Spatiotemporal patterns of dissolved free amino acids in New England rivers could be unique and stable odor signatures for imprinting and homing by Atlantic salmon

Abstract

The phenomenon of homing by salmonid fishes to natal sites for breeding is well established, but the chemicals in rivers that they learn as juveniles and identify as adults are not fully determined. Dissolved free amino acid (DFAA) profiles have been hypothesized to allow salmonids to distinguish their natal river from others nearby. To evaluate this hypothesis, we sampled DFAAs in spring and autumn (when juveniles learn and adults return, respectively) from 3 rivers in New England, USA, that support the landlocked Atlantic Salmon, Salmo salar Linnaeus, 1758. We used 3 approaches to determine the level of consistency between seasons and difference between rivers that would be needed for DFAA to support imprinting and subsequent homing for reproduction: hierarchical cluster analysis, statistical difference tests, and equivalence tests. DFAAs were not detected in the water column of the study rivers, but sediment porewater samples yielded DFAAs at measurable concentrations. Hierarchical cluster analysis, difference testing, and equivalence testing all indicated that some combinations of sediment porewater DFAA concentrations differed among rivers and were similar between spring and autumn within a river. Specifically, equivalence tests revealed subsets of sediment porewater DFAAs that were seasonally equivalent within each river, and none of the seasonally equivalent DFAAs were common among all 3 rivers (i.e., each river had a unique DFAA profile). However, exceptions detected in the cluster analysis and equivalence testing raise questions regarding the extent to which DFAAs might be sufficient for salmon imprinting and homing. Thus, DFAAs may fulfill some of the hypothesized prerequisites as salmon imprinting and homing odor cues, but our lack of understanding of salmon discriminatory abilities and limited DFAA data preclude definitive conclusions about the sufficiency of DFAAs alone as homing cues.