A call to assess the impacts of electromagnetic fields from subsea cables on the movement ecology of marine migrants

The number of subsea cables in the marine environment is likely to grow substantially in the near future. Arrays of energy‐generating windmills or wave power generators are planned for installation in the coastal waters of many countries worldwide. The electricity generated by these and other marine energy sources will be transported to shore through cables with the current and voltage creating electromagnetic fields (EMFs). Furthermore, there are also plans for the installation of undersea cables to interconnect countries and islands for the purpose of sharing power and communications. These also will generate EMFs in the marine environment. While shielding can negate the presence of direct electric fields, induced electric and magnetic fields readily penetrate into the water column. Cables carrying electric current produce anomalies in the earth's main field, which could have the potential for disrupting the migrations of fishes and diverse marine animals that rely on magnetic cues for orientation or navigation. Studies designed to test how these anthropogenic magnetic fields disrupt magnetic orientation have only recently started to be conducted. Given the cultural, economic, and conservation value of many of the species potentially at risk, such work should be immediately prioritized.


| THE EARTH'S MAGNETIC FIELD
Variation in the earth's magnetic field is primarily due to two geologic sources, the main field and crustal anomalies (Kavet, Wyman, & Klimley, 2016;Klimley, 2013). The earth's main magnetic field is generated by the circular convection of molten iron in the earth's outer core 29,000 km from the surface of the earth (Figure 1). The strength of this field from this source decreases as a negative exponent with distance from the source. The total intensity of the magnetic field increases from roughly 30,000 nT at the equator, to 40,000-50,000 nT at the mid latitudes, and to 60,000 at the poles (Skiles, 1985). The inclination, or the angle between the total magnetic field and the earth's surface, of the magnetic field is 0 at the magnetic equator and 90 at the magnetic poles. There is abundant evidence that marine animals derive their direction and even geographic position from features in the main field (for comprehensive review, see Johnsen & Lohmann, 2008;Newton, Gill, & Kajiura, 2019). For instance, scientists have conducted experiments that indicate that yellow stingrays (Urobatis jamaicensis) can distinguish between the magnetic field intensity and inclination angle, the features of the main field that are necessary to form a bi-coordinate geomagnetic map (Newton & Kajiura, 2020a), and may have a polaritybased compass (Newton & Kajiura, 2020b).The second source of variation of the earth's magnetic field is due to local alterations (i.e., anomalies) to the dipole field caused by the extrusion of basalts containing magnetite, a mineral with single pole magnetic moment. These anomalies typically consist of gradients of 10-100 nT/km (Skiles, 1985). There are two sources of these localized alterations in the main field. First, molten basalts are extruded from the mantle at spreading centers to create the oceanic plates, where the magnetite within it orients to the axis and polarity of the earth's main field at the time of extrusion. The polarities of the magnetic particles in successive sections of the crust are either aligned parallel or anti-parallel to the current axis of the earth's dipole field due to the reversal of the earth's axis over geological time every 20,000-200,000 years (Skiles, 1985). The polarity reversals create bands of stronger and weaker magnetization oriented roughly in a north-south direction that are termed magnetic lineations. The strength of these local fields also decreases as a negative exponent, but as their source is nearer the surface, they increase more rapidly with depth than the main field (see upper right, Figure 1). The second source of localized alternations are molten basalts, which are extruded from the mantle during repeated volcanic eruptions. This process creates a dipole at a volcanic crater, due to the separation of basalts with magnetite parallel and antiparallel to the current axis of the main field (Vacquier & Uyeda, 1967), and magnetic minima ("valleys" in a topographic sense) and maxima ("ridges") from the magnetite in lava flows leading away radially from their source (Klimley, 1993).
Evidence exists that marine animals use these magnetic gradients to guide themselves in the oceanic environment. The earth's magnetic field produces compass, map, and topographic (i.e., topotaxis) information that can be used by animals to orient in a highly directional manner during migration. These different navigational tactics and strategies require different sensitivities to magnetic cues (and potentially use different mechanisms of magneto-sensation). A compass sense allows animals to determine directionality and maintain a heading. This can be derived via the polarity or inclination of the magnetic field and requires sensitivity only to the direction of field lines. Magnetic compasses are widespread throughout the animal kingdom (Wiltschko & Wiltschko, 2006). A magnetic map allows animals to derive large-scale spatial information from the magnetic field, for instance, allowing them to extrapolate position relative to a home or target field. Sensitivity to seemingly subtle changes in the magnitude of inclination angle and total intensity is required and diverse marine migrants such as spiny lobsters, sea turtles, European eels, and a variety of salmonids have been shown to use magnetic map cues for orientation (Putman, 2018). Recently, Keller et al. (2021) also demonstrated that the bonnethead (Sphyrna tiburo) uses magnetic cues for homeward orientation, strongly suggesting the presence of a magnetic map. In many instances, animals use a magnetic map and compass F I G U R E 1 Creation of the earth's dipolar main field by convective movement of molten iron the outer core and alterations in the field caused by magnetite extruded from the core. Note the contribution of the local magnetization increases with depth together-the map to assess where they are and the compass to maintain a heading toward their target (Boles & Lohmann, 2003). Animals can also use magnetic topotaxis to identify unique magnetic signatures associated with crustal anomalies as landmarks for guidance along migratory paths (Klimley, 1993). Though magnetic topotaxis has been less well studied in laboratory settings than magnetic compasses and maps, tracking data paired with measurements of magnetic topography imply that the known sensitivity to magnetic intensity in animals would, in many cases, allow precise orientation at relatively fine-scales.

