Mapping elemental contamination on Palmyra Atoll National Wildlife Refuge

Marine Pollution Bulletin 128 (2018) 97–105

Contents lists available at ScienceDirect

Marine Pollution Bulletin journal homepage: www.elsevier.com/locate/marpolbul

Baseline

Mapping elemental contamination on Palmyra Atoll National Wildlife Refuge

T

⁎

Matthew A. Struckhoffa, , Carl E. Orazioa, Donald E. Tillitta, David K. Shaverb,1, Diana M. Papouliasa,1 a b

US Geological Survey, Columbia Environmental Research Center, 4200 New Haven Rd., Columbia, MO 65203, USA US Geological Survey, Mid-Continent Mapping Center, 1400 Independence Rd., Rolla, MO 65401, USA

A R T I C L E I N F O

A B S T R A C T

Keywords: Palmyra Atoll Elemental contaminants WWII Pacific bases Line Islands Geospatial data Hand-held X-ray fluorescence spectrometry

Palmyra Atoll, once a WWII U.S. Navy air station, is now a U.S. National Wildlife Refuge with nearly 50 km2 of coral reef and 275 ha of emergent lands with forests of Pisonia grandis trees and colonies of several bird species. Due to the known elemental and organic contamination from chemicals associated with aviation, power generation and transmission, waste management, and other air station activities, a screening survey to map elemental concentrations was conducted. A map of 1944 Navy facilities was georeferenced and identifiable features were digitized. These data informed a targeted survey of 25 elements in soils and sediment at locations known or suspected to be contaminated, using a hand-held X-ray fluorescence spectrometer. At dozens of locations, concentrations of elements exceeded established soil and marine sediment thresholds for adverse ecological effects. Results were compiled into a publically available geospatial dataset to inform potential remediation and habitat restoration activities.

Numerous Pacific islands served as U.S. Navy facilities during World War II (WWII) and are threatened by legacy contamination in the form of trace metals, persistent organic pollutants, petroleum-derived hydrocarbons, and radioactive materials (Nautilus Institute, 2017). Conservation of fish and wildlife species and protecting human health on lands formerly used for military purposes requires preventing and minimizing adverse effects from exposure to legacy contaminants. In recognition of this fact, contaminated materials have been and continue to be removed from a number of islands where military operations have historically occurred, including Johnston Atoll, Wake Island, Midway, and American Samoa (Smith, 2014). There is a continuing need for geospatial data to support these clean-up efforts, as well as to reduce the risks of exposure to fish, wildlife, and humans (Dell, 2014). Palmyra Atoll is part of the Pacific Remote Islands Marine National Monument, 1773 km (1056 miles) south-southwest of Hawaii at 5° 53′ N latitude and 162′ 5″W longitude, toward the northwest end of the Line Islands (Fig. 1). Maximum natural elevation is approximately 2 m above mean sea level, and in most places the water table lies only 1 m below the surface. Average monthly air temperatures range from 23 to 30° C and average annual rainfall is 4060 mm (www.weatherbase.com). Dominant vegetation types include coconut palm forest (Cocos nucifera), Scaveola – Tournefortia scrub forest, Pandanus forest, and Pisonia

⁎

1

grandis beach forest. The atoll was once comprised of numerous islets in a horseshoe shape open to the west and totaling about 120 ha in total land area. It has been uninhabited throughout most of its known history, but during WWII, the atoll was greatly modified to create a U.S. Naval Air Station. Areas of the reef were dredged, and the materials were used to increase land area for operational facilities and to create a landing strip for aircraft (Fig. 2). An overview of ownership of the atoll islands and waters is important to consider in order to address a range of resource management goals relating to contaminant remediation and habitat restoration (Fig. 2). Between WWII and 2000, the entire atoll was privately owned, and Home Island remains so today (The Nature Conservancy, 2017). In 2000, The Nature Conservancy (TNC) acquired the rest of the atoll to protect its quality reef, and currently owns nearly all of Cooper Island and Menge Island (97 ha). Areas owned and managed by TNC include headquarters for all management and research activities on the atoll, including the operational runway originally built by the Federal Aviation Administration prior to WWII (Platt, personal communication). In 2001, several islets and small portions of Cooper Island (totaling 141 ha) were purchased by the Department of Interior (DOI) U.S. Fish and Wildlife Service (USFWS) to form the Palmyra Atoll National

Corresponding author. E-mail addresses: mstruckhoff@usgs.gov (M.A. Struckhoff), [email protected] (C.E. Orazio), [email protected] (D.E. Tillitt). Retired.

https://doi.org/10.1016/j.marpolbul.2017.12.065 Received 31 May 2017; Received in revised form 27 December 2017; Accepted 30 December 2017 0025-326X/ Published by Elsevier Ltd.

