Risso’s dolphins (Grampus griseus, hereafter referred to as “Risso’s”) are a medium-sized, cosmopolitan odontocete found in temperate and tropical waters. They typically occur between latitudes 60°N to 60°S where sea surface temperatures are greater than 10°C; however, the highest numbers of individuals are found between 30° and 45° latitudes, in waters 15–20°C (Kruse et al., 1999; Baird, 2002; Jefferson et al., 2014b). In the eastern Pacific, Risso’s are abundant and widely dispersed year-round (Leatherwood et al., 1980). They can be found as far north as the Gulf of Alaska and south to Tierra del Fuego (Baird, 2002; Jefferson et al., 2014b). Risso’s are most often associated with shelf-edge and upper continental slope habitat, in areas of steep bottom topography with average depths ranging from 100 to 1000 m (Kruse, 1989; Baumgartner, 1997; Kruse et al., 1999; Baird, 2002; Cañadas et al., 2002; Jefferson et al., 2006; Azzellino et al., 2008). Although highest densities are found along this shelf/slope region, they utilize a wide range of habitats, from shallow, coastal areas (Boer et al., 2013; Jefferson et al., 2014b) to offshore pelagic waters (Baird, 2016). Group size is typically 10–50, although singletons and groups upwards of 4,000 individuals have been documented (Leatherwood et al., 1980; Baird, 2002; Baird, 2016).

Currently, the population abundance of Risso’s along the coasts of Washington, Oregon, and California is estimated at 6,336 (CV = 0.32) individuals (Carretta et al., 2020). Movement patterns within this population have been discerned through sightings and acoustic detections. It is thought that Risso’s in this region demonstrate broadscale seasonal shifts northward from California into Oregon/Washington in the spring/summer months coincident with rising water temperatures (Green et al., 1992). High inter-annual seasonal variability north of (Dohl et al., 1983; Kruse, 1989; Soldevilla et al., 2010), and year-round presence within, the Southern California Bight (Soldevilla et al., 2010), a productive oceanographic region off the coast of the southwestern United States (U.S.), have been documented. Long–term changes in the abundance and distribution of Risso’s have also been documented in the Southern California Bight. In the late 1950s, they were rarely encountered there (Norris and Prescott, 1961). Between 1972 and 1975, sightings increased in frequency as their distribution shifted inshore (Leatherwood et al., 1980). Starting in the El Niño years of 1982–83, occurrence further increased in the Bight and Santa Catalina Island was identified as particularly important habitat (Shane, 1995b). They continue to be regularly documented there both visually and acoustically (Soldevilla et al., 2010; Jefferson et al., 2014a; Bacon et al., 2017).

Risso’s have a primarily teuthophagus diet consisting of both squid and octopus (Clarke, 1996; Baird, 2002; Blanco et al., 2006; Jefferson et al., 2006; Bearzi et al., 2011; Yates and Palavecino-Sepúlveda, 2011; Bloch, 2012; Milani et al., 2017; Luna et al., 2021), though they also feed on fish and crustaceans (Reeves et al., 2002; Bloch, 2012). With the exception of octopus in the western Mediterranean (Blanco et al., 2006), squid appear to be their preferred cephalopod prey (Clarke, 1996; Bearzi et al., 2011; Yates and Palavecino-Sepúlveda, 2011; Bloch, 2012). In California waters, some species of squid that have been identified as important prey for Risso’s, including jumbo squid (Dosidicus gigas) (Orr, 1966; Gilly et al., 2006) and market squid (Loligo opalescens) (Kruse, 1989), perform diurnal vertical migrations (Watanabe et al., 2006). These species serve as both prey and predators within the ecosystem, feeding on deep-, mid-, and shallow-water organisms such as krill and other crustaceans, small fish, and even other cephalopods, with observations suggesting they may feed at the surface in the early evening (Markaida and Sosa-Nishizaki, 2003). Jumbo squid in particular have demonstrated daytime depths exceeding 250 m, near-surface depths at dusk, and variable depths throughout the night (Gilly et al., 2006).

Our understanding of foraging behavior of Risso’s has advanced markedly in recent years. Initial observational studies suggested Risso’s rested or traveled during the day and foraged at night (Shane, 1995b). An increase in both click rates and bouts at night suggested Risso’s foraging activity is related to diurnal patterns in squid behavior (Soldevilla et al., 2010). Most recent work indicates that Risso’s perform foraging dives during the day as well as at night (Benoit-Bird et al., 2017; Arranz et al., 2018; Benoit-Bird et al., 2019) with maximum depths exceeding 500 m particularly common during the day (Benoit-Bird et al., 2019). They regularly switch predation tactics between a near-surface generalist and deep-water squid specialist even within a single dive (Benoit-Bird et al., 2019). They plan foraging dives based on knowledge attained on previous dives (Arranz et al., 2018, Visser et al., 2021) and use a spin dive strategy to decrease transit time to exploit the diurnal mesopelagic layer (Visser et al., 2021).

