New epilithic δ15N-based analytical protocol for classifying Nitrogen impact in Lake Bracciano. (October 2020)

• Record Type: Journal Article
• Authors: Fiorentino, Federico; Jona Lasinio, Giovanna; Careddu, Giulio; Sporta Caputi, Simona; Rossi, Loreto; Calizza, Edoardo; Costantini, Maria Letizia
• Is Part Of: Ecological indicators. Volume 117(2020)
• Publication Date: 2020-10
• Subjects: N impact classification -- Stable N isotope analysis -- Lake Bracciano -- Epilithic periphyton -- Bayesian modelling; Environmental monitoring -- Periodicals; Environmental management -- Periodicals; Environmental impact analysis -- Periodicals; Environmental risk assessment -- Periodicals; Sustainable development -- Periodicals; 333.71405
• DOI: 10.1016/j.ecolind.2020.106663
• Languages: English
• ISSNs: 1470-160X

Abstract

Highlights: A robust analytical protocol for the epilithic δ 15 N analysis is proposed. Bayesian model identifies probability distributions for Nitrogen input classes. Changes of Nitrogen inputs occur during drought at different space and time scales. Two areas of Lake Bracciano are affected by persistent anthropic Nitrogen inputs. Abstract: Nitrogen inputs in aquatic ecosystems are increasing and climate change is likely to exacerbate cultural eutrophication. The recovery of aquatic ecosystem functionality requires strenuous efforts and entails considerable costs. Therefore, the development of early warning ecological indicators that can help arrest the phenomenon in its early stages is highly desirable. Stable isotope analysis of Nitrogen in algal primary producers has proved useful in determining the origins of Nitrogen inputs in several marine and freshwater ecosystems. Nitrogen signatures are often assigned to impact or non-impact classes by comparing the Nitrogen signature of samples with the Nitrogen signature ranges of potential sources, which can hinder objective ecological evaluation when sample signatures are close to the upper/lower boundaries of source ranges. To overcome this problem, we obtained the Nitrogen signatures of the epilithic associations collected in the littoral zone of Lake Bracciano (Central Italy), covering a pre-drought (2015–2016) and ongoing drought (2017–2019) period. The Bayesian Gaussian Mixture Model determined four probability distributions, each associated with a Nitrogen impact class, and assigned the observed epilithic signatures to the most appropriate classes. Application of the approach at various spatial and temporal scales allowed us to compare the pre-drought and ongoing drought Nitrogen input dynamics. At each spatial and temporal scale, we observed differences in the input dynamics arising from the side effects of the drought on human activities, which were reflected in changes in the probability of Nitrogen signatures belonging to one or the other impact class. Based on this probability, the proposed analytical protocol provided a useful tool for prioritizing specific management measures in areas affected by specific Nitrogen inputs. Moreover, with a few recalibrations, the model proposed for Lake Bracciano can be extended to other contexts.