Marine Pollution Bulletin 128 (2018) 97–105

M.A. Struckhoff et al.

most of the Palmyra coral reef system was not disturbed. The reef is considered pristine because it is not subject to overfishing and coastal runoff that affect many reefs throughout the world. The reef system supports 125 coral species and a predator-dominated food web with numerous species of sharks (Stevenson et al., 2007; The Nature Conservancy, 2017). Much of the terrestrial management at Palmyra is intended to support sea bird breeding and resting grounds, and to protect and restore stands of Pisonia grandis, a species in decline throughout its range (Batianoff et al., 2010; Handler et al., 2007). The islands provide nesting habitat for more than one million marine birds, including sooty terns (Onychoprion fuscatus), fairy terns (Sterna nereis), boobies (Sula spp.), bristled-thigh curlews (Numenius tahitiensis), and red-tailed tropic birds (Phaethon rubricauda). Other notable natural resources at Palmyra include the world's largest terrestrial arthropod, the endangered coconut crab (Birgus latro). The atoll is also a nesting site for green sea turtles (Chelonia mydas) (Kropidlowski, personal communication). Chemical contaminants and debris remaining at Palmyra Atoll may threaten the integrity of both terrestrial and marine ecosystems. At the conclusion of WWII, the Navy abandoned many potential sources of contamination in-situ, including electrical transformers, unexploded ordnance, vehicles, fuel tanks, and mechanical and industrial facilities that supported the naval air station. Some buildings were left in place; others were disassembled or bulldozed. Large debris piles were either burned or allowed to deteriorate in both terrestrial and near-shore locations. Although the locations of these piles and buffers around them are in the OIA exclusion areas and not under USFWS ownership, they nonetheless threaten atoll resources based on the exposure potential from the migration of contaminants. There have been attempts both to document and clean up contaminated sites on the Palmyra Atoll. A Defense Environmental Restoration Program review in 1987 identified the presence of legacy contaminants from use by the Navy during WWII (U.S. Army Corps of

Fig. 1. Location of Palmyra Atoll National Wildlife Refuge. Image credit: Palmyra Atoll Research Consortium (Suchanek, 2012).

Wildlife Refuge, which also includes the lagoons and open ocean out to 12 nautical miles from the atoll. The DOI Office of Insular Affairs (OIA) retained 8 near-shore “exclusion areas” because of known or suspected contamination. An additional large exclusion area (652 km2) surrounding a Navy explosives dumping area west-southwest of the atoll partially intersects the 12 mile ownership buffer of the refuge. Dredging during WWII affected only a relatively small area, and

Fig. 2. Satellite image of Palmyra Atoll in 2007 showing current ownership.

98

Marine Pollution Bulletin 128 (2018) 97–105

M.A. Struckhoff et al.

