Lumpfish, Cyclopterus lumpus, distribution in the Gulf of Maine, USA: observations from fisheries independent and dependent catch data

Introduction

Climate change is expected to change the distribution of many marine species, primarily with poleward shifts (Perry et al., 2005; Campana et al., 2020). The Gulf of Maine (GoM) is an interesting case study applicable to this issue as the area is one of the fastest warming bodies of water on the planet (Pershing et al., 2015). Yet, most work on species distribution shifts has focused on species of commercial interest given funding and data availability (Pershing et al., 2015; Free et al., 2019; Goode et al., 2019; Fredston-Hermann et al., 2020). Evaluating changes over time to unexploited species can be telling of ecosystem changes (e.g., marine historical ecology; Engelhard et al., 2016) yet difficult to do because these species are not usually the target of long-term fisheries surveys and are data-limited. Often these data-limited species appear as bycatch but only in small numbers. However, it is possible to combine multiple data sets using different gear types to ensure adequate sample sizes. For example, fisheries-independent trawl surveys, recreational fish surveys, and citizen science dive surveys were aggregated and modeled to predict temporal changes in rockfish distribution in Puget Sound, WA (Tolimieri et al., 2017). Coupling long term catch information with abiotic and biotic variables can provide insight into how these drivers affect ecological communities (e.g., Hampton et al., 2013; Rogers et al., 2019).

While well-studied in other areas within its range, lumpfish (Cyclopterus lumpus) is an example of a species with limited data in the GoM, and one that may be exploited in US waters in the near future for its use as a cleanerfish in northern New England salmonid farming operations (J Ford, Cooke Aquaculture, pers. comm., 2023). Currently, lumpfish is not a regulated species in US waters, therefore, no fishery management plan exists for them. However, in Canada, lumpfish are listed as Threatened due to steep population decreases most likely caused by fishing (DFO, 2021). As they become exploited in US waters, lumpfish will need to be managed and, to do so sustainably, resource managers need to be informed of baseline data on fish biomass, occurrence, and distribution.

Lumpfish are distributed in temperate waters (3−10 °C; Rodríguez-Rey & Whittaker, 2023) in both the eastern and western Atlantic Ocean at high latitudes. In the eastern Atlantic, lumpfish are found in the Barents Sea, around Iceland and Greenland, and as far south as Spain and Portugal (Stein, 1986). In the western Atlantic, lumpfish range from Hudson Bay and Labrador, Canada southwards along the US eastern seaboard to New Jersey (Stein, 1986). Lumpfish are characterized as semi-pelagic but inhabit both the pelagic and demersal zones and undertake diel vertical migrations depending on the season (Kennedy et al., 2016). In Icelandic waters, adult lumpfish typically are found at depths <300 m but have been caught as deep as 518 m (Kennedy et al., 2016). As adults, lumpfish tend to live offshore, preferring cooler (Iceland: 0−6 °C, Kennedy et al., 2016; Barents Sea: 4−7 °C, Eriksen, Durif & Prozorkevich, 2014) waters; however, during the spring spawning season, adults migrate inshore to the coastlines (Goulet, Green & Shears, 1986; Collette & Klein-MacPhee, 2002). Larval lumpfish typically are found in shallower coastal waters, but can occur offshore too (i.e., Barents Sea, Eriksen, Durif & Prozorkevich, 2014). Juvenile lumpfish prefer slightly warmer waters than adult fish (Norway: 6−7 °C, Mortensen et al., 2020; Barents Sea 5−7 °C, Eriksen, Durif & Prozorkevich, 2014). With increasing water temperatures, the probability of occurrence for lumpfish declines (Rodríguez-Rey & Whittaker, 2023).

In the USA, lumpfish have never been exploited, and as a result little is known about lumpfish populations in US waters (Collette & Klein-MacPhee, 2002). Published information is limited to Great Bay Estuary, NH (Rackovan & Howell, 2017) and Schoodic Peninsula, ME (e.g., Moring, 1989; Moring, 2001). Remaining population information is inferred from studies throughout the other areas of their range. Semi- pelagic adults move inshore to rocky coasts to spawn from March to May in the southwestern GoM, and May to June along the northeast Maine coast (Cox & Anderson, 1922; Goulet, Green & Shears, 1986; Collette & Klein-MacPhee, 2002). Females spawn two sticky, demersal egg masses and then move offshore, while the males stay to guard and tend the eggs until hatching, which occurs after approximately six to eight weeks, depending on water temperatures (Cox & Anderson, 1922; Collins, 1978; Goulet, Green & Shears, 1986; Martin-Robichaud, 1991; Kennedy, 2018). Juvenile lumpfish leave the nest area in the early summer and are highly associated with macroalgae, either in tidepools or in the upper 0.5 m of the water column (Daborn & Gregory, 1983; Moring, 1989; Rackovan & Howell, 2017), where they prey on small invertebrates including amphipods, copepods, isopods, and even small fish larvae (Moring, 1989; Tully & O’Ceidigh, 1989; Davenport & Rees, 1993). They depend on seaweed for transportation as it passively drifts, providing protection from predators and an increased food source (Vandendriessche et al., 2007). Lumpfish grow quickly in their first year of life compared to other cold-water marine fishes, reaching approximately 35 to 70 mm in total length (TL; Martin-Robichaud, 1991). During this time, most of their energy is diverted towards growth since the fish cling to algae and wait for prey to pass by (Brown, 1986; Killen, Brown & Gamperl, 2007). At age 1+, lumpfish become mostly pelagic and move into deeper waters (Blacker, 1983). Lumpfish can live up to approximately 10 to 15 years. In the wild, males reach sexual maturity in two to three years while it takes females three to four years (Albert et al., 2002; Hedeholm, ME & Grønkjær, 2014). While GoM lumpfish are considered part of a western Atlantic lumpfish stock unit, composed of US and Canadian fish (Whittaker, Consuegra & Garcia deLeaniz, 2018), they are distinct from the Canadian populations (Langille et al., 2023; Langille et al., 2024).