| EXAMPLES OF NATURAL EMFS' EFFECTS ON ANIMAL MOVEMENT
There have been different reasons proposed for the stranding of whales. One reason suggested for stranding and the death of multiple individuals of different cetaceans was the presence of infectious and parasitic diseases within their bodies, likely caused by human activity in their habitat (Diaz-Delgado et al., 2018). Another reason suggested for the deep diving Cuvier's beaked whale (Ziphius cavirostris) is disorientation caused by the propagation of high levels of low and medium frequencies by the Low Frequency Active Sonar (LFAS) used to detect quiet diesel and nuclear submarines (Frantzis, 2004). Necropsies of eight of the stranded animals in the Kyparissiakos Gulf, Greece were performed, but no apparent abnormalities or wounds were found. Additionally, the proximity of military maneuvers were already suspected in causing mass stranding of Cuviers' beaked whales off the Canary Islands (Simmonds & Lopez-Jurado, 1991;Vonk & Martin, 1989).
Stranding of whales have also been associated with temporal and spatial variation in the magnetic field. Granger, Walkowicz, Fitak, and Johnsen (2020) showed that gray whales (Eschrichtius robustus) are more likely to strand on days with relatively high levels of atmospheric radio-frequency noise, hypothesizing that this natural occurrence resulting from solar storms could disrupt their ability to perceive magnetic information. Stranding locations have been correlated with magnetic lineations that have rotated with plate movements so that the magnetic minimum intersect the coast. Such strandings are most common in Moray Firth and the Wash, located off the northeastern and southeastern coasts of Great Britain, respectively (Klinowska, 1985). Peaks in cetacean stranding also occur along the coasts of Florida and Cape Cod where a magnetic lineation intersects the coastline (Kirschvink, Dizon, & Westphal, 1986). Furthermore, scalloped hammerhead sharks (Sphyrna lewini) make nightly migrations, swimming in a highly directional manner, from a seamount to their feeding grounds 20 km distant only to return on the following morning (Klimley, 1993). The degree of orientation, determined by applying the Rayleigh coefficient to measurements from a heading sensor placed on the sharks, was 0.999, where 1.0 would be perfectly straight and 0.0 would be in random (uniform) directions (Klimley, 1993). These movements were shown with a magnetic survey of the surrounding waters to be along magnetic maxima and minima leading away from the seamount. If magnetic topotaxis requires animals to closely and continuously sense magnetic gradients emanating from the seafloor, this navigational tactic may be particularly susceptible to disruption from EMF anomalies from undersea cables.

| EXAMPLES OF ANTHROPOGENIC EMFS' EFFECTS ON ANIMAL MOVEMENT
There is evidence that EMF anomalies from cables affect the behavior of animals (Table 1). Mesocosm experiments demonstrate conspicuous increases in exploratory behavior and foraging in little skate (Leucoraja erinacea) and an increase in exploratory response in American lobster (Homarus americanus) within enclosures with an energized cable versus a nonenergized cable . Silver-stage European eels (Anguilla anguilla) tracked in the Baltic Sea carrying coded ultrasonic tags swam more slowly in an area crossed by an energized cable than in areas where no cables occurred (Westerberg & Lagenfelt, 2008). Separately, eels carrying ultrasonic tags released distant from a cable and tracked by boat returned to their migratory direction with a delay of 30 min with their trajectories veering during passage over the cable (Öhman, Sigray, & Westerberg, 2007;Westerberg & Begout-Anras, 1999). The outmigration of Chinook salmon (Oncorhynchus tshawytscha) through San Francisco Bay was monitored based on their detection by automated receivers deployed in arrays across the Benicia, Richmond, Bay, and Golden Gate Bridges in San Francisco Bay before and after the installation of a directcurrent transmission cable linking Pittsburg to San Francisco, California. The transmission line created anomalies near the Benicia Bridge, Richmond Bridge, and Bay Bridge in addition to those from the bridges (Klimley et al., 2017). The anomalies created by the cable and bridge are shown for the Richmond Bridge in Figure 2. After the cable was energized, higher proportions of fish crossed the cable and fish were more likely to be detected south of their normal migration route with the times of transit through some regions reduced during cable activity (Wyman et al., 2018).
Indirect evidence of anthropogenic EMFs impacting marine animal movements also come from "magnetic displacement" experiments. In these studies, a carefully constructed array of coiled wires around an orientation arena is used to precisely manipulate the magnetic field around an animal. Typically, the magnetic fields presented to animals exist at some distant site, such as along their migratory route or at the edge of their typical range.
Species like loggerhead sea turtles (Caretta caretta), European eel (Anguilla anguilla), and pink salmon (O. gorbuscha) will adopt headings that, had the animals been at the geographic location of the magnetic field, would guide them along their typical migratory route (Lohmann, Putman, & Lohmann, 2012;Naisbett-Jones, Putman, Stephenson, Ladak, & Young, 2017;Putman, Williams, Gallagher, & Dittman, 2020). The findings from these studies show that (a) specific components of the magnetic field are perceptible to animals, (b) demonstrate that this information is used for orientation, and (c) provide ecological context for the sensory ability (Putman, Ueda, & Noakes, 2019). Interestingly, in the context of anthropogenic EMFs, these studies also show that animals will spontaneously alter their orientation after relatively brief exposures (5-10 min) to relatively subtle changes (<5%) in magnetic intensity and inclination (Naisbett-Jones et al., 2017). Experiments in steelhead trout (O. mykiss) and loggerhead turtles further show that their ability to correctly respond to magnetic displacements is disrupted if exposed to anthropogenic magnetic fields that created unnaturally sharp gradients in the rearing environment (Putman, Meinke, et al., 2014;Putman, Scanlan, et al., 2014;Fuxjager et al., 2014). In both studies, whether the effect on magnetic navigation was permanent or diminished through time was not assessed.

