Flow of terrestrial carbon though aquatic ecosystems (allochthony) is an important but underestimated component of the global carbon cycle. A lack of clear consensus about the importance of allochthonous (terrestrial) organic carbon is sometimes attributed to uncertainties associated with conventional ‘bulk’ isotope data, the most widely used ecological tracer. Amino acid‐specific isotope analysis is an emerging research method promising to address existing limitations of bulk C and N isotope analyses. We tested the efficacy of amino acid δ13C data as a generalizable measure of allochthony by analysing an aggregated dataset (n = 168) of primary and secondary data of carbon sources from disparate geographical locations across the globe. We found the δ13C fingerprints amino acids to be consistently distinct between allochthonous (terrestrial) and autochthonous (aquatic) carbon sources. We also found that our approach is most effective when we use only essential amino acid tracers (i.e. isoleucine, leucine, phenylalanine, threonine and valine). Predictive trends in δ13C fingerprints appear to be largely compatible across studies and/or laboratories. As a case study, we used this approach to quantify the contribution of terrestrial carbon to an endemic cavefish, Cryptotora thamicola, and found that its biomass was comprised largely of autochthonous carbon (~75%).

Compound‐specific isotope analysis of amino acids (CSIA‐AA) is a promising nascent technique that alleviates many shortcomings of conventional bulk‐tissue stable isotope analysis (“bulk SIA”) in ecological studies involving the tracing/reconstruction of carbon and nitrogen pathways. While CSIA‐AA has been increasingly applied to preserved tissue samples (e.g., material in natural history collections), the effects of sample preservation on amino acid δ13C and δ15N profiles are poorly understood. It is therefore unclear if mathematical correction factors are necessary for interpreting isotopic profiles of preserved samples. In this study, we investigated effects of ethanol preservation and formalin fixation on amino acid δ13C and δ15N profiles. We also assess how these effects translate to two ecological applications of CSIA‐AA: quantification of organic carbon sources, and estimation of trophic positions. Results from an 8‐week controlled experiment on freshwater fish tissue show negligible preservation effects on most amino acid δ15N profiles, and results are similar for δ13C profiles of essential amino acids. Findings from mixing models using essential amino acid δ13C profiles similarly show that preserved samples can yield robust estimates of carbon source contributions. We also empirically demonstrate, for the first time, the use of amino acid δ13C profiles to enhance δ15N‐based estimates of trophic position in food webs with multiple producers, and show that these estimates are not compromised by preservation effects. Overall, our findings support the view that amino acid δ13C and δ15N profiles from ethanol‐ and formalin‐treated CSIA‐AA samples can be directly used for addressing ecological questions.

No abstract

Land‐use change is a leading driver of biodiversity loss, especially in tropical fresh waters where the conversion of natural forest to monoculture plantations impacts freshwater fish assemblages. The environmental pathways underpinning shifts in fish assemblages, however, are poorly understood, but could potentially be inferred from trait–environment relationships. We addressed this knowledge gap using eco‐morphological traits to explain fish occurrences in oil palm‐impacted streams of the Endau drainage in Peninsular Malaysia. We also investigated how traits relate to differences in environmental conditions associated with land‐use change. We then integrate findings from the above to test how potential pathways of land‐use driven environmental changes can impact species occurrences through effects on life history, feeding habits, and mobility. Mixed‐effects models show that fishes with superior (upward‐facing) mouths and low body mass were more likely to occur in oil palm streams than forest streams, and these traits were associated with grass‐dominated riparian zones and reduced woody debris in oil palm streams, respectively. Structural equation models show that mouth positions statistically mediated the effect of riparian vegetation on fish species occurrences in oil palm streams. Specifically, fishes with superior mouths were more likely to feed on terrestrial invertebrates. Our analysis of easily measurable traits revealed pathways of land‐use impact that are potentially more widely applicable than conventional taxa‐based approaches. Fishes with superior mouths tended to occur in oil palm streams as they were able to more effectively exploit inputs of terrestrial invertebrates that are potentially associated with grass‐dominated riparian vegetation. Moreover, shifts in traits may suggest land‐use driven changes in stream ecosystem functioning (e.g. in terms of the role of terrestrial subsidies), thereby informing targeted management actions in land‐use impacted habitats (e.g. retention/restoration of riparian trees).

No abstract