• Gutowsky, Lee F. G.
  • Aslam, Waleed
  • Banisaeed, Reza
  • Bell, Lauren R.
  • Bove, Kenzie L.
  • Brownscombe, Jacob W.
  • Burrows, Graham J. J.
  • Chu, Elaine
  • Magel, Jennifer M. T.
  • Rous, Andrew M.
  • Cooke, Steven J.


To generate mortality estimates for fish that are captured and released in recreational and commercial fisheries, it is common to temporarily hold fish in captivity. Typically, captured fish are placed in some form of pen, cage or tank with control individuals, yet little is known about how the type of holding environment influences fish condition or mortality. Here we captured freshwater fish (bluegill; Lepomis macrochirus) via angling and fyke net and retained them in one of four holding environments; a round flow-through tank on shore [TANK], a knotless nylon pen with natural substrate in the lake [PEN], a knotless nylon floating cage with a rigid structure [RCAGE], and a knotless nylon floating cage that lacked rigid structure [CAGE]. Mortality was low (1%) across both capture techniques and holding environments during the 14-day retention period. All mortalities were associated with capture by fyke net. A chronic stress indicator, blood glucose, was determined for a subset of fish on day 5. Although there were significant differences in blood glucose between angled RCAGE and angled PEN (Tukey, P = 0.047) and angled RCAGE and fyke PEN (Tukey, P = 0.015), the observed levels were generally quite low (range: 1.0–3.9 mmol L−1) and the differences were likely associated with differences in feeding; fish in the PEN group with access to substrate (and presumably the most food) had slightly higher glucose levels. At the conclusion of the study Fulton's condition factor was similar among all groups (ANOVA, P > 0.05, all terms). However, fish held in the TANK treatment had the highest levels of external protozoan parasite infection by Ichthyopthirius (Tukey, P < 0.05). This study illustrates that in situ holding environments (rather than tanks) may help reduce mortality, stress, and disease during studies that estimate post-release mortality. We encourage additional research to explore how the holding environment can influence inferences made about post-release mortality and sublethal impacts of fishing.


Study location

The experiment was conducted at Queen's University Biological Station on Opinicon Lake, Ontario (44° 34′ N, 76° 19′ W). Opinicon Lake is a shallow mesotrophic lake located along the Rideau Canal waterway. The lake contains abundant populations of warm-water fish species such as largemouth bass (Micropterus salmoides) and bluegill sunfish. Experimental procedures were carried out between 4 May 2014 and 18 May 2014 when water temperature averaged 14 °C (range: 12–15 °C).

Holding environments

Three replicates of four different holding environment treatments were used in this experiment: floating rigid cages [RCAGE], floating cages [CAGE], pens that reached the substrate [PEN], and tanks [TANK]. Replicates for the RCAGE treatment were assembled using a 1.22 m × 1.22 m × 1.22 m frame of 2.54 cm diameter white CPVC pipe. A knotless nylon net (material obtained from Memphis Net and Twine, Heavy Delta, 12.7 mm sq.) with one unmeshed side was placed over the structure and affixed in place with cable ties. The CAGE treatment was similar except that it had only CPVC pipe around the perimeter of the unmeshed side, i.e., the remaining mesh was free to move in the water. The open side of both the RCAGE and CAGE treatments were kept afloat by pool noodles that were fixed to this part of the structures. Alternating between RCAGE and CAGE, replicates of the pens (n = 6) were attached at their corners using a 61 cm length of twine. The cages were taken to approximately 2.44 m of water and anchored at either end so that the entire structure remained oriented in the same direction during the experiment, i.e., parallel to shore. To construct a PEN replicate, four 2.44 m lengths of rebar were hammered into the substrate approximately 1.22 m apart in 1.07–1.22 m of water. A knotless nylon net (same material as above) with two open sides was placed around the rebar and connected to it using cable ties. The surface side of the mesh was tied to the rebar approximately 1.83–2.44 m from the surface of the water. To reduce the chance of fish escaping near the substrate, loose mesh was covered with sand, gravel, bricks, and rocks. The TANK treatments included a row of three 1000 L outdoor circular fiberglass tanks supplied with flow-through lake water at a rate of 166 L h−1, where water was exchanged 4 times per day (Fig. 1).

Capture of fish

Bluegill were captured 4 May 2014 from Opinicon Lake, Ontario. Fykes nets (n = 3) each had 7 steel hoops that were 0.5 m apart and 0.9 m in diameter. Nets had two wings and a lead that was attached vertically to the mouth of each net. Wings were 4.6 m long by 0.9 m high, leads were 10.7 m long by 0.9 m high, and the mesh 2.54 cm square nylon (See Stoot et al., 2013 for more detail). Fyke nets were set in shallow weedy bays and checked twice during the day to capture bluegill (n = 204). Angling techniques used standard spinning gear with small hooks, bobbers, and a small piece of worm to capture bluegill (n = 204) from shallow weedy bays that were similar to those used during fyke netting. Angling occurred during the day when the fyke nets were fishing. To distinguish fish by a capture technique, individuals were marked by clipping either the upper or lower corner of the caudal fin. Treatment replicates were randomly populated with a similar size range of fish from each capture technique (17 fish per capture technique/replicate or 34 fish total/replicate; Table 1).