Untangling the complex food webs of tropical rainforest streams.

Food webs depict the tangled web of trophic interactions associated with the functioning of an ecosystem. Understanding the mechanisms providing stability to these food webs is therefore vital for conservation efforts and the management of natural systems. Here, we first characterised a tropical stream meta-food web and five individual food webs using a Bayesian Hierarchical approach unifying three sources of information (gut content analysis, literature compilation and stable isotope data). With data on population-level biomass and individually measured body mass, we applied a bioenergetic model and assessed food web stability using a Lotka-Volterra system of equations. We then assessed the resilience of the system to individual species extinctions using simulations and investigated the network patterns associated with systems with higher stability. The model resulted in a stable meta-food web with 307 links among the 61 components. At the regional scale, 70% of the total energy flow occurred through a set of 10 taxa with large variation in body masses. The remaining 30% of total energy flow relied on 48 different taxa, supporting a significant dependency on a diverse community. The meta-food web was stable against individual species extinctions, with a higher resilience in food webs harbouring omnivorous fish species able to connect multiple food web compartments via weak, non-specialised interactions. Moreover, these fish species contributed largely to the spatial variation among individual food webs, suggesting that these species could operate as mobile predators connecting different streams and stabilising variability at the regional scale. Our results outline two key mechanisms of food web stability operating in tropical streams: (i) the diversity of species and body masses buffering against random and size-dependent disturbances and (ii) high regional diversity and weak omnivorous interactions of predators buffering against local stochastic variation in species composition. These mechanisms rely on high local and regional biodiversity in tropical streams, which is known to be strongly affected by human impacts. Therefore, an urgent challenge is to understand how the ongoing systematic loss of diversity jeopardises the stability of stream food webs in human-impacted landscapes.

As teias alimentares representam um emaranhado de interações tróficas associadas ao funcionamento de um ecossistema. Compreender os mecanismos que proporcionam estabilidade a estas teias alimentares é, portanto, vital para os esforços de conservação e gestão dos sistemas naturais. Aqui, primeiro caracterizamos uma meta teia alimentar de riachos tropicais e cinco teias alimentares individuais usando uma abordagem hierárquica Bayesiana unificando três fontes de informação (análise de conteúdo estomacal, compilação de literatura, dados de isótopos estáveis). Com dados sobre biomassa em nível populacional e massa corporal medida individualmente, aplicamos um modelo bioenergético e avaliamos a estabilidade da cadeia alimentar usando um sistema de equações Lotka­Volterra. Em seguida, avaliamos a resiliência do sistema às extinções de espécies individuais usando simulações e investigamos os padrões de rede associados a sistemas com maior estabilidade. O modelo resultou em uma meta teia alimentar estável com 307 ligações entre os 61 componentes. Na escala regional, 70% do fluxo total de energia ocorreu através de um conjunto de dez taxa com grande variação nas massas corporais. Os restantes 30% do fluxo total de energia dependiam de 47 taxa diferentes, apoiando uma dependência significativa de uma comunidade diversificada. A meta teia alimentar foi estável contra extinções de espécies individuais, com uma maior resiliência em teias alimentares que abrigam espécies de peixes onívoros capazes de conectar múltiplos compartimentos da teia alimentar através de interações fracas e não especializadas. Além disso, estas espécies de peixes contribuíram amplamente para a variação espacial entre as cadeias alimentares individuais, sugerindo que estas espécies poderiam operar como predadores móveis conectando diferentes riachos e estabilizando a variabilidade à escala regional. Nossos resultados descrevem dois mecanismos principais de estabilidade da cadeia alimentar operando em riachos tropicais: (i) a diversidade de espécies e massas corporais que protegem contra distúrbios aleatórios e dependentes do tamanho (ii) alta diversidade regional e fracas interações onívoras de predadores que protegem contra a variação estocástica local na composição de espécies. Estes mecanismos dependem de uma elevada biodiversidade local e regional em riachos tropicais, que são conhecidos por serem fortemente afetados pelos impactos humanos. Portanto, um desafio urgente é compreender como a contínua perda sistemática de diversidade põe em risco a estabilidade das teias alimentares em paisagens impactadas pelo homem.

Do microplastics mediate the effects of chemicals on aquatic organisms?

Microplastics are ubiquitous in both marine and freshwater ecosystems, where they can act as a physical contaminant, as well as interact with chemicals present in the environment. It has been suggested that chemical contaminants can sorb to microplastics, such that microplastics act as a vector for chemicals into aquatic biota and enhance their negative effects. It has been repeatedly suggested that the main factors underpinning the binding of chemicals to microplastics are hydrophobic partitioning and the size of microplastic particles. Therefore, we used the hydrophobicity of chemicals, as log Kow, as well as the size of microplastic particles to conduct a quantitative analysis of published results to evaluate the influence of microplastics on chemical toxicity. We collated data from 39 laboratory studies that assessed the effects of microplastics, chemicals and their combination on several ecotoxicological responses of freshwater and marine organisms. Each chemical was assigned the relevant octanol / water partition coefficient (log KOW) as a measure of its hydrophobicity, and the mean size of microplastics particles used in each study was recorded. We found no effect of log KOW or the size of microplastic particles on the interaction between microplastics and chemicals with regards to any of the relevant ecotoxicological responses (behaviour, growth, survival and cellular) considered in this study. These findings are significant in showing that the effect of microplastics on the toxicity of chemicals is more complex than just considering hydrophobicity of chemicals and size of microplastics. We call for more mechanistic experiments to motivate a robust risk assessment and mitigation of microplastic toxicity in the environment.

Climate change impacts on contaminant loads delivered with sediment yields from different land use types in a Carpathian basin.

Soil runoff and sediment transport are considered as an important vector for particle-bound contaminant transfer from source to receiving waters. Under changing climate conditions and rapid basin development, identification of sediment origins is critical for planning further action to reduce erosion effects, and further pollution to surface waters. The goal of this study was to distinguish sediment sources in a Carpathian basin (Wolnica River, southern Poland) and to perform source-oriented contaminant load estimations. Sediment yields (SYLD) and land use specific sediment yields (LUSY) were modeled with the use of the Macromodel DNS/SWAT (Discharge-Nutrients-Sea/Soil and Water Assessment Tool). Sorting of sediment sources was performed by the fingerprinting method using variability of the geochemical composition of soils (Pb, Zn, Cd, Cu, Mn, Ni, Fe, Hg, total N and P, Σ16 PAHs, and 137Cs) of four land use (LU) types: arable lands (A), grasslands (G), residential areas (R), and forests (F). Statistical analysis revealed six metals (Pb, Zn, Cd, Cu, Ni, and Hg) as fingerprint properties providing the best source discrimination in this basin. The contribution of particular land use origin assessed with the use of the mixing model varied in the range of 20-30%. Finally, estimation of land use specific contaminant loads in suspended sediments was performed as a result of a modeling and sediment fingerprinting combination. The final estimates revealed yearly LUSY values varying between 716 t/y for A, 12 t/y for F, and metal loads from 31 kg/y for Zn to values below 100 g/y for Cd and Hg. Long-term predictions (2046-2055) of the metal loads revealed an increase by 75% under the combined RCP 8.5 climate change and land use scenarios. These findings are of great value for land management in the Carpathian basins, especially with regards to the predicted increase of forest cover which significantly alters contaminant signals conveyed through the system.