Despite the few GoM-based lumpfish studies, numerous long-term state and federal surveys that include lumpfish catch data exist which could be used to provide more information about occurrence of lumpfish in the GoM. Therefore, our study goals were to: (1) aggregate GoM lumpfish catch data and characterize lumpfish distribution, and (2) determine if and how water temperature may have affected lumpfish distribution over time. We hypothesized that lumpfish distribution would be correlated to water temperature both temporally and spatially, shifting northeast with increases in water temperature.

Materials & Methods

Results

Across the five datasets, we identified over 12,000 instances of lumpfish being caught across the five datasets, including 9,910 in the 1989–2021 NEFSC Observer data (Table 1). Most of the datasets only indicate positive catch records and not true catch-per-unit effort (although we explore the NEFSC Bottom Trawl data further below to address this point). There was enormous variability in the spatial and temporal scales of catch between the datasets (Fig. 1, Table 1). The NEFSC Observer and Bottom Trawl programs caught lumpfish over the largest range, which is in line with expectations given sampling protocols due to the other dataset’s fishing efforts focused nearshore (Fig. 1). NEFSC Observer data were the most consistent throughout the year. The NEFSC Bottom Trawl and most state surveys were conducted only during spring and fall. Most lumpfish were caught by various types of trawl gear or gill nets (Table 2).

Lumpfish were caught at bottom temperatures ranging from 2  °C to 17.6 °C and at depths ranging from 2 m to 393 m (Fig. 2). Most fish caught were adult individuals, however, NH F&G surveys only caught YOY fish (Figs. 3–4). NH F&G surveys occurred closer to shore and only in estuaries, which often act as nursery habitats for lumpfish (Rackovan & Howell, 2017). Further, NEFSC Observer surveys rarely caught juvenile and YOY individuals due to gear selectivity as they use standard legal fishing gear, which have larger mesh sizes than the fishery-independent surveys (Fig. 4A). Adult lumpfish were caught throughout the GoM whereas juvenile and YOY fish were mostly caught inshore (Fig. 5). Adult individuals were also the age group caught the most throughout each of the seasons, with very few juvenile and YOY individuals caught in the winter season (Figs. 3–4).

Using presence-absence data from the NEFSC BT surveys (which also included effort), we assessed which covariates might affect lumpfish presence in catch. While accounting for spatial autocorrelation, we found that lumpfish presence increased over time (b = 0.029, p < 0.0001). Here, the estimate indicates a 0.037 increase in log-odds (or 1.038 odds) for each increase in year. Lumpfish presence was also less likely in spring (b =−1.41, p < 0.0001) compared to fall surveys. We also found lumpfish presence was more likely in deeper (b = 0.006, p < 0.0001) and colder bottom depths (b =−0.39, p < 0.0001). Although all the variables were significant, the effect size for depth was relatively small compared to the other parameters. Using a separate set of analyses, we found that the yearly latitude of lumpfish presence in catch increased over time (b = 0.039, p < 0.001). In other words, lumpfish presence was more likely at higher latitudes over time (Fig. 6). In addition, lumpfish presence did not shift longitudinally over time (b = 0.001, p = 0.944).

Discussion

Although lumpfish are relatively uncommon in catch data in the GoM, we were able to identify 12,140 instances of catch when combing five disparate datasets spanning five decades (Table 1). Our findings are in line with past studies (e.g., Collette & Klein-MacPhee, 2002) showing that adult lumpfish inhabit the GoM broadly both temporally and spatially, whereas YOY and juvenile lumpfish, for the most part, are found nearshore (e.g., Moring, 2001; Rackovan & Howell, 2017). The predominance of younger fish nearshore during summer and fall, as evidenced from NH F&G and ME DMR surveys, indicates lumpfish spawn and tend egg nests in these waters. Surprisingly, these life stages were also found offshore indicating that larvae drift offshore and adults may be spawning offshore too, similar to Barents Sea lumpfish populations (Eriksen, Durif & Prozorkevich, 2014; Fig. 5C). Multiple studies have shown seasonal inshore-offshore movements of adult lumpfish associated with spawning, both in the GoM (Davenport, 1985) and elsewhere (e.g., Iceland: Kennedy et al., 2014; Kennedy & Ólafsson, 2019; Norway: Mitamura et al., 2012). However, lumpfish spawning and completing the life cycle offshore is not well documented.

