, we tested if this phenomenon occurs for Neotropical ants. Eciton workers rubbed with ants from four sympatric species were released among their colony-mates. Individuals rubbed with Solenopsis saevissima or Camponotus blandus workers were attacked, but not those rubbed with Atta sexdens, Pheidole fallax or with colony-mates (control lot). So, the chemicals of certain sympatric ant species, but not others, trigger intra-colonial aggressiveness in Eciton. We conclude that prey-ant chemicals might have played a role in the evolution of army ant predatory behaviour, likely influencing prey specialization in certain cases.

Abstract: Around the world, the human-induced collapses of populations and species have triggered a sixth mass extinction crisis, with rare species often being the first to disappear. Although the role of species diversity in the maintenance of ecosystem processes has been widely investigated, the role of rare species remains controversial. A critical issue is whether common species insure against the loss of functions supported by rare species. This issue is even more critical in species-rich ecosystems where high functional redundancy among species is likely and where it is thus often assumed that ecosystem functioning is buffered against species loss. Here, using extensive datasets of species occurrences and functional traits from three highly diverse ecosystems (846 coral reef fishes, 2,979 alpine plants, and 662 tropical trees), we demonstrate that the most distinct combinations of traits are supported predominantly by rare species both in terms of local abundance and regional occupancy. Moreover, species that have low functional redundancy and are likely to support the most vulnerable functions, with no other species carrying similar combinations of traits, are rarer than expected by chance in all three ecosystems. For instance, 63% and 98% of fish species that are likely to support highly vulnerable functions in coral reef ecosystems are locally and regionally rare, respectively. For alpine plants, 32% and 89% of such species are locally and regionally rare, respectively. Remarkably, 47% of fish species and 55% of tropical tree species that are likely to support highly vulnerable functions have only one individual per sample on average. Our results emphasize the importance of rare species conservation, even in highly diverse ecosystems, which are thought to exhibit high functional redundancy. Rare species offer more than aesthetic, cultural, or taxonomic diversity value; they disproportionately increase the potential breadth of functions provided by ecosystems across spatial scales. As such, they are likely to insure against future uncertainty arising from climate change and the ever-increasing anthropogenic pressures on ecosystems. Our results call for a more detailed understanding of the role of rarity and functional vulnerability in ecosystem functioning. © 2013 Mouillot et al.

Abstract: Summary While around 20% of the Amazonian forest has been cleared for pastures and agriculture, one fourth of the remaining forest is dedicated to wood production [1]. Most of these production forests have been or will be selectively harvested for commercial timber, but recent studies show that even soon after logging, harvested stands retain much of their tree-biomass carbon and biodiversity [2,3]. Comparing species richness of various animal taxa among logged and unlogged forests across the tropics, Burivalova et al.[4] found that despite some variability among taxa, biodiversity loss was generally explained by logging intensity (the number of trees extracted). Here, we use a network of 79 permanent sample plots (376 ha total) located at 10 sites across the Amazon Basin [5] to assess the main drivers of time-to-recovery of post-logging tree carbon (Table S1). Recovery time is of direct relevance to policies governing management practices (i.e., allowable volumes cut and cutting cycle lengths), and indirectly to forest-based climate change mitigation interventions. © 2015 Elsevier Ltd.

Abstract: Tree growth on weathered soils in lowland tropical forests is limited by low phosphorous (P) availability. However, nutrient manupulation experiments do not always increase the P content in these trees, which raises the question whether trees are taking up added P. In French Guianese lowland rainforest, we measured changes in nitrogen (N) and P availability before and up to two months after N and P fertilizer addition, in soils with intact root systems and in soils where roots and mycorrhizal fungi were excluded by root exclusion cylinders. When the root system was excluded, P addition increased P availability to a much greater extent and for a longer time than in soils with an intact root system. Soil N dynamics were unaffected by root presence/absence. These results indicate rapid P uptake, but not N uptake, by tree roots, suggesting a very effective P acquisition process in these lowland rainforests.

Abstract: Ants are omnipresent in tropical forests, especially territorially dominant arboreal ants whose territories are spatially segregated forming ‘ant mosaics’. These ecologically important species are rarely used in conservation monitoring because of the difficulty in collecting them. We developed a standardised baitline protocol to study the distribution of dominant ants on canopy trees and also a procedure to objectively define species dominance, even in unknown ant assemblages.
Besides eliminating the need to climb trees, this protocol allows live arboreal ant specimens to be sampled at different heights. Behavioural aggressiveness assays between the collected workers provide data on the three-dimensional distribution of colonies and on interactions between species. We compared the results of the behavioural tests to those from null models.
In the New Guinean lowland forest studied, we show that the canopy was either shared by multiple territorial species or inhabited by a single species with a large territory. The baitline protocol collected up to half of the arboreal ant species found in a felling census. However, the proportion of species collected at baits decreased with the increasing spatial dominance of single territorial species.
Behavioural observations used in the protocol allowed a more efficient detection of ant mosaics than null models. Territorially dominant ants were active on both understorey and canopy trees.
The protocol is fast and easy to replicate. It is a potential tool for understanding and monitoring the spatiotemporal dynamics of arboreal ant assemblages and can detect populous colonies, including those of invasive species