Files in this Data Supplement:

• Supplemental Figure 1 -

  Fig. S1. Schematic of the experimental toadfish shelter used in the gray snapper behavior study. The shelter was constructed from 10 cm inner diameter polyvinylchloride (PVC) pipe with one opening capped. A submersible IR camera was mounted on the PVC end-cap and the lens was directed toward the shelter’s open end. The video output cable was connected to a VHS video recorder at a remote location. A clay toadfish model was positioned in their stereotypical posture guarding the shelter entrance. Polyethylene tubing was connected to the odorant injection and water sampling ports.

• Supplemental Figure 2 -

  Fig. S2. Photographs of clay toadfish model and live gulf toadfish (aerial exposure). (A) Toadfish model is set inside an experimental shelter (Fig. S1) and positioned in a stereotypic posture with flared pectoral fins and open mouth, which is associated with guarding a nesting site. The model also possessed glass eyes to create added realism and these clay models were repeatedly struck by gray snappers (Tables 1−4). (B). Live gulf toadfish inside the experimental shelter for comparison. (C) The clay model (center) flanked by live toadfish of comparable size.

• Supplemental Figure 3 -

  Fig. S3. Photograph of the rhodamine dye plume mapping experiment showing the propagation of the red dye plume. Water samplers were positioned inside the experimental shelter (point-source) and at the 50 cm perimeter of the cleared behavioral arena.

• Supplemental Figure 4 -

  Fig. S4. Behavioral responses to 2.5 μmol N l⁻¹ amino acid treatments and seawater controls during 60 min observation period. Amino acids selected were l-proline (P), l-alanine (A), and l-glycine (G), with a 1:1 AG mix and a 1:1:1 PAG mix. Baseline is a control for pump noise during a 60 min period prior to odorant delivery. The behavioral index is the sum of the number of snapper responses per 60 min, where snapper entry into the shelter, passing 0−25 cm in front of shelter, passing 25−50 cm in front of shelter, hovering near a shelter and visually scanning the contents are awarded 5, 2, 1 and 1 points, respectively. Data were log(x+1) transformed for equal variance. Significant differences between each treatment and its shelter control were determined by a paired t-test and denoted by (*). Significant differences among treatment groups where determined with a repeated measures ANOVA with the Holm−Sidak post-hoc test and are indicated with different lowercase letters (F=2.300, P=0.104, N=6).

• Movie 1 -

  Movie 1. Video images of gray snapper recorded from a submersible camera mounted at the rear of an experimental shelter. The odorant point source was below the clay toadfish model in the foreground. Black and white striped PVC pipes were in a cross-hair arrangement; the center post was 50 cm from the shelter opening whereas the two outer posts were 25 cm away. The gray snapper was presumably tracking an odorant plume to its source by chemoreception. After repeated approaches, the snapper entered the experimental shelter and eventually struck at a clay toadfish model.