Hawaiian Monk Seal ( <i>Neomonachus schauinslandi</i> ) Marine Debris Entanglement Risk Varies With Size Class and Sex

Plastic pollution is a recognized threat to global oceans (National Academies of Sciences, Engineering, and Medicine 2022; United Nations Environment Programme 2021; United Nations Environment Programme 2023). Hawaiian monk seals (Neomonachus schauinslandi), listed as endangered under the US Endangered Species Act and currently numbering approximately 1600 individuals (Carretta et al. 2024), have been the subject of long-term demographic study and conservation action. Since the early 1980s, when regular monitoring of the species began, entanglement of monk seals in marine debris has been recognized as a serious threat, and efforts to disentangle seals and remove debris from their habitat have been major conservation activities (Donohue et al. 2001; Henderson 2001). Baker et al. (2024) showed that large-scale marine debris removal efforts have effectively reduced this threat to monk seals in some locations. Henderson (2001) analyzed monk seal entanglements observed during 17 years (1982–1998) and reported that pup and juvenile monk seals were more likely to become entangled than older seals, and the relative frequency of entangling debris types (i.e., nets versus lines) varied with seal age class. Here, we analyze more than 40 years of entanglement and population monitoring data collected by the US National Oceanic and Atmospheric Administration (NOAA), revealing new information about the relative entanglement risk faced by different demographic groups. We also summarize the number of documented entanglements, their outcomes, and the types of entangling material involved. The Hawaiian monk seal metapopulation comprises nine geographically defined subpopulations spanning the Archipelago (Figure 1). Eight subpopulations are in the Northwestern Hawaiian Islands (NWHI). The ninth subpopulation in the main Hawaiian Islands (MHI), consists of eight primary islands (Figure 1), where seals were rarely seen until the late 20th century, but now number approximately 400 individuals (Baker and Johanos 2004; Carretta et al. 2024). In the early 1980s, monk seal population monitoring was established in the NWHI by NOAA and has since been conducted nearly annually during seasonal field camps deployed at the six subpopulation locations from Kure Atoll to French Frigate Shoals (Figure 1). Regular monitoring at Midway Atoll was initiated in the mid-1990s when that seal subpopulation began to repopulate. Pups are weaned at roughly 5–7 weeks of age and subsequently marked with plastic flipper tags and implanted Passive Integrated Transponder (PIT) tags (Henderson and Johanos 1988; Johanos et al. 1994). During field seasons each year, standardized beach surveys were conducted regularly during which seals were resighted and identified using tags, temporary pelage bleach marks, and photographic identification (Harting et al. 2004). The reproductive status of every female sighted was recorded in each year she was resighted. Because parturition is asynchronous in Hawaiian monk seals, occurring all year with a peak from March to August, some reproductive events are not detected because monitoring typically occurs during just a portion of the period from March to August (Harting et al. 2007; Johanos et al. 1994). The annual surveys and life-long encounter histories of individual seals have yielded key demographic information such as abundance, age-specific survival and fecundity, age-sex structures, and movement among subpopulations (Baker and Thompson 2007; Baker et al. 2016; Harting et al. 2007; Johanos et al. 2014). Because of their ecological sensitivity and inaccessibility, Necker and Nihoa Islands (Figure 1) are typically surveyed only a few days per year and consequently detailed monk seal demographic data are lacking for those subpopulations. In the MHI from Kauaʻi to Hawaiʻi Island, monk seals have been monitored year-round primarily through a sighting network of volunteers, reports from the public, and government partners since the early 2000s (Baker et al. 2011). Since 2013, at privately-owned Niʻihau Island and nearby Lehua Rock, aerial, boat, and ground surveys have been conducted 1 or 2 days per year through a collaboration between the US Government and Niʻihau residents. Hawaiian monk seal entanglements were documented during the population monitoring described above and also opportunistically during 1974 to 2022. Monitoring protocols specific to entanglement included collection of information on date, location, debris type, and seal identity, size class, and sex. Wounding and deaths caused by entanglement were also noted. Entanglements were scored as due to active fishing gear, debris, or unknown. To the degree possible, the same data were obtained when entanglements were reported by non-NOAA observers. In some cases, the entangling debris was not directly observed on seals, but recent entanglement was inferred from the appearance of characteristic wounding or scarring on known individuals. In such cases, the type of debris could not be determined. We also recorded whether seals were manually disentangled by people, their debris was shed without human intervention, the entanglement resulted in death, or the seal remained entangled when last observed. Previously entangled seals were judged to have shed