%0 Journal Article
%T Developing δ15N and δ13C isoscapes using whole blood from Magellanic penguins, Spheniscus magellanicus.
%A Gonzalez JF
%A Sánchez-Carnero N
%A Frere E
%A Yorio P
%A Ciancio JE
%J Rapid Commun Mass Spectrom
%V 38
%N 18
%D 2024 Sep 30
%M 38989637
%F 2.586
%R 10.1002/rcm.9860
%X <B>BACKGROUND: </B>Understanding the migration of marine animals is hindered by the limitations of traditional tracking methods. It is therefore crucial to develop alternative methods. Stable isotope-based tracking has proven useful for this task, although it requires detailed isoscapes in the focal area. Here, we present predator-based isoscapes of the coastal zone of the Patagonian Shelf Large Marine Ecosystem (PSLME), which offers a novel tool for geolocation.<BR><B>METHODS: </B>Whole-blood samples from breeding Magellanic penguins nesting at 11 colonies were used to create δ15N and δ13C isoscapes. Isotopic values were assigned to random positions inside their corresponding foraging area. Spatial analysis and data interpolation resulted in δ15N and δ13C isoscapes for the coastal zone of the PSLME, which were validated through cross-validation.<BR><B>RESULTS: </B>The isoscapes mean standard error ranged from 0.05 to 0.41 for δ15N and from 0.07 to 0.3 for δ13C, similar to the error range of the mass spectrometer used for measuring isotope ratios. Predictive surfaces reflected the latitudinal trends, with δ13C and δ15N values increasing northwards. δ13C values showed a strong latitudinal gradient, while δ15N values had two distinct domains, with higher values in the north. The error surface indicated the highest certainty within 130 km from the shore and within the reported Magellanic penguin foraging areas.<BR><B>CONCLUSIONS: </B>Both isoscapes revealed strong spatial variation. The δ13C isoscape showed a latitudinal gradient, consistent with patterns in other oceans. The δ15N isoscape clearly separated northern and southern colonies, likely influenced by nitrogen sources. The error obtained fell within the measurement error ranges, adding credibility to the models.