Thermal selection and delayed migration by adult sockeye salmon (Oncorhynchus nerka) following escape from simulated in-river fisheries capture

Abstract

Two hypotheses were tested concerning the consequences to adult sockeye salmon (Oncorhynchus nerka) of escape from commonly used fishing gear (gillnet, seine net and tangle net). First, by experimentally exposing 214 fish to three commonly used fishing gear types (gillnets, tangle nets, or seine nets) and releasing to complete migration after PIT-tagging, we tested the hypothesis that migration success and behaviour are differentially impacted according to the type of fishing gear interaction and escape. Also, by fitting salmon with thermal loggers, we tested whether salmon behaviourally selected cooler lake water following an escape encounter. Migration success was unaffected after escape from the simulated gear types compared to control fish. Surviving sockeye salmon that had gillnet or seine net encounter, but not a tangle net encounter, on average took approximately 2 days longer to migrate to the spawning area (averaging 14.7 days and 14.9 days, respectively) compared to control fish (12.6 days). Furthermore, escaped fish migrated at cooler temperatures through a lake system (average of 14.0 °C) compared to control fish (15.3 °C), which would reduce their absolute cost of transport during this migration by reducing the standard metabolic rate by about 10%. Consequently, in addition to demonstrating that fish escapement from a fishing gear will alter their subsequent migration behaviour, we introduce the possibility that the associated increase in energy expenditure and migration delay can be partially compensated for by behavioural selection of cooler water, if it is available, to lower basic energy turnover in a fish that is entirely reliant on energy stores to fuel its spawning migration.

Introduction

Pacific salmon (Oncorhynchus spp.) are important culturally (Lichatowich, 2013), ecologically (Cederholm et al., 1999), and economically (Gislason et al., 2017a, Gislason et al., 2017b) and are targeted by numerous marine and freshwater fisheries. In riverine areas, the most common harvesting of salmon targets adult fish during their spawning migration using either gillnets, tangle nets, beach seine nets, or rod-and-reel angling (Patterson et al., 2017a). As a result, incidental encounters and by-catch are common, with many fish being released after capture, while others escape from fishing gear (Raby et al., 2015, Bass et al., 2018a, Bass et al., 2018b, Kanigan et al., 2019), which we collectively defined here as ‘non-retention’ fish. The frequency and fate of non-retention fish post-net encounter is of importance for fisheries managers who need to quantify total fishing-related mortality for specific fisheries and estimate predictions of spawner abundance to manage salmon populations (Ricker, 1976, Patterson et al., 2017a, Patterson et al., 2017b).

Most research to date has focused on the consequences of release from common fishing gears (reviewed in: Raby et al., 2015; Patterson et al., 2017a) and estimates of short-term post-release mortality are included in some fishing impact models used to predict spawner abundance (DFO, 2019). In contrast, the incidence and consequences of escape are poorly understood (Patterson et al., 2017a), and most fisheries models quantifying fishing-related incidental mortality for salmon do not explicitly account for escapees (Baker et al., 2014, Patterson et al., 2017b). The importance of escapees is perhaps best understood for the gillnet fishery because survivors of encounters carry characteristic entrapment wounds. In Bristol Bay, Alaska, for example, 6–44% of sockeye salmon (Oncorhynchus nerka) were observed on spawning areas with scars consistent of a gillnet entanglement, and of fish with these injuries, approximately half failed to reproduce (Baker and Schindler, 2009). Similar scarring was seen on 10 to > 40% of certain populations of Fraser River sockeye salmon in British Columbia (BC), Canada (Clarke et al., 1994), on their spawning grounds between 1987 and 1994. Between 2014 and 2016, 19–27% of sockeye salmon captured on the Seton River (a tributary of the Fraser River) had gillnet scars and compared with uninjured fish, their en-route and pre-spawn mortality increased by 16–18% (Bass et al., 2018a). Moreover, a higher gillnet fishing effort in the lower Fraser River was positively correlated with a higher proportion of gillnet injuries to sockeye salmon captured in Seton River (Kanigan et al., 2019). This high prevalence of gillnet wounds could, in fact, suggest a current underestimation of the percentage of salmon escaping gillnets if fish die post-entanglement and pre-observation. Clearly, further experimental study is needed on salmon escapements from common fishing gears to more accurately estimate the frequency and consequences of gear escape at different locations.