“waste” (unmapped debris not in exclusion areas), or “other”. Each area was physically visited by our team and searched for potential sources of contaminants (e.g., piles of waste or equipment). The local topography and building features (e.g., trenches) were used to determine where contaminants, if present, were likely to have collected over time. From this location, 5 sub-samples approximately 8 cm deep were collected with a large stainless steel spoon over an area approximately 4 m2. The sub-samples were combined and homogenized in a clean 1-gallon plastic container, and approximately 100 g to 200 g was transferred to a clean polyethylene bag for XRF analysis. GPS coordinates were collected from the center of each sampled area. Samples were classified as “marine sediment” if they were collected from intertidal areas or from the coral reef; otherwise, they were classified as “soil”. Elemental concentrations were measured using a hand-held XRF (Niton XLt 793Y) according to manufacturer recommendations and internally calibrated once each day prior to sampling. For each reading, the XRF was set to integrate for 60 s on each of two internal filters (one for heavier atomic mass elements, one for lighter elements). Three replicate measurements were collected for each soil or sediment sample, with bag contents shaken to remix between readings to analyze a different portion of the sample. The three readings were averaged for each sample. Elements below the limit of detection (LOD) in one or more of the measurements were considered not detected. The XRF analyses were of freshly collected samples of varying moisture content and concentrations are reported as milligrams per kilogram (mg/kg) on a wet-weight basis. A subset of seven samples were air-dried and reanalyzed to allow comparisons of wet and dry concentration measurements; these are presented as separate records within the data [dataset] (Struckhoff et al., 2017). The XRF field sample measurements were compared to either U.S. Environmental Protection Agency Ecological Soil Screening Levels (Eco-SSL) (2005a, 2005b, 2005c, 2005d, 2007a, 2007b, 2007c, 2007d, 2008) or marine sediment Threshold Effects Levels (TEL) (MacDonald et al., 1996). Concentrations of elements below Eco-SSLs are considered protective of terrestrial plants, birds, and invertebrates that are likely to come into contact with or to consume soil contaminated with that element. Mammalian Eco-SSLs were not considered because there is no mammalian wildlife on the atoll. Threshold Effects Levels applied to marine sediments are considered protective of marine species generally and not specifically by taxa. For simplicity, all terrestrial soil samples were compared to the lowest established Eco-SSL for a given element regardless of taxonomic group (Table 1). Soil moisture values between 15 and 25% can reduce measured sample concentrations from 70 to 80% relative to lab-confirmed concentrations (U.S. Environmental Protection Agency, 2015). Given that most measurements were made on moist samples, XRF values reported above TELs or SSLs are conservative estimates of threshold exceedance. Geospatial data from this effort include the scanned and georectified 1944 map of Navy facilities, digitized and attributed features from the map, elemental analysis sample locations, and the locations at which the SPMDs were deployed. These data are publicly available [dataset] (Struckhoff et al., 2017). The data for samples analyzed using XRF include concentrations for select metals and descriptions of the location from which samples were taken. Summaries of the 1944 map and the XRF results are presented in the sections that follow. The 1944 map of naval facilities yielded 208 buildings and other features related to U.S. Navy operations on the atoll during WWII. Seventeen road segments were also mapped. Most of the activity on the atoll was located on the conjoined Cooper and Menge Islands. A powerhouse and munitions storage magazines were located at the west end of Menge. The north part of Menge was dominated by living quarters, but also included a powerhouse and photo processing laboratory. The area surrounding the current base camp for TNC included many machine shops, maintenance garages, and marine operations. The runway was flanked on the lagoon side by numerous aircraft and vehicle fuel

Engineers, 1987). This effort initiated an additional survey that documented fuel and oil in tanks and barrels, transformers, asbestos, and collapsing buildings of environmental concern; an analysis of container materials and surrounding soils documented very high levels (parts per thousand) of total petroleum hydrocarbons, toluene, and lead (U.S. Army Corps of Engineers (by R.M. Towill Corporation), 1993). From this information, a remediation plan was developed (U.S. Army Corps of Engineers (by R.M. Towill Corporation), 1993). Selective decontamination of the worst areas occurred in 1997 and 1998; however, further remedial action was recommended at numerous sites (U.S. Army Corps of Engineers (by Environmental Chemical Corporation [ECC]), 1998). Later, an extensive visual survey of known and potential sources of contamination and limited sampling and analysis was conducted by USFWS and others; lead, cadmium, copper, zinc, arsenic, PCBs, and polycyclic aromatic hydrocarbons (PAHs) were measured at levels exceeding ecological threshold criteria in soils or sediments from some islets (U.S. Fish and Wildlife Service, 2000). As of 2017, assessment, clean-up, and reclamation of the atoll remains incomplete. The goal of this project was to survey concentrations of elements in soils and sediments of Palmyra Atoll. Contaminant data have been made publically available in machine readable format to support wildlife and habitat management and research [dataset] (Struckhoff et al., 2017). The data will inform the identification of remediation locations, establishment of habitat management goals and objectives, prioritization and design of restoration projects, and development of monitoring and research plans. Concentrations of elements were measured using a portable, hand-held X-Ray fluorescence spectrometer (XRF) in 2008 and 2010. During 2008 sampling, passive samplers (semi-permeable membrane devices, SPMD) were deployed to detect bio-available persistent organic contaminants in terrestrial and nearshore aquatic environments, and element concentrations in sediment from the off-shore reef system were measured using XRF. Locations for the deployment of SPMDs and off-shore sediment collection are included with the publically available dataset [dataset] (Struckhoff et al., 2017), but they are not addressed further in this report. The present report focuses on the terrestrial and near-shore XRF data. Suspected areas of contamination were located from a georeferenced digital copy of a 1944 atoll map and an inventory of naval facilities that was produced by the present authors. Digital photos of the map were mosaicked and georeferenced against a 2007 orthorectified satellite image. Identifiable features and structures on the map were digitized and attributed with information from the map legend for use in the field on a handheld global position system (GPS) receiver. These spatial data served as the basis for locating potential sources of contamination while on the atoll in October 2008. It turned out that the oblique angles of the original map photos yielded a poorly georeferenced low-quality image, so this process was repeated using scanned images of the same map, and map features were updated based on the second georeferenced image and field-collected data. Soil and sediment sampling focused on areas that had facilities most likely to be contaminant sources based on their function as indicated in the legend for the 1944 Navy map. Areas such as barracks, offices, and recreation facilities were generally ignored. Sampling also included areas not identified on the map but that were known to be, or suspected of being, contaminant sources from previous sampling by others, or because the activities or materials in the area were known sources of contamination. These included miscellaneous debris (e.g., electrical transformers) and the nine OIA exclusions areas. It also included buildings having unidentified purposes, as well as known but unmapped facilities where chemicals may have been used or industrial processes performed (i.e., radio facilities and a hospital). Reference samples, far from suspected areas of contamination, were opportunistically collected on many islets and at near-shore locations to provide a measure of natural background levels of the elements. Sampled areas were broadly categorized in the dataset as either “reference”, “exclusion”, “buildings” (on 1944 map or discovered while sampling), 99

