Food web C and N stable isotope variability in high mountain lakes throughout an altitudinal gradient

Background/Question/Methods

High mountain lakes are increasingly used as sentinels of present Global Change, particularly for climate change, long-range pollution transfer and interactions between them. They are also suitable for reconstructing environmental fluctuations of the past using their sediment record. Understanding C and N flow through their food webs becomes necessary to interpret observations on past, and present changes. C an N stable isotopes are extremely useful to investigate biochemical pathways in food webs. Here, we present the results from four lakes in the Pyrenees (Spain), encompassing an elevation gradient from 1,620 to 2,688 m above sea level (a.s.l.). Lakes were shallow, softwater, and oligotrophic, with long ice cover periods (from ca. 4 to 7 months) and cold water temperatures during ice-free periods. The two lowest lakes were located below tree line and had watersheds with well developed soils and vegetation. In contrast, the two highest lakes were located above tree line and presented watersheds with a large proportion of bare rock and thin and poor soils. The lakes were sampled from phytoplankton to fish and from littoral to deep benthic zones in a single visit during summer 2004.

Results/Conclusions

No major differences were found in food web taxonomic composition, excluding the occurrence of aquatic macrophytes at lower lakes. However, a large variability was found in δ¹³C and δ¹⁵N within food web primary carbon sources (seston, biofilms, macrophytes, organic sediment), which propagated throughout the whole food web. An expected gradation from benthos to plankton was found for δ¹³C, reflecting differences in boundary-layer diffusivity. More interesting was the finding that bottom organisms from lakes below tree line were ¹⁵N depleted in comparison to littoral comparable organisms. These C and N differences propagated up to brown trout individuals, indicating that they specialize either in littoral or bottom preys.

Altitudinally, δ¹⁵N increased in all compartments. A reduction in the discrimination against ¹⁵N during primary production when nitrogen demand increases relative to nitrogen supply could be the reason for the δ¹⁵N observed tendency. Brown trout δ¹³C and δ¹⁵N were positively correlated within a lake, and the slope of the regression decreased lake altitude, reflecting the lower differences in δ¹⁵N between littoral and bottom. The existence of within-lake (littoral/bottom) and altitudinal patterns in δ¹⁵N without major changes in food web structure, suggest that δ¹⁵N signatures are suitable for monitoring Global Change effects in mountain systems.