their debris unaided when they were resighted free of debris and no human intervention was recorded. While possible, it is very unlikely that anyone other than NOAA and its partners conducted disentanglements that were unreported. When seal sightings were recorded, a seal size-class was assigned based upon both visually assessed size and morphological characteristics (Stone 1984). Size classes included nursing pups, weaned pups, juveniles (typically 1 or 2 years old), subadults (typically 3–4 years old), and adults (typically age 5 years and older). Pups were designated as weaned from the first day they were seen without an attending adult female until December 31 of their birth year. Seal identities were classified as permanent if they were sufficiently marked (usually with tags or natural marks) to allow re-identification from year to year. Temporary identities, in contrast, indicated that seals were identifiable only during the present field season, usually by applied temporary pelage bleach marks. Entangling debris type was recorded as line, net, strap, ring, other, or unknown (Figure 2). These categories were broad and usually did not allow for resolution of the debris source. For example, “line” included all types from thick rope down to monofilament fishing line. “Net” included all types of fishing nets such as trawl, seine, and monofilament gillnet. “Strap” primarily included plastic packing bands but could include other flexible circular objects. “Ring” denoted relatively rigid circular objects. “Other” was assigned to an assortment of objects that did not fit into the aforementioned categories. In some cases, a field note helped clarify debris type in more detail, but these were not recorded with sufficient regularity and were not deemed sufficiently reliable to, for example, distinguish trawl from seine nets. However, a subtype of “ring” or “other” debris categories was separated as its own type named “eel cone.” These are highly distinctive conical objects that are part of trap gear used to catch eels and hagfish. Typically, only weaned monk seal pups were found with eel cones affixed to their muzzles. Sometimes just the eel cone base ring without the conical portion was involved. All NWHI entanglements in fishing gear were assumed to have involved abandoned, lost, or otherwise discarded fishing gear (ALDFG) rather than active fishing gear because no fisheries have operated in the NWHI since 2009. Prior to that, three NWHI fisheries (longline for tuna and billfish, handline for bottomfish, and a lobster trap fishery) did occasionally interact with monk seals (Nitta and Henderson 1993). However, with the exception of one case in which a monk seal drowned after becoming entangled in the main line and bridle of a lobster trap in 1986 (Nitta and Henderson 1993), no cases of entanglement in the types of active fishing gear used in these NWHI fisheries was recorded. In the MHI, monk seal entanglements occur in both active fishing gear and ALDFG. Thirteen MHI entanglement incidents clearly involved actively fished gillnets or monofilament line. These were omitted from our analyses as this study is focused on debris entanglement rather than direct fishery interaction. Remaining MHI entanglement incidents were classified as not involving active fishing gear or, when a determination could not be made, as unknown status. Some MHI cases were determined not to have involved active fishing gear because the entangling items were either not fishing gear or were gear types not used in Hawaiʻi (e.g., trawl or seine nets). In other cases, active fishing gear was ruled out when the material was heavily weathered, indicating it had been lost or discarded long ago. We evaluated patterns in observed entanglements in the context of available demographic data. For example, we compared the proportion of observed entanglements by size class and sex to the relative proportion of identified seals in those same categories summed across years and subpopulations. We used this approach because abundance estimates by size class and sex were often unavailable. It is possible that variation in field effort, seal identification success, and distribution of entangled seals over time and space may have influenced proportions calculated by summing across years and subpopulations. However, given that monitoring protocols were consistent, we expect these influences to be minor. Because these proportions were calculated from counts spanning many years in which individuals occurred multiple times, lack of independence precluded statistical contingency table testing. We compared the number of days individuals were seen during each year according to size and sex class as a proxy for detectability. When calculating the number of days seen, we only included seals with permanent identities because those with temporary identities were more likely to be less well-marked and therefore less likely to be identifiable if present on shore on a given day. We also tested for independence of entangling debris type and seal size and sex class using a χ2 test with p-values estimated by simulation (R Core Team 2023). For this analysis, we categorized debris as net, line, and other, lumping all remaining categories (ring, strap, eel cone, other) as other in order to achieve sufficient cell counts to allow for robust testing. Unknown debris types were omitted