Marine debris accumulation in the nearshore marine habitat of the endangered Hawaiian monk seal, Monachus schauinslandi 1999–2001

Marine Pollution Bulletin 46 (2003) 1385–1394 www.elsevier.com/locate/marpolbul

Marine debris accumulation in the nearshore marine habitat of the endangered Hawaiian monk seal, Monachus schauinslandi 1999–2001 Raymond C. Boland

a,*

, Mary J. Donohue

b

a

b

Pacific Islands Fisheries Science Center, National Marine Fisheries Service, 2570 Dole Street, Honolulu, HI 96822, USA University of Hawaii Sea Grant College Program, School of Ocean and Earth Science and Technology, 2525 Correa Rd., HIG 238, Honolulu, HI 96822, USA

Abstract Large amounts of marine debris are present in shallow reefs adjacent to beach haulouts of the critically endangered Hawaiian monk seal, Monachus schauinslandi. These areas serve as seal pup nurseries, and injury and death caused by entanglement in marine debris are undermining population recovery efforts. We investigated the extent of this threat by measuring the accumulation of potentially entangling derelict fishing gear in nursery zones, 1999–2001. Plots of reef 1.0–1.3 km2 at three Northwestern Hawaiian Islands were initially cleaned of derelict fishing gear in 1999 then resurveyed in 2000 and 2001. Submerged debris densities across sites ranged from 16 to 165 debris items/km2 . Resurveyed sites yielded annual marine debris accumulation rates from 0 to 141 debris items/km2 . This large range was attributed to the physiography of reef areas surveyed. Trawl net webbing was significantly more common than other types of debris recovered and represented 84% of all debris encountered, suggesting that much of the debris originated from distant North Pacific Ocean fisheries. The likely source of most debris is the multinational trawl fisheries of the North Pacific Ocean. An international solution to this problem is needed. Targeted marine debris removal is a short-term, successful, entanglement mitigation strategy. Published by Elsevier Ltd. Keywords: Marine debris; Derelict nets; Coral reefs; Accumulation rate; Hawaiian monk seal; Northwestern Hawaiian Islands

1. Introduction Marine debris is arguably the largest documented anthropogenic impact to the recovery of the endangered Hawaiian monk seal, Monachus schauinslandi (Donohue et al., 2001a; Henderson, 2001). Breeding colonies of the Hawaiian monk seal are limited to six small islands and atolls in the Northwestern Hawaiian Islands (NWHI) (Johanos and Baker, 2001) (Fig. 1). The total Hawaiian monk seal population is estimated at 1400 individuals (Forney et al., 2000) and entanglement in marine debris, particularly derelict fishing gear, is a threat to the recovery of this species (Henderson, 2001; Donohue et al., 2001a). Entanglement has been documented for 58% of all pinniped (seal and sea lion) species with detrimental effects for both individuals and populations (Laist, 1997;

*

Corresponding author.

0025-326X/$ - see front matter Published by Elsevier Ltd. doi:10.1016/S0025-326X(03)00291-1

Fowler et al., 1990; Fowler, 1987, 1985). Henderson (2001) reported a mean annual entanglement rate of 0.70% for the Hawaiian monk seal population between 1982 and 1998. This rate is near the upper range value (0.16–0.80%) for reported entanglements in other pinnipeds (Stewart and Yochem, 1990; Arnould and Croxall, 1995), and exceeds the mean annual entanglement rate (0.40%) reported for juvenile male northern fur seals (Callorhinus ursinus) for which marine debris is considered a significant source of mortality (Fowler et al., 1994; Stewart and Bengston, 2000). Bias is undoubtedly associated with published entanglement rates as these are almost exclusively determined by counting only those animals that return to shore (Laist, 1997). Most entanglement rates are conservative estimates, since animals that become entangled and die at sea are not accounted for. The Hawaiian monk seal is the only pinniped for which entanglements at sea have been documented and included in entanglement rate calculations (Henderson, 2001). At-sea entanglements of Hawaiian monk seals are observed opportunistically by

1386

R.C. Boland, M.J. Donohue / Marine Pollution Bulletin 46 (2003) 1385–1394

Fig. 1. Hawaiian Archipelago. Insets indicate high entanglement risk zone at Kure Atoll, Pearl and Hermes Atoll, and Lisianski Island.