| IDENTIFYING THE IMPACT OF EMFS ON ANIMAL ORIENTATION
Though sufficient to demonstrate the plausibility of impacts of EMFs on animal movements, generalizing information from the types of studies conducted to date is not yet possible (Hutchison, Secor, et al., 2020). Either studies were not designed to distinguish what aspect of the navigational system was being disrupted (e.g.,  or the anthropogenic magnetic fields may not be directly relevant to those produced by undersea cables (e.g., Putman, Meinke, et al., 2014;Putman, Scanlan, et al., 2014). Additional studies, both in the field and laboratory, need to be completed to understand to what extent anthropogenic alterations to magnetic conditions in these animals' habitats affect migrations (Table 2). There are two general ways that anthropogenic EMFs could cause problems for animals attempting to navigate using magnetic cues: (a) EMFs might disrupt the ability of an animal's magnetoreceptors to function (Engels et al., 2014) or (b) EMFs might cause the magnetic information detected by the animal to be unreliable or misleading (Putman, Meinke, et al., 2014;Putman, Scanlan, et al., 2014). Assessing impacts on magnetoreception would be difficult given that no magnetoreceptor has been definitively found in animals and thus a mechanistic understanding of how EMFs would disrupt that process cannot be satisfactorily established. In contrast, measuring the distortion of magnetic intensity, inclination, and declination over the areas occupied by cables and associated structures is quite feasible (Figure 1) and would yield valuable information as to what aspects of the magnetic navigation system of animals could be disrupted (Hutchison, Secor, et al., 2020). In either case, a key aspect of future work should be to investigate the functional implications of anthropogenic EMFs by assessing under what conditions they might alter movement patterns and, more specifically, limit the ability of animals to use magnetic cues for a map, a compass, or topotaxis.
We encourage field studies that rely on monitoring the movements of naturally migrating fish and other animals across power cables, such as those leading from arrays of wind-generating sources in the ocean. For these studies, species for which the natural path of migration is known should be prioritized to determine whether the installation of a cable affects the distribution and timing of movements. Particularly powerful to address this question would be to complete the following experiment. Individual animals in migratory condition carrying ultrasonic transmitters with accelerometer and magnetometer sensors could be tracked in a boat as they pass over energized cables while recording telemetered measurements of the local magnetic field and other environmental conditions. In tandem with field studies, laboratory experiments under controlled conditions should be carried out to investigate the magnitude and duration of exposure to EMFs from cables impact magnetic navigation. The paradigm of magnetic displacements, discussed above, would likely be particularly powerful to assess EMFs effects on the magnetic map and compass (Putman, 2018). Experiments could be conducted in which animals are exposed to EMFs (like those associated with undersea cables) for different lengths of time and then tested in magnetic displacements over a range of recovery periods to determine what level of exposures impact magnetic orientation and for how long. To assess impacts on magnetic topotaxis, the ability for animals to be conditioned to follow either magnetic minima or maxima (produced by alternately energized cables, perhaps in separate arms of a Y-maze) should be investigated with and without the addition of EMFs. With such experiments in the lab and the field focusing on a few key model species, it would be possible to make predictions about how a variety of animals that use magnetic cues for similar navigational tasks would likely be impacted by anthropogenic EMFs. With data from these empirical approaches, the effect of EMFs from cables and the associated infrastructure that are proliferating across coastal areas throughout the world can be determined on a population-and ecosystem-level for marine animals (Putman, 2018).