Reliable assessments of Risso’s movements and diving behavior are challenging yet essential for understanding population structure, stock boundaries, habitat use, and potential exposure to anthropogenic impacts. Satellite tags can provide a more objective, detailed depiction of individual movements and behavior than visual or acoustic surveys alone. As part of a long-term research effort to examine cetacean movements, habitat use, and ecology within the Southern California Bight, we remotely deployed 16 Argos-linked satellite tags on Risso’s between 2009 and 2019, and here report on habitat use, dive behavior, and movements of these tagged individuals.

As regular inhabitants of the Southern California Bight, Risso’s have been frequently documented on line-transect surveys throughout the region (e.g., Barlow and Forney, 2007; Jefferson et al., 2014a) and across seasons in recent years (e.g., Schorr et al., 2019). Habitat used by tagged dolphins in this study ranged from the deep waters of the numerous basins to the shallow waters inshore of San Clemente and Santa Catalina Islands along the mainland coast (9–2,006 m), with Argos locations sometimes clustered around steep slope features ( Figure 1 ). Globally, densities of Risso’s appear higher over the outer continental shelf and upper regions of the continental slope (Jefferson et al., 2014b). The patchy distribution and local abundance of this species may be driven by enhanced productivity associated with currents and upwellings characteristic of these steep topographic areas (Kruse et al., 1999). Outside of the Southern California Bight, they are documented in various habitats ranging from shallow waters such as 2–5 m in Monterey Bay, CA (Jefferson et al., 2014b) to offshore pelagic waters of Hawai'i in depths exceeding 3,000 m (Baird, 2016). In the Azores around Pico Island, they can be found from 0–10 km from shore (Hartman et al., 2016) where steeply sloped nearshore topography results in water depths increasing from 200 m to 1000 m within 5 km (Hartman et al., 2008).

While relatively short deployment durations of this study preclude any meaningful discussion of large-scale movements into and out of the Southern California Bight, they do provide new insights into localized movements. Changes in Risso’s density indicate latitudinal seasonal shifts occur along the U.S. west coast (Green et al., 1992; Forney and Barlow, 1998). In the Bight, Santa Catalina Island has repeatedly been identified as important habitat for Risso’s (Shane, 1995a; Shane, 1995b; Soldevilla et al., 2010), with acoustic detections occurring on 75% of the days at a site on the south end of the island (Soldevilla et al., 2010). The near year-round acoustic presence at Santa Catalina Island may suggest that an island-associated population exists and/or a greater Southern California Bight resident population, with possible seasonal influx from a larger migratory population (Soldevilla et al., 2010). In this study, all Risso’s were tagged in the outer waters (defined as waters offshore of the mainland coast) of the Southern California Bight, mainly in the greater surrounding waters of the outer Channel Islands. Nearly all individuals spent at least some of their time within the shallower nearshore waters of these islands ( Figure 1 ). Of the 16 Risso’s, 12 approached within 20 km of Santa Catalina Island ( Figure 1 ), in all seasons and in eight different years. Of these 12, only half were tagged within 20 km of this island (min = 3 km). All but one of these six tags also left this island; the brief transmission for the sixth tag prevents further discussion of its movements. In fact, the only animal out of the 16 that demonstrated close association with Santa Catalina Island for most of its tag transmission was actually tagged in the San Nicolas Basin on the west side of San Clemente Island. While these tags may not confirm or refute the presence of island-associated individuals or populations, they do further demonstrate the importance of Santa Catalina Island for the broader, regional population throughout the year.

While animals in this study were tagged in all but one month of the year, all remained within the Southern California Bight despite demonstrating varied and often highly mobile movement patterns, and most remained relatively local to their tagging location ( Figure 3 ), suggesting year-round residence within the Bight is plausible. Alternatively, if some of these individuals were members of the larger Washington-Oregon-California migratory stock, it is entirely feasible that the limited number of tags and their generally short durations simply did not capture movements out of this region. Unfortunately, durations like these are not uncommon for remotely deployed tags on small cetaceans, where factors including surface activity, interactions with conspecifics, and the challenge of achieving optimal tag placement often lead to premature detachments.

The seasonality of inshore/offshore movements were most apparent when considering distance to any landmass, rather than to the mainland coast. According to our models, Risso’s were significantly closer to land, though this was typically an island, from September through January (p < 0.05). Although market squid spawn in the Southern California Bight year-round (Navarro et al., 2018), concentrated spawning aggregations occur nearshore over shallow sandy substrate in southern California along the mainland coast and the Channel Islands (Zeidberg et al., 2012; Van Noord, 2017) from October through April/May (Fields, 1965). Past observational research suggested that both Risso’s and short-finned pilot whales (Globicephala macrorhynchus), another squid-eating species common within the Southern California Bight prior to the 1982–83 El Niño event, shift into the Channel Islands during the winter months to feed on spawning market squid (Sinclair, 1992; Shane, 1995b). Our data suggest this seasonal, prey-driven shift has continued in the decades since. These may be fine-scale, localized shifts by animals that remain within the greater Southern California Bight throughout the year, those of migratory animals shifting into the Bight, or both.