the National Marine Fisheries Service (NMFS), the United States Fish and Wildlife Service (USFWS), and the United States Coast Guard (USCG) and account for 20 of the 204 entanglements observed from 1982 to 2000. Hawaiian monk seal entanglement rates, therefore, may be more accurate than those reported for pinnipeds with unobservable at-sea entanglement. The magnitude of debris hanging on the reefs of the NWHI was revealed during preliminary diver surveys of coral reef tracts at French Frigate Shoals (FFS) and Pearl and Hermes Atoll (PHA) (Boland, 1997) and removal efforts by divers in subsequent years (Donohue et al., 2001a). A multiagency removal effort (1996–2000)

resulted in the surveying and clearing of >63 tons of derelict fishing gear from approximately 9 km2 of reefs, primarily focusing on shallow reef tracts adjacent to the sand islets where female monk seals bear and nurse their pups (Westlake and Gilmartin, 1990). The shallow bathymetry that traps drifting debris, coupled with the routine use of these areas by seals, result in high entanglement risk zones (HERZ). The large amounts of debris recovered annually suggest that debris accumulation has not diminished despite the widespread ratification of MARPOL Annex V (an international treaty prohibiting at-sea plastic disposal) (Henderson, 2001). In this work, we resurveyed the HERZ at three NWHI

R.C. Boland, M.J. Donohue / Marine Pollution Bulletin 46 (2003) 1385–1394

to generate the first estimates of annual derelict fishing gear accumulation in coral reef habitats of the Hawaiian monk seal. We also report annual accumulation of derelict fishing gear on adjacent beaches noting its type, density, and size.

2. Methods This work is part of a multiagency effort to address the problem of marine debris in the NWHI and to mitigate entanglement mortality and morbidity of Hawaiian monk seals and other wildlife. We used the methodology detailed in Donohue et al. (2001a) to survey, remove, and analyze derelict fishing gear in the HERZ of Kure Atoll (KUR), PHA, and Lisianski Island (LIS; Fig. 1). HERZ were established at PHA and LIS in 1999 and at KUR in 2000. The three defined HERZ ranged in area between 1 and 1.3 km2 and were of similar mean depth. Each HERZ was located on the northeast side of the reef complex exposed to direct trade winds, between the seal haulout areas and the seaward barrier reef. Divers were towed behind small boats in a series of systematic transects to identify reefhung derelict fishing gear. To evaluate debris accumulation, the HERZ were resurveyed in 2000 and 2001, after initial cleaning in 1999. Accurate assessment of the seafloor surveyed in prior years was obtained by continuous logging of diversÕ tracks using global positioning system (GPS) units and a Geographic Information System (GIS). 2.1. In-water survey methods The National Oceanic and Atmospheric AdministrationÕs (NOAA) 164-ft ship Townsend Cromwell and the 225-ft USCG Cutter Kukui were support platforms during the study, 9 October–5 November, 2000, and Townsend Cromwell was used 22 October–20 November, 2001. Small craft and divers were dispatched from the support vessels to conduct submerged derelict fishing gear surveys (for a detailed description of the methodology, see Donohue et al., 2001a). The HERZ were surveyed first, followed by opportunistic surveys in additional atoll or island areas. Tracks of survey transects were logged with GPS units (Garmin 12, Garmin International) and downloaded to GIS software (ARCVIEW, ESRI Inc.) daily. Swath width was determined at the beginning and ending of each transect using a measuring tape to determine underwater visibility. Swath width never exceeded 15 m, even if measured visibility was greater. Area surveyed was calculated as the product of the transect length and swath width. Debris density was determined by dividing the total number of debris items encountered by the area size surveyed.

1387

When divers encountered derelict fishing gear, its location was documented with GPS, and the debris was marked with a buoy to facilitate subsequent removal. In situ observations of debris size were noted visually. Debris size was classified using the criteria of Donohue et al. (2001a): Class 1. Fragments of net or line less than 5 m2 . Class 2. Small amount of net or line approximately 5–10 m2 . Class 3. Moderate amount of net or line approximately 11–25 m2 . Class 4. Large amount of net or line greater than 25 m2 .

2.2. Removal and processing methods Divers carefully cut the debris free from the substrate to minimize coral damage. To preserve associated coral growth, nets completely incorporated into the reef structure and no longer an entanglement hazard were neither removed nor included in debris density estimates. Debris encountered during removal efforts, but not recorded in transect surveys, was removed but not included in debris density estimates (see Ribic et al., 1992). Scales suspended from support vessel cranes were used to determine mass of recovered debris. Due to the large volume of debris recovered the analysis of debris type was conducted on only 25% of total debris collected. Insufficient debris samples for some gear types in 2001 prevented statistical comparison of gear types. Therefore, information on debris types is not presented for 2001. Recovered debris was sorted into five categories by type: trawl net, monofilament gill net, multifilament gill net, seine net, and line (rope). All coral fragments in recovered debris were returned to the sea. 2.3. Accumulation rates A GIS overlay procedure was used to compare the initial 1999 transects at LIS and PHA to transects completed in 2000. For KUR which was not surveyed in 1999, GIS overlay was compared between 2000 and 2001. The area of overlap between the 1999 and 2000 transects was defined as the area resurveyed in 2000. The same process was followed for all three sites to estimate accumulation between 2000 and 2001 (Figs. 2–4), any debris encountered in resurveyed areas was deemed accumulated in the intervening year.