Marine Pollution Bulletin 128 (2018) 97–105

M.A. Struckhoff et al.

Table 1 Marine sediment Threshold Effects Levels (TEL) and terrestrial Eco-SSLs (mg/kg dry weight) for elements with an Eco-SSL that were most frequently detected by hand-held X-Ray fluorescence spectrometer at Palmyra Atoll, 2008. NA indicates that a toxicity threshold has not been established. Element

Antimony Arsenic Chromium Cobalt Copper Lead Manganese Nickel Zinc a

Symbol

Sb As Cr Co Cu Pb Mn Ni Zn

TELa

NA 7.24 52.30 NA 118.7 30.24 NA 15.90 124.00

Eco-SSL

Eco-SSL reference

Plant

Soil invertebrate

Avian

NA 18 NA 13 70 120 220 38 160

78 NA NA NA 80 1700 450 280 120

NA 43 26 120 28 11 4300 210 46

U.S. U.S. U.S. U.S. U.S. U.S. U.S. U.S. U.S.

Environmental Environmental Environmental Environmental Environmental Environmental Environmental Environmental Environmental

Protection Protection Protection Protection Protection Protection Protection Protection Protection

Agency, Agency, Agency, Agency, Agency, Agency, Agency, Agency, Agency,

2005a 2005b 2008 2005c 2007a 2005d 2007b 2007c 2007d

(Macdonald et al., 1996).

have Eco-SSLs; they are included in the dataset, but the relative hazards of these three elements are not addressed in this report. Iron, is generally considered “nontoxic”, and neither TELs nor Eco-SSLs are available; however, iron results are discussed because marine systems are typically iron-limited and can be variously stressed through enrichment from abundant iron sources such as those found in military dumps (Wells et al., 1995). In terrestrial settings and freshwater wetlands, excessive iron can differentially affect plant and microbial species (Guasch et al., 2012; Saaltink et al., 2017; Snowden and Wheeler, 1993; Wheeler et al., 1985) with documented adverse effects on restoration efforts (Lucassen et al., 2000). The 12 other elements detectable with the XRF are not reported because they lack both an Eco-SSL and a TEL (calcium, chlorine, potassium, rubidium, scandium, strontium, and tin) or were always below the limit of detection (mercury, nickel, selenium, silver, and vanadium). Samples that were believed to represent uncontaminated, natural concentrations of elements were collected at 21 locations for marine sediment and 7 locations for soil; these are referred to as “reference” samples herein. Seven reference marine sediment samples had Cr concentrations that exceeded the TEL of 52 mg/kg (Table 2). Antimony was detected in two samples at concentrations in excess of the terrestrial Eco-SSL for soil invertebrates (78 mg/kg); no marine sediment TEL exists for Sb. Iron concentrations in reference marine sediment samples ranged from 106 mg/kg to 146 mg/kg. No other elements of concern were above the limits of detection in marine sediment. Reference samples taken from terrestrial locations included Cr concentrations 3 to 4 times the Eco-SSL of 26 mg/kg (Table 2). One sample (Paradise 02) had concentrations of Pb and Zn of 25 mg/kg and 88 mg/kg, approximately two times greater than the relevant Eco-SSLs (11 mg/kg and 46 mg/kg, respectively); however, this may have been a poor reference sample because it was taken near a utility pole. The only other reference samples to exceed the Zn Eco-SSL were from Sand Island. Iron concentrations in reference soils ranged from 139 mg/kg to 236 mg/kg. All 20 samples from the OIA “exclusion areas” were intertidal sandy sediment except for Strawn 04 (Table 3). At least one sample from each of the four sampled exclusion areas had at least one element concentration that exceeded a TEL for marine sediment. The mean concentrations of Cr in 15 exclusion area samples exceeded the TEL (52.3 mg/kg), but their range (56 mg/kg to 117 mg/kg) was similar to that of near-shore reference samples (39 mg/kg to 107 mg/kg). Other than Cr, no elements were detected by XRF near Kaula Island, nor was there physical evidence of potential contaminant sources. One sample from the edge of the exclusion area at Aviation Island also had elevated Zn concentrations above the TEL; scattered metal debris was present. The exclusion areas at the west end of Strawn and at Quail Islands both included abundant metal debris from dumps that had apparently been burned during the Naval Air Station period. Samples from these areas

