Gadus morhua, Linnaeus, 1758, Linnaeus, 1758

Green, John M., Schornagel, Dustin, Nguyen, Khanh Q., Pennell, Curtis & Morris, Corey J., 2024, Field observations of the movements of locally adapted Atlantic cod (Gadus morhua) living in zero and sub-zero centigrade temperatures for half the year, Canadian Journal of Zoology 102 (3), pp. 253-263 : 255-259

These 49 fish

had a mean total length of 54.5 cm (range 43– 75 cm), and were likely mature and of migratory size ( Morris and Green 2002; Morris et al. 2014). Three fish with less than 50 positions were excluded from analysis. Some tagged fish were detected outside The Shinneys typically from June until September resulting in fewer VPS positional data during this time ( Fig. 3 ). Of fifteen fish that moved out of The Shinneys in May and June as water temperatures increased above 0 ◦ C, nine returned before the water temperature became sub-zero ( Fig. 3 ). Commercial and/or recreational fishers likely caught some or all of those that did not return. There was no evidence of mortality among tagged fish at the overwintering site during summer or winter based on continued detection and changes in positions of tags.

Fish tracked in this study inhabited sub-zero temperatures continuously from December to June each year ( Fig. 4A ). Comparison between warm and cold (sub-zero) periods showed that the cumulative movement of tagged fish in The Shinneys was greater during the cold periods than the warm periods (ANOVA, F = 1421, DF = 4, p <0.001) ( Fig. 4B ). When the water temperature at 10 m depth was greater than 0 ◦ C (June–November), the cumulative distances that tagged fish moved within the tracking array were less than those during the cold period (December–May). Monthly comparison, taking into consideration lower numbers of fish providing positional data in June and July also measured lower cumulative distances during the September to November period (>0 ◦ C) than during the colder months (December–May) (ANOVA, F = 2.92, DF = 24, p = 0.001) ( Fig. 4B ). Pairwise comparison of cumulative distances moved between seasons is shown in Fig. 5A for fish in the tracking array.

Rate of movement

The proportion of time Gilbert Bay cod spent in each rate of movement category was similar during both the warm and cold periods (p> 0.05 for all comparisons; Fig. 6 ). For both cold and warm conditions, fish spent most time in the stopped category (62.8%) or moving slowly (26.2%), and far less time moving at a medium (7.6%) or fast (3.4%) rate (ANOVA, F = 254.8, DF = 3, p <0.001) ( Fig. 6 ). Pairwise post hoc comparisons for the four rate of movement categories are shown in Fig. 5B . During monthly periods from September until November (>0 ◦ C) and from December until May (<0 ◦ C), fish were mostly inactive, moving on average 0.1 m /s (±0.06SD) ( Fig. 6 ). However, there was considerable individual variability in movement rates, punctuated by short periods of relatively rapid movement by some fish throughout the cold period ( Fig. 7 ). The average rate of movement increased from March to May while water temperature was still sub-zero ( Fig. 4C ), and was significantly more variable when water temperature was greater than 0 ◦ C (ANOVA, F = 187.4, DF = 3, p <0.001) ( Fig. 5 ).

Depth usage inside the array

Gilbert Bay cod selected shallower depths during cold periods (mean of 8.3 m ± 6.2 SD) compared to warm periods (mean of 10.2 m ± 6.4 SD) (ANOVA, F = 9719, DF = 24, p <0.001) ( Fig. 4D ). The range in depths used during the coldest months (January–March) was also smaller than that in June and July although the mean depth of 10 m was similar. Tracked fish within the VPS area used comparatively deeper and more variable depths during July to September (mean of 11.4 m ± 5.7 SD), followed by slightly shallower depths from October to December (mean of 9.2 m ± 7 SD) and April to June (mean of 8.5 m ± 5.7 SD) (ANOVA, F = 3956, DF = 24, p <0.001) ( Fig. 4D ). The greater variability in depth use, including some of the deepest sites available within the VPS area, prior to the onset of sub-zero temperatures in November and December 2013, contrasts with use of shallower depths, with relatively little variability, during the coldest months from January to March. No tagged fish overwintered in the deepest parts (i.e.,> 25 to 80 m) of The Shinneys located within 1–3 km of the VPS array. While these sites were outside the VPS array, other receivers monitored them.

Figures

Fig. 3. Tagged Atlantic cod (Gadus morhua) detections over time. Colours represent occupied zones, corresponding to areas identified in Fig. 1, and indicate movement between The Shinneys VEMCO Positioning System array (orange), Bar Tickle West (red), George’s Cove (green), Magashu Tickle (purple), and River Out (blue). Gray blocks represent periods when Atlantic cod (Gadus morhua) are out of The Shinneys.

Fig. 4. Average temperature-use (A), cumulative distance (B), movement speed (C), and depth-use (D) by tracked Atlantic cod (Gadus morhua) in Gilbert Bay from September 2013 to September 2015. The number of tagged individuals contributing data during each month of tracking is indicated along the top of each panel. The boxes represent the distribution of 50% of the data considered, the red or blue squares represent the mean, and the horizontal black line represents the median data value, while the vertical “whiskers” represent the range of the data. Round black circles are outliers. Blue and red squares in the top panel indicate when Atlantic cod (Gadus morhua) utilized cold (≤0 ◦C) and warm (>0 ◦C) water temperatures, respectively. Red squares of other panels indicate mean values of each variable.

Fig. 5. Pairwise post hoc comparisons of Atlantic cod (Gadus morhua) cumulative movement distances between seasons (A) and different movement categories (B), using Tukey’s “honestly significant differences” test, showing mean values that are significantly different from each other. Vertical bars are 95% confidence intervals (CIs). Open points indicate the predicted mean values of tested variables. The horizontal red dashed line at 0 is the reference line, and overlapping CIs indicate no statistically significant difference.

Fig. 6. Boxplots representing the percentage of time in each speed category of Atlantic cod (Gadus morhua) in the VEMCO Positioning System array. The boxes represent the distribution of 50% of the data considered, red square represents the mean, and horizontal black lines represent the median data value, while the vertical “whiskers” represent the range of the data. Round black circles are outliers. The numbers of tagged Atlantic cod (Gadus morhua) contributing data are indicated above each box. Rates of movement are based on consecutive positions only (no missed tag transmissions). Speed categories include stopped (0–0.1 m/s), slow (>0.1–0.2 m/s), medium (>0.2–0.3 m/s), and fast (>0.3 m/s).

Fig. 7. Examples of four Atlantic cod (Gadus morhua) exhibiting periods of fast rates of movement during winter 2014 and 2015. These movements lasted several minutes based on multiple consecutive position estimates. Distances moved (UTMs) while moving at medium (blue) and fast (red) rates of movement show examples of the spatial coverage for these individual Atlantic cod (Gadus morhua) during bursts of movement.