Identifying the Molecular Signatures of Agricultural Expansion in Amazonian Headwater Streams

Abstract

Agricultural impacts on aquatic ecosystems are well studied; however, most research has focused on temperate regions, whereas the forefront of agricultural expansion is currently in the tropics. At the vanguard of this growth is the boundary between the Amazon and Cerrado biomes in Brazil, driven primarily by expansion of soybean and corn croplands. Here we examine the impacts of cropland expansion on receiving lowland Amazon Basin headwater streams in terms of dissolved organic carbon (DOC) concentration and dissolved organic matter (DOM) composition via ultrahigh‐resolution mass spectrometry. Streams draining croplands had lower DOC concentrations and DOM molecular signatures enriched in N‐ and S‐containing formula in comparison to forested streams. Cropland streams were also enriched in aliphatic, peptide‐like, and highly unsaturated and phenolic (low O/C) compound categories in comparison to forest streams (enriched in polyphenolics, condensed aromatics, and highly unsaturated and phenolic [high O/C] compound categories) indicative of the shifting of sources from organic‐rich surface soils and litter layers to autochthonous and more microbial biomass. Distinct molecular assemblages were strongly correlated with cropland and forest catchments, highlighting headwater streams as sentinels for detecting change. On investigation of unique molecular formulae present in only cropland sites, four cropland markers provided the ability to track agricultural impacts in the region. Overall, these patterns indicate reduced organic matter inputs in croplands and greater microbial degradation at these sites leading to declining DOC concentrations, and DOM of more microbial character in receiving streams that is more biolabile, with clear ramifications for downstream ecology and biogeochemical cycles.

Stormflows Drive Stream Carbon Concentration, Speciation, and Dissolved Organic Matter Composition in Coastal Temperate Rainforest Watersheds

“…The molecular properties of DOM in the stream draining the upland forest converged on those of the wetland sites during the storm (except for percent aliphatic and sugar‐like formulae). These findings are consistent with other temperate and tropical forested and agricultural watersheds showing elevated export of aromatic molecular formulae (e.g., condensed aromatics and polyphenolics, and as indicated by AI mod ) during high flow events (Eckard et al, 2017; Spencer et al, 2019; Wagner et al, 2019). The upland forest had the greatest CV for DOC and POC suggesting that organic forms of C vary more in uplands during storms than in the wetland streams, which in turn showed the greatest variation in DIC over the storm.…”

“…Prestorm, peak discharge, and poststorm samples (8/6/18, 8/9/18 a.m., and 8/11/18) were examined to assess unique molecular formulae present in only one of the three hydrographic sampling events and in at least two samples from that event to ensure that analysis focused on nonlocalized storm processes (e.g., Spencer et al, 2019; Table 1). Number of molecular formulae identified as unique diminished from prestorm (99) to peak (78) to poststorm (15).…”

“…Catchment landscape heterogeneity is one of the main mechanisms controlling stream C biogeochemistry because landcover impacts the molecular composition (Mosher et al, 2015; Seifert et al, 2016; Spencer et al, 2019) as well as the magnitude and partitioning of C exported across the terrestrial‐aquatic interface (Ågren et al, 2014; Goni et al, 2013; Perdrial et al, 2014). In our study, stream water DOC:DIC generally decreased along the hydric soil gradient (poor fen‐forested wetland‐upland forest) during peak flows resulting in DOC comprising a decreasing fraction of lateral C concentrations from the more highly drained upland forest soils during the storm.…”

Decreasing organic carbon bioreactivity in European rivers

“…As expected, in this study the absolute amounts of reactive organic matter (measured as BOD 5 ) in the merged data increased with bulk TOC, which is a proxy variable for colour in natural (Laudon, Köhler, & Buffam, 2004) The reactivity could to a certain extent be explained by the quality of the organic matter, as the low-TOC rivers with the highest k had the lowest organic C:N ratios, and vice versa. Organic matter with a low C:N ratio is highly biodegradable (Fellman et al, 2008;Islam et al, 2019), owing to its low contents of recalcitrant aromatic carbon rings (Hood, Williams, & McKnight, 2005), but high content of aliphatic and peptide-like structures (Kellerman et al, 2018;Spencer et al, 2019). High N contents of organic matter is common in agriculture-influenced rivers (Spencer et al, 2019) and typically originates from labile benthic algal and phytoplankton sources, whereas low N contents (high C:N) is linked to detritus from terrestrial sources (Balakrishna & Probst, 2005;Kaiser, Arscott, Tockner, & Sulzberger, 2004).…”

“…Organic matter with a low C:N ratio is highly biodegradable (Fellman et al, 2008; Islam et al, 2019), owing to its low contents of recalcitrant aromatic carbon rings (Hood, Williams, & McKnight, 2005), but high content of aliphatic and peptide‐like structures (Kellerman et al, 2018; Spencer et al, 2019). High N contents of organic matter is common in agriculture‐influenced rivers (Spencer et al, 2019) and typically originates from labile benthic algal and phytoplankton sources, whereas low N contents (high C:N) is linked to detritus from terrestrial sources (Balakrishna & Probst, 2005; Kaiser, Arscott, Tockner, & Sulzberger, 2004). Most terrestrial plant detritus that enters rivers has C:N > 20, but variability in this ratio is large and some terrestrial organic matter sources such as fens, marshes, or sewage have characteristically lower C:N ratios (Fellman et al, 2008; Lehman, Mayr, Liu, & Tang, 2015; Vaquer‐Sunyer et al, 2016).…”