While earlier research in the Bight suggested that Risso’s foraged at night (Shane 1995a; Soldevilla et al., 2010) with slow travel/rest the predominant daytime behaviors (Shane, 1995b; Smultea et al., 2018), more recent studies have documented daytime foraging excursions (Arranz et al., 2018; Arranz et al., 2019; Benoit-Bird et al., 2019). Dives in excess of 500 m recorded on these tags occurred across all times of day (max = 607 m day, 656 m night) ( Figure 5B ) and are comparable to results from Benoit-Bird et al. (2019). Dives were predicted to be 1.27 times deeper (e ^β; β = 0.24, SE = 0.02, p-value < 2e-16) and 4.70 times shorter (e ^β; β = -0.14, SE = 0.01, p-value < 2e-16) at night versus day ( Figures 6B , 7B ), and to occur nearly twice as often. This depth difference is potentially driven by both a greater number of dives at < 100 m depth during the day (63% day versus 35% night) and > 100 m at night (65% night versus 37% day) ( Figure 5B ). These results would align with increased nighttime foraging activities that have been described previously (Soldevilla et al., 2010; Arranz et al., 2019; Benoit-Bird et al., 2019). While Benoit-Bird et al. (2019) documented elevated Risso’s echolocation in shallow waters at night, we are unable to make further comparisons due to the limited temporal and spatial coverage of acoustic data and the lack of nighttime tag data from that study.

Described as dynamic foragers, Risso’s target different prey layers, switching between shallow water generalist and deep water specialist strategies (Benoit-Bird et al., 2019) and appear to consider optimal energy investments and tradeoffs prior to the next dive (Arranz et al., 2018; Visser et al., 2021). Jumbo squid are at depths > 250 m during the day, travel to near-surface waters at dusk, and then demonstrate highly variable dive behavior until they presumably find a zone to exploit for food (Gilly et al., 2006). In Hawai'i, the mesopelagic layer migrates from the 400–700 m depth during daylight hours to 0–400 m at night, though pilot whales appear to target the deeper 400 m layer at night, potentially in search of higher caloric prey (Owen et al., 2019). In the Azores, Risso’s target two deep scattering layers, a broad one at 500–700 m and a narrow one at 400 m, with both layers performing diel vertical migration (Visser et al., 2021). Off Santa Catalina Island, Risso’s were detected in three layers: a deep static layer at 450 m, a mid-water migrating layer at 100–300 m, and an intermittent shallow layer at 50 m along with scattered patches consisting of fish and squid (Benoit-Bird et al., 2019). It’s possible that Risso’s utilize the photic advantage of daylight to capitalize on the low energetic costs of shallow layer exploitation in between bouts of deep layer explorations (the switching behavior documented in Benoit-Bird et al., 2019) and thereby conserve energy for nighttime hunts that are directed at preferential prey, such as jumbo squid, that while vertically migratory, are not necessarily reaching shallow waters throughout the nighttime hours.

Lunar phase has been linked to changes in zooplankton biomass (Hernández-León et al., 2002) and depth (Boden and Kampa, 1967) and horizontal displacement in the mesopelagic layer in Hawaiʻi (Benoit-Bird et al., 2009). In this study, lunar phase predicted both horizontal and vertical shifts, albeit with large confidence intervals. Risso’s were located 1.45 times further away from shore (including islands) ( Figure 4E ) during a full moon than a new moon, a pattern similar to that observed in Hawaiian pilot whales (Owen et al., 2019). Lunar phase has been associated with changing use of the water column in Cuvier’s beaked whales (Ziphius cavirostris) (Barlow et al., 2020) and jumbo squid (Gilly et al., 2006), and like the Risso’s tagged in this study, pilot whales (Owen et al., 2019), Galápagos fur seals (Arctocephalus galapagoensis) (Horning and Trillmich, 1999), and bigeye tuna (Thunnus obesus) (Musyl et al., 2003) all appear to use deeper waters during a full moon. Full moon light levels at 100–150 m are equivalent to mid-day light levels at ~500 m (Clarke and Denton, 1962). Deep scattering layer species seek lower light levels to avoid predators, and are thus found at deeper depths during brightly moonlit nights (Boden and Kampa, 1967), increasing use of near-surface waters at night as the moon wanes (Hernández-León et al., 2002). It may be a cost-effective strategy for Risso’s to pursue prey at deeper depths around a full moon—as suggested by our model—when layers may provide additional biomass to target. The question also should be raised whether changes in dive behavior associated with lunar phases are driven by illumination levels themselves, or whether they are the result of an endogenous signal that is associated with the 29.5 day lunar cycle, or a combination of both. It seems plausible that light may have significant biological influence on marine systems with predominantly clear water and skies but in geographic locations where waters are more turbid and skies are obscured by clouds for at least part of the year, as they are in the Southern California Bight (e.g., Rastogi et al., 2016), an endogenous cue may play a more significant role. In (Benoit-Bird et al., 2009), lunar phase accounted for more variability in micronekton migration than surface irradiance, suggesting that lunar effects other than light levels effect these prey, and in turn, the predators that pursue them.

