Using stable isotopes to investigate movement of fish in rice paddy fields

We evaluated movement of fish, especially Misgurnus spp. (loach), in paddy fields and irrigation ponds by conducting an inventory of δ13C and δ15N of fish, potential food sources, and soil organic matter in two irrigation pond–paddy field systems in Korea. The pond–paddy systems differ with respect to the presence or the absence of a ridge between the paddy field and the irrigation pond and also whether or not livestock are present in their watersheds. The ridge prevents the free movement of fish between paddy field and irrigation pond habitats in one of the pond–paddy watersheds, but not in the other watershed. We found differences in loach δ13C and δ15N inhabiting the paddy fields compared to those in loach δ13C and δ15N in the irrigation ponds. In irrigation ponds, loach δ13C was lower in September (average −27.9 and −27.7 ‰) compared to July (average −26.2 and −26.3 ‰) in the watershed with a ridge (station 1) and without a ridge (station 6), respectively. Loach δ13C in irrigation ponds in September was similar to loach δ13C in the paddy field in July at both sampling sites, indicating loach might have moved into irrigation pond from paddy field. Loach δ15N in the watershed with livestock was significantly higher (average 18.2 ‰) in the irrigation ponds than loach δ15N in the watershed with no livestock present (average 11.3 ‰), probably reflecting higher anthropogenic nitrogen inputs from livestock. Differences in δ13C and δ15N in the loach reflected changes in habitat utilization of loach between paddy fields and irrigation ponds. Aquatic insect δ13C differed from loach δ13C but were more similar to carp δ13C. The stable isotope inventory approach used in this study could be used to augment or replace a more traditional field based mark–recapture approach.