Functional traits are commonly used in predictive models that link environmental drivers and community structure to ecosystem functioning. A prerequisite is to identify robust sets of continuous axes of trait variation, and to understand the ecological and evolutionary constraints that result in the functional trait space occupied by interacting species. Despite their diversity and role in ecosystem functioning, little is known of the constraints on the functional trait space of invertebrate biotas of entire biogeographic regions. We examined the ecological strategies and constraints underlying the realized trait space of aquatic invertebrates, using data on 12 functional traits of 852 taxa collected in tank bromeliads from Mexico to Argentina. Principal Component Analysis was used to reduce trait dimensionality to significant axes of trait variation, and the proportion of potential trait space that is actually occupied by all taxa was compared to null model expectations. Permutational Analyses of Variance were used to test whether trait combinations were clade‐dependent. The major axes of trait variation represented life‐history strategies optimizing resource use and antipredator adaptations. There was evidence for trophic, habitat, defence and life‐history niche axes. Bromeliad invertebrates only occupied 16%–23% of the potential space within these dimensions, due to greater concentrations than predicted under uniform or normal distributions. Thus, despite high taxonomic diversity, invertebrates only utilized a small number of successful ecological strategies. Empty areas in trait space represented gaps between major phyla that arose from biological innovations, and trait combinations that are unviable in the bromeliad ecosystem. Only a few phylogenetically distant genera were neighbouring in trait space. Trait combinations aggregated taxa by family and then by order, suggesting that niche conservatism was a widespread mechanism in the diversification of ecological strategies. A plain language summary is available for this article.

Little is known of how linkage patterns between species change along environmental gradients. The small, spatially discrete food webs inhabiting tank-bromeliads provide an excellent opportunity to analyse patterns of community diversity and food-web topology (connectance, linkage density, nestedness) in relation to key environmental variables (habitat size, detrital resource, incident radiation) and predators:prey ratios. We sampled 365 bromeliads in a wide range of understorey environments in French Guiana and used gut contents of invertebrates to draw the corresponding 365 connectance webs. At the bromeliad scale, habitat size (water volume) determined the number of species that constitute food-web nodes, the proportion of predators, and food-web topology. The number of species as well as the proportion of predators within bromeliads declined from open to forested habitats, where the volume of water collected by bromeliads was generally lower because of rainfall interception by the canopy. A core group of microorganisms and generalist detritivores remained relatively constant across environments. This suggests that (i) a highly-connected core ensures food-web stability and key ecosystem functions across environments, and (ii) larger deviations in food-web structures can be expected following disturbance if detritivores share traits that determine responses to environmental changes. While linkage density and nestedness were lower in bromeliads in the forest than in open areas, experiments are needed to confirm a trend for lower food-web stability in the understorey of primary forests.

Summary: The duration of the dry seasons in south‐eastern Amazonia is expected to increase. Little is known of how freshwater assemblages respond to drought in the humid rainforests and of the extent to which they resist the absence of rainfall before the collapse of the system. We manipulated rainshelters over tank‐forming bromeliads (i.e. the interlocking leaf axils of these plants form wells that collect rainwater) to simulate an exceptionally long dry period (49 days, compared with a 10‐year mean ± SD annual maximum number of 17 ± 5.3 days without rainfall at the study site) and then a rewetting period. By sampling weekly over 3 months, we followed the dynamics of the representation of abundance‐weighted traits in invertebrate assemblages in these treatment plants and in a control group. The functional structure of assemblages was drought resistant until the water volume in the bromeliad pools dropped by 90%, when there was a sudden shift in the functional trait structure due to the loss of most populations except the drought‐resistant culicids. Traits related to life history, body size and preferred food showed significant responses to drought. There was a convergence in the functional traits of species surviving in dry plants, strengthening the idea that environmental filtering, rather than stochasticity, determines the functional trajectory of aquatic assemblages during drought. At the end of the dry period, samples of the detritus potentially containing drought‐resistant eggs/cysts (and eventually live larvae) were taken from the dry plants and rewetted in the laboratory, allowing us to distinguish resistant species from those requiring recolonisation via subsequent oviposition by adults from elsewhere. Patches of water‐filled bromeliads persisting in the area provided the most important pool of colonists, and communities returned to the pre‐disturbance state within 1–2 weeks of rewetting. Our results suggest that the functional trait structure of invertebrate assemblages in bromeliads could remain stable under scenarios of precipitation change that would triple the duration of current dry periods at a local scale. Future experiments should evaluate how environmental factors might alter the tipping point between resistance to drought and a collapse in ecosystem processes.

Summary: Tank bromeliads form a conspicuous, yet neglected freshwater habitat in Neotropical forests. Recent studies driven by interests in medical entomology, fundamental aspects of bromeliad ecology and experimental research on food webs have, however, prompted increasing interest in bromeliad aquatic ecosystems. As yet, there is nothing in the literature about the life histories and environmental drivers of invertebrate population dynamics in tank bromeliads. Based on fortnightly samples taken over one year, size frequency plots and individual dry masses allowed us to establish the life cycles and growth rates of the dominant aquatic invertebrates in a common bromeliad species of French Guiana. Linear mixed‐effect models and Mantel tests were used to predict changes in density, biomass, and growth rates in relation to temperature, rainfall, humidity and detrital resources. Annual variations in invertebrate densities and biomasses could be described according to three types of distribution: unimodal, bimodal or almost constant. Despite seasonal variations, precipitation, temperature, relative humidity and detritus concentration accounted significantly for changes in density and biomass, but we found no significant responses in growth rates of most invertebrate species. Species rather displayed non‐seasonal life cycles with overlapping cohorts throughout the year. There was also a trend for delayed abundance peaks among congeneric species sharing similar functional traits, suggesting temporal partitioning of available resources. Beyond novel knowledge, quantitative information on life histories is important to predict food‐web dynamics under the influence of external forcing and self‐organisation. Our results suggest that changes in species distribution that will affect population dynamics through biotic interactions in space and/or time could have greater effects on food webs and ecosystem functioning than changes in environmental factors per se.

There is growing recognition that ecosystems may be more impacted by infrequent extreme climatic events than by changes in mean climatic conditions. This has led to calls for experiments that explore the sensitivity of ecosystems over broad ranges of climatic parameter space. However, because such response surface experiments have so far been limited in geographic and biological scope, it is not clear if differences between studies reflect geographic location or the ecosystem component considered. In this study, we manipulated rainfall entering tank bromeliads in seven sites across the Neotropics, and characterized the response of the aquatic ecosystem in terms of invertebrate functional composition, biological stocks (total invertebrate biomass, bacterial density) and ecosystem fluxes (decomposition, carbon, nitrogen). Of these response types, invertebrate functional composition was the most sensitive, even though, in some sites, the species pool had a high proportion of drought‐tolerant families. Total invertebrate biomass was universally insensitive to rainfall change because of statistical averaging of divergent responses between functional groups. The response of invertebrate functional composition to rain differed between geographical locations because (1) the effect of rainfall on bromeliad hydrology differed between sites, and invertebrates directly experience hydrology not rainfall and (2) the taxonomic composition of some functional groups differed between sites, and families differed in their response to bromeliad hydrology. These findings suggest that it will be difficult to establish thresholds of “safe ecosystem functioning” when ecosystem components differ in their sensitivity to climatic variables, and such thresholds may not be broadly applicable over geographic space. In particular, ecological forecast horizons for climate change may be spatially restricted in systems where habitat properties mediate climatic impacts, and those, like the tropics, with high spatial turnover in species composition.