from this analysis. Additionally, we assessed whether entanglements were associated with reproductive state by calculating the proportion of entangled and non-entangled adult females that were documented as pregnant or post-partum each year. In making this comparison, we assumed that the probability that females were observed to be pregnant or post-partum was independent of whether they were entangled in a given year. For those who were known to have given birth in the year when they were observed entangled, we assessed the proportion of incidents that were detected before parturition, during the nursing period, and after weaning. A total of 474 Hawaiian monk seal entanglements were documented from 1974 to 2022. The true number of entanglements was undoubtedly much greater given that surveillance of seals occurred during only a small portion of the year at most sites. Further, an unknown number of entangled seals are unable to return to shore and perish at sea unobserved. In 456 (96%) of the documented incidents, the debris was observed attached to the seals, whereas 18 were inferred from distinctive wounding or scarring. In most (370) incidents, observers recorded that no wounding was associated with the entanglement, whereas injuries were observed and documented in 66 incidents. Wound status was not recorded in the remaining 38 incidents. People intervened to remove the debris from 321 entangled seals, 129 seals shed their debris unaided, 10 seals died from entanglement, and outcomes were unknown for 14 incidents. Of the 10 dead seals, nine were found deceased with debris attached. Although the remaining seal was not entangled when found, its death was attributed to peracute underwater entrapment (drowning, most likely due to net entanglement) based upon post-mortem examination findings consistent with this conclusion (Moore et al. 2013). In 15 of 474 cases (3%) we could not definitively determine whether the entangling material was debris versus active gear, and we chose to include these cases in our analyses. The number of entanglement incidents varied among demographic (size and sex class) groups (Figure 3). Weaned pups were the most frequently observed entangled size class (42% of all entanglements), followed by adults (30% of entanglements). Only 1.5% of entanglement incidents involved nursing pups (7 of 474), and these were pooled with weaned pups in subsequent analyses because nursing and weaned pups were very close in age and the sample size of nursing pups was too small for meaningful analysis on its own. Many more adult females (102) than adult males (33) were observed entangled, whereas males had somewhat higher representation among entangled subadults (40 males, 21 females) and juveniles (30 males, 25 females). Roughly equal numbers of male and female weaned pups (98 and 93, respectively) were observed entangled. Comparing the proportions of demographic groups among identified entangled seals to all identified seals provides further evidence that entanglement risk varied with size class and sex (Figure 4). While pups (sexes pooled) composed 16% of all seals identified, they accounted for 44% of entanglements. Our finding that weaned pups have elevated entanglement risk compared to older seals is consistent with Henderson (2001). Strikingly, adult females were observed entangled far more commonly than adult males. Adult females accounted for 23% of entanglements, whereas only 7% of entanglement cases involved adult males, though adult females and males represent similar proportions of identified seals (29% and 25%, respectively). The positive female bias in entanglement risk among adult seals was not detected among younger age classes (Figure 4). The disproportionate representation of specific demographic groups among entangled seals depicted in Figure 4 could have arisen from disparities in detection probabilities. To evaluate that possibility, we examined distributions of the number of days per year that individuals were seen by demographic group (Figure 5). Individual adult females tended to be resighted somewhat more frequently than individual males—a median of 14 versus 11 times per year, respectively. Similarly, individual pups (sexes pooled) were seen more frequently than seals of all other sizes (median of 19 times per year versus 12, 15, and 13 times for juveniles, subadults, and adults, respectively). Our most significant new finding is that adult female monk seals accounted for a higher proportion of entanglements than any other single size class and sex category (Figure 4). Adult females are at far greater risk of entanglement than adult males. Differences in detection probabilities, as indicated by the median number of times that individuals from various demographic groups were resighted, were too small to explain apparent differences in entanglement risk. While individual weaned pups tended to be resighted more frequently than older individual seals (Figure 5), this higher detectability alone cannot account for the fact that the proportion of weaned pups among entangled seals was nearly threefold greater than their representation among identified seals (Figure 4). Likewise, whereas individual adult females were seen somewhat more often than adult males, this cannot explain that more than three times as many adult females than adult males were observed entangled. To our