Comparing trophic position estimates using bulk and compound specific stable isotope analyses: applying new approaches to mackerel icefish Champsocephalus gunnari.

Quantifying the tropic position (TP) of an animal species is key to understanding its ecosystem function. While both bulk and compound-specific analyses of stable isotopes are widely used for this purpose, few studies have assessed the consistency between and within such approaches. Champsocephalus gunnari is a specialist teleost that predates almost exclusively on Antarctic krill Euphausia superba. This well-known and nearly constant trophic relationship makes C. gunnari particularly suitable for assessing consistency between TP methods under field conditions. In the present work, we produced and compared TP estimates for C. gunnari and its main prey using a standard bulk and two amino acid-specific stable isotope approaches (CSI-AA). One based on the difference between glutamate and phenylalanine (TPGlx-Phe), and the other on the proline-phenylalanine difference (TPPro-Phe). To do that, samples from C. gunnari, E. superba and four other pelagic invertebrate and fish species, all potential prey for C.gunnari, were collected off the South Orkney Islands between January and March 2019, analyzed using standard isotopic ratio mass spectrometry methods and interpreted following a Bayesian approach. Median estimates (CI95%) for C. gunnari were similar between TPbulk (3.6; CI95%: 3.0-4.8) and TPGlx-Phe(3.4; CI95%:3.2-3.6), and lower for TPPro-Phe (3.1; CI95%:3.0-3.3). TP differences between C. gunnari and E. superba were 1.4, 1.1 and 1.2, all compatible with expectations from the monospecific diet of this predator (ΔTP=1). While these results suggest greater accuracy for Glx-Phe and Pro-Phe, differences observed between both CSI-AA approaches suggests these methods may require further validation before becoming a standard tool for trophic ecology.