3. Results Derelict fishing gear totaling 10.8 metric tons was recovered from reef habitats at PHA, LIS, and KUR.

1388

R.C. Boland, M.J. Donohue / Marine Pollution Bulletin 46 (2003) 1385–1394

Fig. 2. Overlap of survey transects for derelict fishing gear within the Lisianski Island HERZ conducted from 1999 to 2001. Stippling indicates HERZ area. Shaded areas represent survey overlap.

Fig. 4. Overlap of survey transects for derelict fishing gear within the Kure Atoll HERZ conducted from 2000 to 2001. Stippling indicates HERZ area. Shaded areas represent survey overlap.

3.1. Area surveyed, debris density, distribution, and accumulation HERZ area ranged from 1.00 to 1.26 km2 of coral reef habitat at the sites. Surveys were conducted for three consecutive years, and annual survey coverage ranged from 27% to 69% of the HERZ area. Resurveyed areas ranged from 27% to 59% of the areas previously surveyed (Table 2). Within the three HERZ by site, debris density ranged from 16 to 165 debris items/km2 , and accumulation rates within resurveyd sites ranged from 0 to 141 debris items/ km2 per year during the three survey years (Table 3). 3.2. Debris size

Fig. 3. Overlap of survey transects for derelict fishing gear within the Pearl and Hermes Atoll HERZ conducted from 1999 to 2001. Stippling indicates HERZ area. Shaded areas represent survey overlap.

Over the two years debris was recovered, weight ranged from as little as 80 kg to as much as 7875 kg (Table 1).

Size of debris recovered at the three sites for both years is presented in Table 4. Three debris size classes were present within the PHA HERZ. Class 1 was the most common, representing 74% and 88% of recovered debris in 2000 and 2001, respectively. Class 2 debris was the second most common size, accounting for 24% in 2000 and the remaining 12% in 2001. Class 3 debris was present in 2000 and represented 2% of debris items recovered within the HERZ. KUR debris recovered included classes 1, 2, and 3 items. Class 1 accounted for 81% and 80% of debris recovered in 2000 and 2001, respectively. Size class 2 was the second most predominant debris size, representing 15% of debris within the HERZ in 2000 and 2001, and

R.C. Boland, M.J. Donohue / Marine Pollution Bulletin 46 (2003) 1385–1394

resenting 83.5% of all netting recovered for all sites combined. Excluding nets of unknown type at PHA and KUR (N ¼ 2 for each site), the proportion of net types was significantly different from a uniform distribution which would be expected if all gear types examined were contributing equally to NWHI derelict fishing gear (v2 ¼ 33:19, df ¼ 3, P < 0:001), with predominantly greater trawl net. Samples of all net types except trawl net at LIS were insufficient to statistically examine net type proportion. Monofilament and seine netting were the second and third most frequently encountered net types, representing 6.6% and 6.1% of nets by type. Multifilament gillnet accounted for 2.7% of net fragments, and the remaining 1.1% of nets were of other or unknown types. Trawl net also represented the most common net type recovered by weight, accounting for 46.9% of all debris recovered, followed by multifilament gillnet (13.6%), monofilament gillnet (12.0%), seine net (5.3%), and other or unknown net (2.9%). Miscellaneous maritime line accounted for 19.3% of debris recovered by weight. Net type by frequency encountered and weight for each site are presented in Figs. 5 and 6, respectively.

Table 1 Marine debris removed from the Northwestern Hawaiian IslandsÕ high entanglement risk zone (HERZs) by year Year

2000 2001

Site Pearl and Hermes Atoll (kg)

Kure Atoll (kg)

Lisianski Island (kg)

7875 660

1664 440

81 80

1389

class 3-sized debris items accounted for 5% of all debris recovered at KUR for both years combined. Debris items recovered at LIS also included size classes 1, 2, and 3. Class 1 debris was most common in both years, representing 80% and 55% of all recovered debris in 2000 and 2001, respectively. In 2000, the only other size class represented was class 3 (N ¼ 1). In 2001, class 2 (N ¼ 4) and class 3 (N ¼ 1) were present. 3.3. Debris type Recovered derelict fishing gear from coral reef habitats was composed primarily of netting and line. Trawl net was the most commonly encountered net type, rep-

Table 2 High entanglement risk zone (HERZ) area and portion resurveyed by site, 1999–2001 Site Kure Atoll Pearl and Hermes Atoll Lisianski Island

Total area of HERZ surveyed km2 (%)

a

km2 (%)

HERZ area (km2 )

1999

2000

2001

2000

2001

1.26 1.00 1.17

NA 0.66 (66) 0.44 (38)

0.87 (69) 0.64 (64) 0.31 (27)

0.77 (61) 0.57 (57) 0.39 (33)

NA 0.39 (59) 0.12 (27)

0.42 (49) 0.23 (59) 0.04 (33)

b

HERZ area surveyed

a The HERZ area resurveyed is the overlap of the area surveyed and the previous year(s) tracks. In 2001, the HERZ area resurveyed value includes only those areas surveyed in all three years. b From Donohue et al. (2001a).

