Fish Diversity, Biomass and Trophic Dynamics in Relation to Lanscape Gradients in the Floodplain of the Amazon River
Year: 2017
Type: dissertation
Abstract: Floodplains are among the most biologically productive, but also threatened ecosystems in the world with global loss of natural systems exceeding 60%. Human development activities, including the construction of dams, mining and conversion of forests to croplands and pastures, are rapidly changing floodplains landscapes, likely impacting fish diversity, aquatic food webs, and fishery production. My dissertation investigates the possible effects of land-cover changes on fish diversity and biomass in the lower Amazon River floodplain, and evaluates how food webs vary across different scales in this changing landscape. Field expeditions were conducted during four phases of the annual hydrological cycle in several habitats across a gradient of landscape conditions, from highly deforested to mostly forested areas, along the Amazon River floodplain. The study surveyed fish assemblages and obtained data on local environmental variables and fisheries management practices as well as samples of fish muscle tissue and basal production sources for analysis of stable isotope ratios. In addition, landscape features were mapped based on satellite imagery.\n\nSpatial patterns of fish biomass distribution and biodiversity were strongly associated with forest cover as well as local environmental conditions linked to landscape gradients. Several species and functional groups defined by life history, feeding, swimming/microhabitat-use strategies were positively associated with forest cover. Other species, including some that would usually be considered habitat generalists and species directly dependent on autochthonous resources (e.g., planktivores), were most common in areas dominated by herbaceous vegetation or open-water habitats associated with the opposite extreme of the forest cover gradient. Beta-diversity and the degree of uniqueness of species combinations within habitats were also positively associated with forest cover. Basal production sources and fishes showed considerable variation in C and N stable isotope ratio values across the landscape. Analysis of stable isotope data using Bayesian mixing models indicated that aquatic macrophytes and suspended particulate organic material (POM) were the principal sources contributing to the biomass of most fish species at the regional scale; however, contributions of different production sources to fish biomass varied within a given local hydrological unity area. These results suggest that spatial variation in the isotopic ratios and proportional contributions of basal sources to fish biomass are probably influenced by multiple factors, including variation in the quality and quantity of basal sources associated with differences in watershed vegetation across the landscape. In addition, these results demonstrate that because estimated contributions of production sources to fish biomass varied according to the spatial scale of analysis, study designs should carefully consider not only replication and temporal scale, but also spatial scale. Finally, results showing that spatial patterns of fish diversity and biomass are associated with forest cover, indicate that forest loss in the Amazonian floodplain results in spatial homogenization of fish assemblages and reduced functional diversity and fishery production. Maintaining fish diversity and fisheries production in floodplain ecosystems of the Amazon River requires protecting mosaics of both aquatic habitats and floodplain vegetation, with sufficient forest cover being critically important.
Author: Arantes, Caroline Chaves
Language: English
Subfield: Nature and Landscape Conservation
Field: Environmental Science
Domain: Physical Sciences
SDG: Life on land
Open Access status: green
Awards: THE FLOODPLAIN (VARZEA) OF THE MAIN STEM AMAZON RIVER EXTENDS OVER NEARLY 100,000 KM2, CONSTITUTING THE WORLD'S LARGEST RIVERINE WETLAND. THE CO, THE ANNUAL FLOOD OF THE AMAZON RIVER IS THE WORLD S LARGEST INUNDATION EVENT, FLOODING ABOUT 300,000 KM^2 FOR SIX OR MORE MONTHS EACH YEAR, WITH WATE, G584417851, G7080527179
OpenAlex ID: W7028380688
