Changes in rainfall level and litter stoichiometry affect aquatic community and ecosystem processes in bromeliad phytotelmata

• Published in: Freshwater biology Vol. 64; no. 8; pp. 1357 - 1368
• Main Authors: Benavides‐Gordillo, Sandra, Farjalla, Vinicius F., González, Angélica L., Romero, Gustavo Q.
• Format: Journal Article
• Language: English
• Published: Oxford Wiley Subscription Services, Inc 01.08.2019
• Subjects: Abundance; Additives; Algae; Aquatic communities; Aquatic ecosystems; Aquatic habitats; Aquatic invertebrates; Aquatic organisms; Atmospheric precipitations; Bacteria; bromeliad food web; Colour; Communities; Community structure; Cycles; Decomposition; Detritivores; Detritus; Ecological function; Ecosystem structure; Ecosystems; Freshwater ecosystems; Habitat availability; Inland water environment; Keystone species; Leaching; Leaf litter; Levels; Litter; Macroinvertebrates; Microcosms; Microorganisms; Mineral nutrients; N:P ratio; Naididae; Nutrient cycles; Organic matter; Particulate emissions; Particulate matter; Precipitation; Quality; Rain; Rainfall; Rainfall levels; Stoichiometry; Structure-function relationships; Suspended particulate matter; Tanks; Water tanks; Zoobenthos
• ISSN: 0046-5070; 1365-2427
• DOI: 10.1111/fwb.13310

Both N deposition and changes in precipitation amount are important components of global change and affect aquatic ecosystem functioning by altering the detrital quality and detrital processing rates by macroinvertebrates and microorganisms. Changes in precipitation pattern can also modify the physical structure of the ecosystem, determining habitat availability for aquatic organisms. Although these drivers can individually affect ecosystem structure and functioning, their interactive effects are poorly understood. To better understand the combined effects of detritus quality and precipitation regime on the structure and function of aquatic ecosystems we manipulated litter quality, by modifying N:P stoichiometry and rainfall amount on tropical natural aquatic microcosms (water‐holding epiphytic bromeliads). We performed an orthogonal manipulation of litter N:P ratio (natural N:P and high N:P) and rainfall levels (current level and predicted 40% increase for south‐eastern Brazil) Changes in litter quality affected the abundance of the aquatic organisms inhabiting bromeliads, including algae and bacteria. However, changes in litter quality combined with changes in precipitation levels affected only one group of macroinvertebrates, the Naididae worms (Oligochaeta) an important group of detritivores in the bromeliad systems. These combined drivers also affected the decomposition of organic matter by changing coloured organic matter concentrations, fine particulate matter biomass, and NH4+ concentrations in bromeliad tanks. The NH4+ concentration in bromeliad water was positively related with Naididae abundance, suggesting that these worms can act as ecosystem engineers by boosting N cycling. In both N:P treatments we observed a 50% decrease in N flux from litter to bromeliad leaves due to increases in nutrient leaching from the tanks caused by rainfall. The combined effects of litter quality and changes in precipitation regimes altered the decomposition process and nutrient cycling in tank bromeliads, probably through changes in the abundance of some keystone species (e.g. Naididae). Aquatic macroinvertebrate community structure remained similar after manipulating precipitation, despite changes in the abundance of some species. Although some microorganisms were washed out of the system, their abundance inside the bromeliads remained stable. We conclude that the abundance of individuals belonging to aquatic communities in bromeliad phytotelmata are generally stable and resistant to changes in rainfall levels, but are sensitive to changes in litter quality. This experiment showed that the interactive effects of increased precipitation and N concentrations differed from the effects of either factor alone. As responses to changes in these factors were complex rather than additive, caution is required in predicting the outcomes of changes in both factors in freshwater ecosystems.