WESPEN Online Order Form print this page
US Fish & Wildlife Service FieldNotes

STOCKTON FWO: Evaluation of Natural Marks for Adult Delta Smelt

Region 8, September 26, 2013
Figure 1. Dorsal view of the head, including 3 AOI: pre-, inter- and post-orbital.
Figure 1. Dorsal view of the head, including 3 AOI: pre-, inter- and post-orbital. - Photo Credit: n/a
Figure 2. Measurements considered for morphometric analyses in adult delta smelt - 1: Post adipose fin length, 2: Pre-dorsal fin length, 3: Head length, 4: Orbital length (snout to center of eye), 5: Maximum body depth (location may vary among fish), 6: Adipose fin length, 7: Adipose fin height, 8: caudal peduncle depth (minimum depth of caudal peduncle), 9: Pre-anal fin length, 10: Post-anal fin length, 11: Standard length, 12: Fork length, 13: Total length.
Figure 2. Measurements considered for morphometric analyses in adult delta smelt - 1: Post adipose fin length, 2: Pre-dorsal fin length, 3: Head length, 4: Orbital length (snout to center of eye), 5: Maximum body depth (location may vary among fish), 6: Adipose fin length, 7: Adipose fin height, 8: caudal peduncle depth (minimum depth of caudal peduncle), 9: Pre-anal fin length, 10: Post-anal fin length, 11: Standard length, 12: Fork length, 13: Total length. - Photo Credit: n/a

By Gonzalo Castillo

External pigmentation patterns or scars, are potentially useful as natural marks to identify individual animals, eliminating the need for invasive marking or tagging. Following initial evaluation of natural marks for delta smelt (Hypomesus transpacificus) in 2012, staff from the Stockton Fish and Wildlife Office, the UC Davis Fish Conservation and Culture Lab (FCCL), Biopar, and Reclamation began a project to continue evaluating the use of natural marks in this species.

Researchers used cultured adult delta smelt produced at the FCCL to evaluate the use of external pigmentation (melanophores) as natural marks. Examination of adult delta smelt revealed several potential areas of interest (AOI) for natural marks (three dorsally in the head, one dorsally in the caudal peduncle, one ventrally in the mandible and one laterally in the head). The three dorsal AOI of the head were selected as primary AOI for subsequent analyses as they showed extensive and diverse pigmentation (Figure 1).

To independently evaluate the effectiveness of natural marks, researchers tagged 300 fish with an individual alphanumeric code (VIA tags, Northwest Marine Technologies). A digital camera equipped with a macro lens was used to acquire the selected AOI. A second digital camera and a floatarium were also used to acquire whole body images for obtaining 13 morphometric measurements (Figure 2). Initial evaluation of natural marks involved visual (naked eye) matching of digital images. A qualitative matching-grade criteria was used to measure the confidence for the manual matching between images assumed to correspond to the same fish (4: excellent; 3: good; 2: fair and 1: poor).

Visual matching of the dorsal head AOIs in tests 1 (sessions 1-2) and 2 (sessions 1-3) showed the following results: 1) highly diverse pigmentation patterns in each AOI, which enabled matching images based on the location and size of individual pigments and clusters of pigments, 2) higher reliance on the inter-orbital AOI for matching images, 3) a 100 percent correct visual matching for the images of 30 fish examined by two trained individual in blind tests, 4) average matching-grades per AOI from 2.3 to 3.3 with the highest average grade for the inter-orbital area, 5) a general decrease in average matching grade for the 3 AOI between sessions 1-2 and 1-3.

Further analyses are being conducted to evaluate changes in pigmentation over time and the potentially role of sex, spawning condition and light exposure on skin pigmentation. In addition, further evaluation and development of automated methods (algorithms) are planned to attempt matching the pigmentation patterns of individual delta smelt from digital images taken over three photo session.; hence, potentially enabling faster automated recognition of individual fish by means of natural marks.

Gonzalo Castillo is a fish biologist at the Stockton Fish and Wildlife Office in Lodi, California.

Contact Info: Gonzalo Castillo, 209-334-2968x323, gonzalo_castillo@fws.gov