Different fishing gear types can cause different types of injuries, levels of exhaustion, and levels of stress for fish that escape or are released, which then may trigger different consequences for survival and spawning (Davis, 2002, Cooke et al., 2013; Broadhurst et al., 2006). Fraser River sockeye salmon, for example, can encounter gillnets, tangle nets, or beach seines during their spawning migrations. Gillnet entanglement can cause exhaustion, suffocation, lacerations (sometimes extensive and severe), and mucus removal that provides an entry point for infectious agents (Kojima et al., 2004, Baker and Schindler, 2009). Post-release survival for Fraser River sockeye salmon following manual release after an experimental gillnet entanglement can differ among populations (17.8% for Harrison and 34.2% for Weaver populations; Donaldson et al., 2012), between sexes (30% survival for females and 66% survival for males experiencing 20 min gillnet entanglement; Teffer et al., 2017), and with maturation state (35% for less mature and 75% for more mature; Bass et al., 2018b). The smaller mesh size and larger hang-ratio of tangle nets compared with gillnets is intended to entangle a fish’s fins, mouth, or teeth rather than ensnaring and constricting their body. Nevertheless, while an entanglement with a tangle net can still exhaust, lacerate, and cause mucus loss, survival has been shown to be higher than with gillnet encirclement (Vander Haegen et al., 2004, Donaldson et al., 2012) with an estimated 68.6% survival of adult Columbia River Chinook salmon (Oncorhynchus tshawytscha) released from a tangle net (Ashbrook et al., 2008). Beach seines potentially harm fish by removing mucus and possibly through suffocation if oxygen is depleted by a high fish density in the net sets (Raby et al., 2014). However, injures following beach seine escape are typically mild, and post-release survival for Pacific salmon has been shown to be higher for than gillnets (Bass et al., 2018b). Survival to spawning areas after a beach seine encounter is reported from 33% to 82% for adult sockeye salmon (Donaldson et al., 2011, Robinson et al., 2015, Bass et al., 2018b).

Despite this knowledge, no study has directly compared the effects of the three fishing gear types commonly used to intercept adult sockeye salmon during their river migration (gillnet, tangle net, and seine net) within a single fishery. Therefore, we aimed to investigate the consequences of simulated escapes from these three gear types. This study focused on a subset of a population of Fraser River sockeye salmon that had likely not encountered previous net entanglement. A volitional escape from entrapment in fishing gear prior to landing differs from by-catch release because escapees do not experience air exposure or direct human handling (Patterson et al., 2017a). Also, because the escape is most often unobserved, the exact process of escape is largely unknown. Increased energy expenditure and stress for the fish are expected as they struggle to free themselves, typically using burst swimming to push their way through, or out of, mesh entanglement. Beach seine escaped fish would likely swim under the lead line or jump over the float line. We tested the hypothesis that migration success and behaviour are differentially impacted according to the type of fishing gear interaction by PIT-tagging fish before they were released back to the river for short (200 m to negotiate a dam raceway) and a longer (50 km to their spawning area) migrations. Consistent with previous studies (Mäkinen et al., 2000, Frank et al., 2009, Nguyen et al., 2014, Bass et al., 2019), we predicted that escapees would have reduced migration success and an increase in migration duration for fish that escape gear compared with control fish, and that interaction with gillnets would have a greater impact compared with tangle net and seine net interactions.

Also, by fitting the fish with a thermal logger, we could test the hypothesis that escapees would behaviourally select cooler water while migrating through deep lakes enroute to their spawning area compared with fish that did not experience a gear escape simulation. Behavioural thermoregulation of Pacific salmon is well documented (Newell and Quinn, 2005; Mathes et al., 2009; Keefer and Caudill, 2015), with fish potentially seeking out cooler water in response to supra-optimal water temperature (Goniea et al., 2006), and for energetic and reproductive benefits (Roscoe et al., 2010). Of course, these behaviours, which can increase survival to spawning grounds (Mathes et al., 2009), may become increasingly important in an era of warming water temperatures. Moreover, recent evidence suggests that gillnet-injured adult sockeye salmon seek cool water during their up-river migration (Bass, 2018). Escaping fishing gears requires increased energy expenditure and migration delay (Mäkinen et al., 2000, Nguyen et al., 2014, Bass et al., 2019) and often leaves fish with lacerations (which may serve as an entry point for opportunistic infectious agents). Behavioural thermoregulation following fishing gear interactions may therefore offer benefits for a migrating salmon using a fixed energy store as basic energy needs (the standard metabolic rate) would be reduced, and it would slow the proliferation of opportunistic pathogen infections.