Changes in feeding niche widths of perch and roach following biomanipulation, revealed by stable isotope analysis

1. We made an empirical test of a recent proposal that feeding niche widths might be determined as variance of stable isotope values. We determined delta C-13 and delta N-15 values of perch (Perca fluviatilis), roach (Rutilus rutilus) and their prey from a biomanipulated lake, when the mass removal of fish led to reduced inter- and intra-specific competition and increases in zooplankton abundance and body size. 2. After the first fish removals, both perch and roach mean delta C-13 values decreased and mean delta N-15 values increased, indicating a greater diet contribution from pelagic sources. 3. Variances of both delta C-13 and delta N-15 values first increased in both fish populations, indicating a wider food spectrum and expanded feeding niche width following reduced fish abundances. Observed changes were greater for the perch population than for roach. 4. In 2006, the perch population abruptly changed its diet so that most individuals were primarily consuming the abundant young-of-the-year fish, and this was reflected in significantly reduced variances of both delta C-13 and delta N-15 values. 5. We conclude that isotopic variance can indeed reflect changes in feeding niche width and offers a promising way to study such general ecological concepts.