Results from this study suggest seasonal trends in movements, demonstrate year-round presence within the Bight, and further describes lunar and diel influences on behavior. We have presented our statistical findings and suggest biological explanations based on findings in the literature. However, we must acknowledge that statistical significance according to model output should be interpreted with caution from a biological perspective, as these data have limitations. Although several of these tags lasted nearly three weeks, the kernel density analysis was likely biased by deployment location. In addition to sample size limitations, some of these data included gaps from failed transmission and from short tag durations that didn’t span across seasons and lunar phases. The effects of Argos location quality and resolution, and the resulting prediction errors from modeled animal tracks based on Argos data, further limit interpretation of these results. While we filtered Argos locations to reduce error before using them in standardized movement models, predicted locations had standard errors ranging from 0 to 26 km. There were occasional adjustments to tag programming to optimize data collection that may have affected some statistics. When looking at environmental variables such as sea surface temperature and chlorophyll-a, in-situ measurements collected around an animal is the only way to provide accurate, real-time representation of that variable with high resolution, but this was not available for these tags. Remote sensing can capture patterns in these data over a broad spatial region, but cloud cover is an inherent limitation. Although we were able to combine a remote sensing and an in-situ dataset to derive sea surface temperature at modeled locations, we had to create 8-day averaged composites for chlorophyll-a to create gap free composites. This mismatch in resolution between animal location estimates and environmental variables is another potential source of error. It should be noted that remote sensing can underestimate values for chlorophyll-a and sea surface temperature (Gomez et al., 2020; Keates et al., 2020), and absolute values should be interpreted with caution. These aforementioned factors may have influenced the significance of these predictors in the dive models.

We believe one of the greatest values of these analyses are their ability to aid formulation of new hypotheses and research directions. Additional targeted data collection will strengthen model results and allow for more concrete biological interpretations of these findings. Deployments in other regions of the Southern California Bight as well as during specific periods where data gaps currently exist [e.g., gaps in month ( Figure 2 ) and lunar phase ( Figure 6 )] would provide a more comprehensive regional dataset. The short duration of these tags highlights the need for more tag development to improve attachment while continuing to minimize impact. A regional photo-identification catalog, genetic analyses, and the use of medium-duration, dart attached archival acoustic tags (e.g., Sweeney et al., 2022) along with prey mapping would provide meaningful connections between the existing research within this region. It is our hope that this work will foster purposeful discussions and ideas for future work to manage and protect this species.

The raw data supporting the conclusions of this article will be made available by the authors, without undue reservation.

The animal study was reviewed and approved by Cascadia Research Collective Institutional Animal Care and Use Committee.

BR, EF, GS, and SW conducted the fieldwork. BR, GS, and DS conducted the data analysis. All authors contributed to the article and approved the submitted version.

Funding for fieldwork and tags was provided by the U.S. Navy, Pacific Fleet (N66604-14-C-0145, N66604-18-P-2187) and via the Californian Cooperative Ecosystems Studies Unit (N62473-19-2-0025), the Chief of Naval Operations Environmental Readiness Division (N45) via a grant from the Naval Postgraduate School (Grant No. N000244-10-1-0050) and the Environmental Security Technology Certification Program (N66604-14-C-2438). Additional support was provided by the Office of Naval Research and the Living Marine Resources program.

We would like to thank the U.S. Navy, Pacific Fleet, N45, the Environmental Security Technology Certification Program, the Living Marine Resources program, and the Office of Naval Research for support of this research. We thank all participants from the Marine Mammal Monitoring on Navy Ranges program at the Naval Undersea Warfare Center, Jeff Foster for contributions in the field, and the Naval Postgraduate School and Californian Cooperative Ecosystems Studies Unit for contract facilitation. We thank Stacy DeRuiter for assistance with analyses and Chip Johnson, Jessica Chen, and Shannon Coates and three reviewers for their thoughtful comments and suggestions. This study was conducted under U.S. National Marine Fisheries permits No. 540-1811, 16111, 15330, 20605. Tagging was approved by the Cascadia Research Collective Institutional Animal Care and Use Committee.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.