Table 3 Debris density and accumulation rate in high entanglement risk zones (HERZs) by site and year Site

Kure Atoll Pearl and Hermes Atoll Lisianski Island


1999


2000

Density (items/km2 )

Density (items/km2 )

Accumulation (items/km2 /year)

Density (items/km2 )

Accumulation (items/km2 /year)

NA 28

165 41

NA 36

116 47

141 40

63

17

17

16

0

2001

From Donohue et al., 2001a.

Table 4 Size classes of marine debris items removed from high entanglement risk zones (HERZs) by site and year Size class 1 2 3 4

Pearl and Hermes Atoll

Kure Atoll

Lisianski Island

2000

2001

2000

2001

2000

2001

49 16 1 0

37 5 0 0

116 22 6 0

74 14 5 0

7 0 1 0

6 4 1 0

1390

R.C. Boland, M.J. Donohue / Marine Pollution Bulletin 46 (2003) 1385–1394

Towboard surveys using divers either on scuba or snorkeling gear have been used extensively in the NWHI (Parrish and Polovina, 1994; Donohue et al., 2001a). Divers are trained to deal with the specific hazards (entanglement, collision with coral heads, seperation) associated with towboarding. The danger of shark strike does exist but this risk is minimized by towing in favorable conditions. Despite this, the author has had several encounters with large sharks while towing, including a Tiger shark striking his towline. While the danger exists, most sharks seem more curious than predatory when investigating towed divers.

Fig. 5. Frequency of derelict fishing gear types encountered at Lisianski Island, Pearl and Hermes and Kure Atolls in 2000. At Pearl and Hermes and Kure Atolls, there was significantly greater trawl net than other debris types (P < 0:001 for both). Insufficient samples of net types at Lisianski Island prevented statistical testing of net type proportion at that site.

Fig. 6. Percent frequency of derelict fishing gear types recovered by weight at Lisianski Island and Pearl and Hermes and Kure Atolls in 2000.

4. Discussion 4.1. Potential sampling bias in methods Towboarding has been determined to be an effective method to survey benthic targets (Fernandes, 1990; Fernandes et al., 1990; Moran and DeÕath, 1992). Fernandes (1990) tested differences in the sightability of small targets (40 cm diameter) using different survey widths. A survey width of 9–15 m, consistent with our methods, had the highest correlation of sighted targets vs. true targets. Nearly all pieces of marine debris encountered in this study were relatively large targets that tended to float up from the seafloor, making them conspicuous and difficult to miss.

4.2. Area surveyed, debris density, distribution, and accumulation Of all NWHI sites where underwater debris surveys have been conducted, the greatest debris density was documented in the KUR HERZ (165 debris items/km2 ) in 2000 (see Donohue et al., 2001a). The debris density at KUR outside the HERZ in 2000 (78 debris items/ km2 ) was also the highest reported for any area exclusive of HERZ at any other NWHI surveyed. Resurveys of the Kure HERZ in 2001 suggested that debris accumulation at this site was high. HERZ debris density in 2001 (116 debris items/km2 ), while lower than that reported in 2000, was greater than that ever reported for LIS or PHA. Annual debris accumulation at KUR was also an order of magnitude greater than at any other site examined. The high debris density at KUR, particularly within the HERZ, is likely the result of KURÕs relative geography, physiography, and the absence of any previous underwater cleanups of derelict fishing gear at this site. KUR is the most northwesterly emergent atoll in the Hawaiian Archipelago. The convergence of oceanic surface waters as a result of atmospheric forcing by the North Pacific Ocean subtropical high has been proposed as a mechanism for the disproportionately large debris accumulation in the NWHI (Ingraham and Ebbesneyer, 2001; Donohue et al., 2001a,b). This convergence zone is regularly defined from January to March in the region immediately north of KUR, latitude 30–42°N and is known to seasonally oscillate as far south as latitude 28°N, directly intersecting KUR (Polovina et al., 2001). KUR would be expected to encounter and filter aggregated floating debris more frequently than other NWHI, resulting in high debris densities. The physiography of the KUR HERZ, consisting of patch reefs situated between the main island and the extensive barrier reef, results in waters relatively protected from wave surge, causing debris forced over the barrier reef to remain in the HERZ. Additionally, initial debris levels documented in 2000 resulted from many years of marine debris accumulation.