knowledge, elevated entanglement risk among adult females compared to adult males has not been reported in other pinnipeds. Hawaiian monk seals are not sexually dimorphic in terms of length (Baker et al. 2014), suggesting that some factor related to female behavior or profound changes in female body morphology associated with pregnancy, lipid storage, fasting, and lactation may explain why females were observed entangled so much more commonly than adult males. If higher entanglement risks were associated with reproduction, we would expect that a higher proportion of entangled adult females would be documented as pregnant or post-partum than non-entangled seals. Yet, observed birth rates of entangled and non-entangled adult females were essentially equal, suggesting no association with reproductive state. Forty-six percent of adult females were known to give birth in the year when they were observed entangled; nearly identical to the 45% of non-entangled adult females noted as having reproduced. Pregnant monk seals gain massive lipid reserves prior to parturition, then essentially fast during lactation, losing mass and remaining on or near shore from parturition to weaning, after which they begin foraging at sea and regaining mass. We considered that entanglement risk could be affected during reproductive stages when they are fattening versus fasting and spending more or less time at sea versus on shore. Forty-four females were known to have given birth during a year when they were entangled. Of these, the date when they were first observed entangled was prior to parturition in 11 cases, during lactation in 18 cases, after weaning in 13 cases, and undetermined in two cases. Although somewhat more were first observed entangled during the lactation period compared to before birth or after weaning, this may be due to observation bias; field seasons were generally timed to intersect with pupping and nursing (as opposed to gestation). Females remain on or very near shore during lactation and are highly available for detection, and field teams intensively monitored mother-pup pairs to ascertain their status. Consequently, we cannot conclude which reproductive stage, if any, contributes to adult females' greater vulnerability to entanglement. Regardless of the factors involved, the population consequences of adult female monk seals being especially susceptible to entanglement are very concerning. Relatively small proportional changes to adult survival have relatively large effects on population growth rates and therefore the recovery potential of long-lived species (Heppell et al. 2000). Net and line accounted for of classified entangling debris items (Figure When eel cones are pooled with net and line, of documented entangling items are attributed to ALDFG. were the most entangling category and of a of plastic items typically of undetermined The of items categorized as indicated nearly all were the plastic packing bands often associated with fishing categorized as “other” of a a a of a and only In 21 cases, the type of entangling debris was 18 of these cases were inferred from or scarring after debris was were unable to ascertain the debris type in the remaining three cases. The type of entangling debris items varied among size classes of seals (Figure seals and were most commonly observed entangled in line, whereas seals and were most commonly in net were most commonly found to Of the documented eel cone entanglements, 18 were weaned one was on a and one was on a type line, or other) was not independent of seal size class and were no significant differences in the frequency of entangling debris types by sex either across or size The proportional differences we observed in types of debris that various monk seal size classes may be by the size of the debris relative to the size of the seals. The is that eel cones were found on weaned pups and seen on adults, likely because adult are than the It is also possible that the size of nets the size of in net which be than the more over small and this could explain why pups and juveniles were most commonly in whereas seals were most commonly in (Figure in marine debris, primarily but also other is a and threat to individual Hawaiian monk seals, the and to many other species and et al. far the most items that have been documented to of abandoned, lost, or otherwise discarded fishing gear (ALDFG) and plastic packing to be associated with fishing and this is a threat that is to though not without and removal of marine debris from the NWHI habitat has in the entanglement to monk seals (Baker et al. 2024). involved and with partners (Donohue Since the has primarily been conducted by a the which a total of of debris from the and islands of the NWHI in and to remove debris from seals directly the of entanglement. Harting et al. showed that such direct to individual monk seal survival have the species status. large-scale debris removal and to entangled seals, to be for as long as the Hawaiian to be the for of and other plastic was by of an early We for statistical analyses and for debris removal in the Northwestern Hawaiian Islands was by many many of We also field who have monitored NWHI Hawaiian monk seal and and partners who over MHI monk seals. These individuals and have of entangled seals from and for which we are The no of The data that the findings of this study are available in at