We show clear differences in catch composition given differences in sampling design and between the five monitoring programs. There were differences in the catch composition for nearshore versus offshore surveys, by season, and with gear type. Most gear used in the surveys was bottom gear fished during the daytime. Because lumpfish exhibit diel vertical behavior occupying the demersal zone during the daytime, gears used by the monitoring programs (e.g., bottom trawl net, gillnet, dredge) are effective for characterizing lumpfish distribution (Eriksen, Durif & Prozorkevich, 2014; Kennedy et al., 2016). However, since the survey gears differed by mesh sizes (thus selected different sized lumpfish), survey depths (estuarine, nearshore, offshore), and areas (by state or federal waters), all catch data should be considered to more completely depict lumpfish distribution across depths, space, seasons, and life history stages in GoM waters (Kennedy & Jónsson, 2017).

We found some, albeit weak, support for our hypothesis that lumpfish distribution has shifted northeast in the past few decades (Fig. 6). Lumpfish were generally caught at deeper depths and colder bottom temperatures (Fig. 2). These results should be interpreted with caution as it is not clear how the distribution is shifting, given the limited data, despite significant increases in latitude. The GoM is one of the fastest warming bodies in the world (Pershing et al., 2015; Balch et al., 2022; GMRI, 2024). Lumpfish are a cold-adapted species, so a distributional shift with temperature aligns with their life history characteristics (Collette & Klein-MacPhee, 2002) and has been projected for more northern lumpfish populations (Rodríguez-Rey & Whittaker, 2023). Within the GoM, there have been other accounts of species moving in relation with warming waters (Le Bris et al., 2018; Friedland et al., 2023). Past work has shown that some species may see range expansions (e.g., spiny dogfish and American lobster) while other more northern species (e.g., Acadian redfish, American plaice, Atlantic cod, haddock, and thorny skate) will experience range constrictions (Kleisner et al., 2017). For lumpfish, the GoM is towards the southern end of their range and, as it continues to warm, will likely become increasingly less suitable for lumpfish (Rodríguez-Rey & Whittaker, 2023). Additional work is needed to understand how changes in other oceanographic variables (e.g., nitrate, salinity, productivity; Rodríguez-Rey & Whittaker, 2023) may interact with temperature increases (Pershing et al., 2021) and changes to fishing pressures to affect GoM species in the future.

Over the course of five decades of data, there were changes in sampling protocols within and across our five datasets. Also, effort data were not associated with all the datasets, which limited our ability to conduct more detailed analyses. Future work could use more sophisticated approaches (Fletcher Jr et al., 2019) to combine these disparate datasets more formally. We had associated metadata (e.g., bottom temperature, salinity) for only some of the datasets we used. Future work could collate and combine similar types of metadata from other available sources. Further, consideration could be given to sea temperature data as bottom temperature may not be the strongest driver to predict marine species’ shifts, but rather a temperature composite of the water column (Friedland et al., 2023). We also did not address how large perturbations (e.g., storm events) may affect population trends differently than long-term oceanographic changes. We also did not study how behavioral responses by harvesters may change with seafood demand. There could also be additional work to understand how socio-ecological dynamics may interact with extreme events (White & Wulfing, 2023) to affect lumpfish. Finally, we only examined linear trends in lumpfish catch and distribution over time. Future work could examine how population dynamics may be changing in nonlinear ways (Bruel & White, 2021; Boënnec, Dakos & Devictor, 2024). Given predicted demand for lumpfish in aquaculture, our findings highlight the need for further research on the status of lumpfish in the GoM. If exploited, proper management must ensure lumpfish are harvested responsibly and overfishing prevented, especially because adults return to the same spawning areas (Davenport, 1985; Kennedy et al., 2014) at the same times (Kennedy & Ólafsson, 2019) each year and discrete lumpfish populations exist (Langille et al., 2023; Langille et al., 2024).

Conclusions

Lumpfish in the USA present a rare opportunity to understand the population dynamics of a species that has never been exploited and provide information for sustainable harvesting practices. We found limited support for our hypothesis that lumpfish occurrences are shifting northeast along with increases in temperature. We also found that the probability of catching lumpfish increased over time and there were higher catches during Fall, at greater depths, and colder bottom temperatures. We hope this paper provides a foundation for future work on lumpfish, geared towards this emergent aquaculture sector, including lumpfish movements, genetic structure, stock assessments, and latitudinal population effects.