R.C. Boland, M.J. Donohue / Marine Pollution Bulletin 46 (2003) 1385–1394

Despite the annual removal of marine debris from the PHA HERZ, debris densities were higher in 2000 (46%) and 2001 (68%) than the initial survey (1999). HERZ survey coverage in 2000 and 2001 (64% and 57% of HERZ surveyed, respectively) was similar to that in 1999 (66%; Donohue et al., 2001a). In contrast to increasing debris densities at PHA, debris accumulation rates in the PHA HERZ were nearly identical for the 2 years measured (mean debris accumulation ¼ 38.7 ± 0.2 (SD) debris items/km2 ). Two possible explanations for the increased debris density exist. The 1999 effort may have missed areas that held significant amounts of debris that were later surveyed in 2000 and 2001, when nearly 40% and 60% of the area surveyed had not been previously surveyed. The relative fouling of debris by encrusting organisms supported this notion. In 1999, over 50% of debris recovered at PHA exhibited light or no organic fouling (Donohue et al., 2001a). In contrast, over 60% of debris recovered at PHA in 2000 was moderately to heavily fouled (Donohue, unpublished data). Heavy fouling indicates a longer in situ time, and it indicates that the debris is old and likely outside previously surveyed areas. These areas were also outside of the resurveyed areas and thus not part of the accumulation rate estimation. A second possible explanation for a temporal increase in the debris density may be related to debris size. In 2000 and 2001, most debris recovered was predominately in smaller size classes, while in 1999 most debris recovered was larger (Donohue et al., 2001a), and in 2001 no debris items of classes 3 or 4 were recovered. The largest measured increase in debris density occurred between 1999 and 2000. Less variation in debris density was measured between 2000 and 2001: 41.0 and 47.0 debris items/km2 of coral reef habitat, respectively. Additional debris density studies planned for 2002 may indicate whether densities measured in 2000 and 2001 are more characteristic of overall PHA debris densities. In contrast to PHA, resurveys of the LIS HERZ resulted in lower debris densities than reported during the initial year of surveys, which included debris accumulated for multiple years. Subsequent surveys produced a mean coral reef debris density of 20.7 ± 7.5 (SD) debris items/km2 as compared to the initial 1999 value of 62.2 debris items/km2 (Donohue et al., 2001b). Further, from 2000 to 2001 estimated marine accumulation decreased to zero from 18 debris items/km2 for the period 1999– 2000. The relatively low underwater debris densities and minimal marine accumulation at LIS reflected the islandÕs southeastern position relative to PHA and KUR and a physiography atypical of a classic atoll. Thus, LIS is subject to debris carried into the NWHI region by the North Pacific Ocean subtropical high less frequently than PHA or KUR. LISÕs physiography is also such that much debris approaching the island is eventually deposited on island beaches.

1391

The difference in marine debris accumulation rates by site reported here is partly explained by the differences in reef morphology of the three sites. LIS does not have an extensive barrier reef bordering the HERZ, whereas PHA and KUR do. The environment on the shallow interior of barrier reefs is characteristically quiescent water, where debris items can become entangled and settle after being forced over the barrier reef by waves and winds. At PHA and KUR, debris items likely settle and remain in the HERZ because of this physiography. In contrast, the LIS HERZ is bordered by little to no barrier reef, with more turbulent water flow and less aquatic debris accumulation. Although all HERZ are located in the northeast areas of the islands and atolls, and experience similar weather conditions and sea states, the barrier reefs at PHA and KUR protect the HERZ, effectively slowing wind, and sea. At LIS, however, the HERZ is subject to significant wind, and sea, particularly during winter months. Overall, this results in relatively more debris deposition on the coral reefs at PHA and relatively more beach deposition at LIS. Significant amounts of debris were also recovered on PHA, LIS, and KUR beaches. Terrestrial debris accumulation at PHA from 1999 to 2000 (1991 kg) was greater than that for the previous (1140 kg; Donohue et al., 2001a) or subsequent year (900 kg). Although terrestrial debris accumulation at LIS remained high from 1999 to 2001, it decreased linearly by approximately half each year, with the greatest terrestrial accumulation noted from 1998 to 1999 (4533 kg; Donohue et al., 2001a). Terrestrial debris accumulation at KUR was similar for the 2 years measured (357 ± 9.9 (SD) kg), but only represents accumulation during the months of August to October. For the rest of the year derelict fishing gear that washes up on KUR beaches are incinerated by the State of Hawaii without recording weights. Annual terrestrial debris accumulation at KUR is unknown. The influence of site physiography on whether debris is likely to settle on coral reefs or wash ashore is demonstrated by the relative amount of debris recovered onshore as compared to that underwater. Debris recovered from LIS waters accounted for a mean of just 13% of the total weight of debris recovered from the island during 1999 and 2001 (1999 data from Donohue et al., 2001a). In contrast, debris recovered in the water at PHA accounted for a mean of 80% of all debris recovered at this site for these same years (1999 data from Donohue et al., 2001a). Incomplete data on terrestrial debris accumulation at KUR prevent similar comparisons at that site. Debris densities reported here generally exceed those reported for pelagic waters of the North Pacific Ocean, an area of high fishing effort. Day et al. (1990) reported debris densities as high as 36.7 debris items/km2 ,

1392

R.C. Boland, M.J. Donohue / Marine Pollution Bulletin 46 (2003) 1385–1394

quantified debris type, and found that the greatest densities of derelict gill net and unidentified net were immediately north of the Hawaiian Archipelago. Trawl netting density was greatest in subarctic waters, followed by waters near the Hawaiian Islands (Day et al., 1990). A similar study by Matsuura and Keiichi (1992) noted that marine debris density was greatest near continental land masses. However, the Matsuura and Keiichi (1992) study included debris items originating terrestrially. When land-based items were excluded and debris exclusively associated with fishing efforts examined, the density was greatest immediately northeast of the Hawaiian Archipelago (Matsuura and Keiichi, 1992). 4.3. Debris size The predominance of items in the smaller debris categories recovered underwater suggests that the inwater deposition of larger debris items is infrequent in the HERZ, that larger items are broken up into smaller items by wave action, or that larger items may be more readily deposited on beaches. In 2000, just 15 of 315 debris items found were of the larger size classes 3 or 4. All of the class 4 items (N ¼ 3) and one of the class 3 items were located outside the PHA HERZ. In 2001, the lack of any size class 4 items and the low number of size class 3 items (N ¼ 12, for all sites combined) in the HERZ support the supposition that larger debris items do not frequently remain there. 4.4. Debris types Debris types found in the HERZ are consistent with commercial fishing gear used in the North Pacific Ocean. Trawl net webbing represented the majority of sampled net webbing encountered at all sites (83.5%). This is consistent with prior studies examining recovered derelict fishing gear in the NWHI (Donohue et al., 2001a). Since no trawl net fisheries occur in the region of the Hawaiian Archipelago, the source of these nets is likely the subarctic waters near trawl fishery locations that have the highest reported densities of floating trawl net webbing. The second and third highest densities occur just north of the Hawaiian Archipelago (Day et al., 1990). The seasonal north–south oscillation of the north Pacific convergence zone deposits this debris in the critical habitat of the Hawaiian monk seal. 4.5. Population dynamics of debris In 1985, Gerrodette suggested that debris be thought of as a dynamic population, and theorized that debris was a population affected by births and deaths (Gerrodette, 1985). Previous papers on debris density have addressed only the population size aspect of this theory. We were able to study three separate debris populations

for 2 or 3 years. Birth or recruitment of debris was defined as debris settling on the reef. Death was defined as the removal of the debris item from the reef. Our efforts provided an estimated population density and birthrate through debris density and accumulation rate. These are two key elements in determining cleanup efforts, or in terms of population dynamics, population control. We treated our data in a similar population-based analysis. The yearly survey tracks were divided into 200 m segments, and the amount of debris within segments was tallied and treated as a debris population sample. Samples were then graphically compared at all sites. The segment density was plotted as a Poisson distribution (Fig. 7). This sampling indicated that the debris removal efforts kept up with the accumulation rate, or ‘‘recruitment rate’’, at all HERZ. At KUR and LIS, the sampling detected a decline in the debris ‘‘population’’. In addition to direct mortality resulting from drowning or injury from lacerations and infection, fitness costs have been associated with pinniped marine debris entanglement. Such costs have been documented in Northern fur seals. Seals suffering entanglement have lower reproductive success, represented by reduced preweaning pup growth (Delong et al., 1988). Other studies have demonstrated that entangled pinnipeds must increase metabolism to compensate for increased drag during swimming caused by entangling debris (Feldcamp et al., 1988). Contending with increased stress as a result of entanglement may be difficult or impossible. One hypothesis for the lack of recovery of the Hawaiian monk seal is prey limitation, which may be exacerbated by entanglement in marine debris. Because of these direct and indirect effects, continual accumulation of marine debris in the critical habitat of the Hawaiian monk seal likely contributes to this speciesÕ lack of recovery. Unlike prey limitation, habitat loss, or other ecological factors that may be inhibiting the recovery of the Hawaiian monk seal, marine debris is a tangible factor that can be addressed and mitigated. Mitigation actions should vary in relation to the effects of physiography and relative geography. For those sites without a welldefined barrier reef (LIS, Laysan Island), we propose targeting debris removal from shore, with less emphasis on submerged debris. For those sites with classic atoll configurations (PHA, KUR and Midway Atolls, and French Frigate Shoals) we propose intensive survey and removal of reef-hung derelict fishing gear. Overall, more marine debris mitigation efforts in the Northwestern portion of the NWHI chain are dictated by the behavior of the North Pacific Ocean subtropical high and associated greater relative aggregation of marine debris. The predominance of nonlocal sources of debris in the critical habitat of the Hawaiian monk seal necessitate a Pacific-wide approach to the ultimate mitigation of Hawaiian monk seal entanglement.

R.C. Boland, M.J. Donohue / Marine Pollution Bulletin 46 (2003) 1385–1394

1393

Fig. 7. Population variability of: (A) Lisianski Island, (B) Pearl and Hermes and (C) Kure Atolls from 1999 to 2001. As 0 approaches 1.0, there are more 200 m segments with no debris, indicating a decrease in debris ‘‘population’’. At Lisianski and Kure Atoll the debris removal effort was large enough to cause a decline in the ‘‘population’’.

Acknowledgements We thank the crew of the NOAA ship Townsend Cromwell for logistical support necessary to complete this study and the crew of the United States Coast Guard (USCG) Cutter Kukui for logistical and dive support, including the assistance of USCG divers. The City and County of Honolulu, Browning-Ferris Industries/Horizon Waste Services Inc. and Covanta Energy provided critical support through disposal and recycling of recovered debris. We thank the US Fish and Wildlife Service (USFWS), The Ocean Conservancy, and The Ocean Futures Society for providing field personnel. We also thank Athline Clark, Spencer Frary, Ron Hoeke, Nancy Hoffman, Stephani Holzwarth, Karen Geisler, Scott Godwin, Rob Marshall, Erin McCarthy, Jen Schorr, Cindy Thomas, Dean Uyeno, and Nina Young for field assistance. Editorial and graphics assistance was provided by Judith Kendig and Deborah Yamaguchi, respectively. This work was conducted under the USFWS National Wildlife Refuge special use permits

00019 and 01012, and the State of HawaiiÕs Department of Land and Natural Resources permit numbers SEPO 00-918.2 and 2002-23.

References Arnould, J.P.Y., Croxall, J.P., 1995. Trends in entanglement of Antarctic fur seals (Arctocephalus gazella) in man-made debris at South Georgia. Mar. Poll. Bull. 30 (11), 707–712. Boland, R.C., 1997. A preliminary survey of the underwater accumulation of derelict nets at French Frigate Shoals. Honolulu Laboratory, Southwest Fisheries Science Center, National Marine Fisheries Service, NOAA, Honolulu, Hawaii 96822-2396, Administrative Report H-97-13, 9 p. Day, R.H., Shaw, D.G., Ignell, S.E., 1990. The quantitative distribution and characteristics of marine debris in the North Pacific Ocean, 1984–1988. In: Shomura, R.S., Godfrey, M.L. (Eds.), Proceedings of the Second International Conference on Marine Debris, 1989. US Department of Commerce, NOAA Technical Memo. NOAA-TM-NMFS-SWFSC-154, pp. 182–211. Delong, R.L., Dawson, P., Gearin, P.J., 1988. Incidence and impact of entanglement in netting debris on northern fur seal pups and adult females, St. Paul Island, Alaska. In: Kozloff, P.,

1394

R.C. Boland, M.J. Donohue / Marine Pollution Bulletin 46 (2003) 1385–1394

Kajimura, H. (Eds.), Fur Seal Investigations 1985. US Department of Commerce, NOAA Technical Memo. NOAA-TM-NMFS-F/ NWC-146, pp. 58–68. Donohue, M.J., Boland, R.C., Sramek, C.M., Antonelis, G.A., 2001a. Derelict fishing gear in the Northwestern Hawaiian Islands: diving surveys and debris removal confirm threat to coral reef ecosystems. Mar. Poll. Bull. 42 (12), 1301–1312. Donohue, M.J., Polovina, J., Foley, D., Brainard, R., Laurs, M., 2001b. Hawaiian monk seal entanglement and El Ni~ no: a link between and endangered species, pollution, and oceanography? In: 14th International Conference on the Biology of Marine Mammals, Vancouver, BC, 28 November–3 December 2001 (Oral presentation abstract). Feldcamp, S.D., Costa, D.P., DeKrey, G.K., 1988. Energetic and behavioral effects of net entanglement of juvenile northern fur seals, Callorhinus ursinus. Fish. Bull. 87 (1), 85–94. Fernandes, L., 1990. Effect of the distribution and density of benthic target organisms on manta tow estimates of their abundance. Coral Reefs 9, 161–165. Fernandes, L., Marsh, H., Moran, P.J., Sinclair, D.F., 1990. Bias in manta tow surveys of Acanthaster planci. Coral Reefs 9, 155–160. Forney, K.A., Barlow, J., Muto, M.M., Lowry, M., Baker, J., Cameron, G., Molby, J., Stinchcomb, C., Carretta, J.V., 2000. US Pacific Marine Mammal Stock Assessments: 2000. US Department of Commerce, NOAA Technical Memo. NOAA-TM-NMFSSWFSC-300. Fowler, C.W., 1985. An evaluation of the role of entanglement in the population dynamics of northern fur seals on the Pribilof Islands. In: Shomura, R.S., Yoshida, H.O. (Eds.), Proceedings of the Workshop on the Fate and Impact of Marine Debris, 1984. US Department of Commerce, NOAA Technical Memo. NOAA-TMNMFS-SWFSC-54, pp. 291–307. Fowler, C.W., 1987. Marine debris and northern fur seals; a case study. Mar. Poll. Bull. 18 (6B), 326–335. Fowler, C.W., Merrick, R., Baker, J.D., 1990. Studies of the population level effects of entanglement on northern fur seals. In: Shomura, R.S., Godfrey, M.L. (Eds.), Proceedings of the Second International Conference on Marine Debris, 1989. US Department of Commerce, NOAA Technical Memo. NOAA-TM-NMFSSWFSC-154, pp. 453–474. Fowler, C.W., Baker, J.D., Ream, R., Robson, B.W., Kiyota, M., 1994. Entanglement studies on juvenile northern fur seals, St. Paul Island. In: Sinclair, E. (Ed.), Fur Seal Investigations, 1992, US Department of Commerce, NOAA Technical Memo. NOAA-TMNMFS-AFSC-45, pp. 101–161. Gerrodette, T., 1985. Toward a Population Dynamic of Marine Debris. In: Shomura, R.S., Yoshida. H.O. (Eds.), Proceedings of the workshop on the fate and impact of marine debris, 1984, US Department of Commerce, NOAA Technical Memo. NOAA-TMNMFS-SWFSC-54, pp. 291–307.

Henderson, J.H., 2001. A pre- and post-MARPOL Annex V summary of Hawaiian monk seal entanglements and marine debris accumulation in the Northwestern Hawaiian Islands. Mar. Poll. Bull. 42 (7), 590–597. Ingraham Jr., W.J., Ebbesneyer, C.C., 2001 Surface current concentration of floating marine debris in the North Pacific Ocean: 12-year OSCURS model experiments. In: Proceedings of the International Conference on Derelict Fishing Gear and the Ocean Environment, 2001, Hawaiian Islands Humpback Whale National Marine Sanctuary Publication. Johanos, T.C., Baker, J.D., 2001. The Hawaiian monk seal in the Northwestern Hawaiian Islands, 2000. US Department of Commerce, NOAA Technical Memo. NOAA-TM-NMFS-SWFSC-340. Laist, D.W., 1997. Impacts of marine debris: entanglement of marine life in marine debris including a comprehensive list of species with entanglement records. In: Coe, J.M., Rogers, D.B. (Eds.), Marine Debris Sources, Impacts, and Solutions. Springer-Verlag, NY. Matsuura, S., Keiichi, N., 1992. Distribution of floating debris in the North Pacific Ocean: sighting surveys 1986–1991. In: Coe, J.M., Rogers, D.B. (Eds.), Marine Debris Sources, Impacts, and Solutions. Springer-Verlag, NY. Moran, P.J., DeÕath, G., 1992. Suitability of the manta tow technique for estimating relative and absolute abundances of crown of thorns starfish (Acanthaster planci L.) and corals. Aust. J. Mar. Freshwater. Res. 43, 357–378. Parrish, F.A., Polovina, J.J., 1994. Habitat thresholds and bottlenecks in production of the spiny lobster (Panulirus marginatus) in the Northwestern Hawaiian Islands. Bull. Mar. Sci. 54 (1), 151– 163. Polovina, J.J., Howell, E., Kobayashi, D.R., Seki, M.R., 2001. The transition zone chlorophyll front, a dynamic global feature defining migration and forage habitat for marine resources. Prog. in Oceanogr. 49, 469–483. Ribic, C.A., Dixon, T.R., Vining, I., 1992. Marine debris survey manual. US Department of Commerce. NOAA Technical Report NMFS 108. Washington, DC, GPO. Stewart, B.S., Bengston, J., 2000. Foraging effort responses of northern fur seals to entanglement in derelict fishing gear fragments. Poster abstract. International Marine Debris Conference on Derelict Fishing Gear and the Ocean Environment, Honolulu, HI, 7–11 August 2000. Stewart, B.S., Yochem, P.K., 1990. Pinniped entanglement in synthetic marine debris in the Southern California Bight. In: Shomura, R.S., Godfrey, M.L. (Eds.), Proceedings of the Second International Conference on Marine Debris, 1989. US Department of Commerce NOAA Technical Memo. NOAA-TM-NMFS-SWFSC-154, pp. 554–561. Westlake, R.L., Gilmartin, W.G., 1990. Hawaiian monk seal pupping locations in the Northwestern Hawaiian Islands. Pac. Sci. 44 (4), 366–383.