Fig. 3. Satellite image of Menge Island at Palmyra Atoll National Wildlife Refuge, with overlay of 1944 map of U.S. Navy structures and digitized locations of structures (white).

storage facilities and associated pump houses and pipelines; the other side was flanked by more fuel storage pits (at the west end), and airfield support buildings, including paint and maintenance shops. A lumber yard and aggregate plant were located east-northeast of the fuel pits. Between the airstrip and a road running parallel to the north shore of the atoll were numerous buildings dedicated to industrial activities; on the shore side of the road were barracks, hospitals, and communications buildings. Other buildings not on the 1944 map include a hospital on Engineer Island, radio communications facilities on Paradise Island, and scattered defensive batteries of various sizes. An example of the georeferenced 1944 map and digitized features is provided in Fig. 3. In 2008, concentrations of elements were measured in samples from 155 unique locations, yielding 166 records (7 samples include dual wet and dry measurements; 4 records for Cooper 09 indicate the widelyspaced sources for a single composite sample; and the location of batteries at the north end of Barren Island was marked using GPS, but no XRF measurement was taken). In 2010, an additional 41 samples of near-shore marine sediment were collected and analyzed using XRF. Error estimates generated by the XRF instrument indicate a typical trend in analytical measurements: error values were less than 20% for most readings and less than 5% for the highest concentrations measured, while error values associated with concentrations approaching the limit of detection were greater than 20% (data not shown). Concentrations are reported for antimony (Sb), arsenic (As), chromium (Cr), cobalt (Co), copper (Cu), lead (Pb), manganese (Mn), and zinc (Zn). Titanium, sulfur and phosphorus were detected but do not

100

Marine Pollution Bulletin 128 (2018) 97–105

M.A. Struckhoff et al.

Table 2 Concentrations of selected elements detected in background “reference” samples using a hand-held X-Ray fluorescence spectrometer at Palmyra Atoll, 2008. — Indicates below limit of detection in at least one of three XRF readings for the sample. Sample

Matrix

Concentration (mg/kg wet weight) As

Cooper 16 Holei & Bird Forereef End Holei & Bird Forereef Start Home & Paradise Forereef End Home & Paradise Forereef Start Menge 07 North Barren North Barren Buoy Paradise 03 Penguin Spit Forereef Penguin Spit Inner End Penguin Spit Inner Start Penguin Spit Middle Start Sand 03 Sand 05 Sand 06 Strawn 01 Strawn Forereef End Strawn Forereef Start Tortugonias Bouy Start Western Terrace Bouy Median

Sand Coral/calcified algae Coral/calcified algae Sand/calcified algae/coral rubble Coral/calcified algae Sand Coral Sand Sand Sand/calcified algae Fine sand Sediment Sand/calcified algae Sand Sand Sand Sand Sand/calcified algae Sand/calcified algae Sand Coral/calcified algae

Menge 08 Paradise 02 Paradise 04 Paradise 05 Sand 01 Sand 02 Strawn 02 Median

Soil Soil Swamp sediment Soil Soil Soil Soil

Co

Cr

Cu

Fe

Mn

Pb

Sb

Zn

Marine sediment – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

93 – – – – 96 39 – 107 – – – – 93 83 88 – – – – 75 91

– – – – – – – – – – – – – – – – – – – – – –

140 106 121 117 122 141 132 123 146 114 127 124 120 141 138 142 129 125 115 123 108 124

– – – – – – – – – – – – – – – – – – – – – –

– – – – – – – – – – – – – – – – – – – – – –

– – – – – – – – – 115 – – – – – – – – 95 – – 105

– – – – – – – – – – – – – – – – – – – – – –

Soil – – – – – – – –

131 82 97 171 77 70 67 82

– – – – – – – –

148 148 161 152 236 172 139 152

– – – – – – – –

– 25 – – – – – 25

– – – – – – – –

– 88 45 13 55 59 – 55

– – – – – – – –

Bold indicates calculated median for indicated medium (sediment or soil).

those at reference sites. All Quail Island Fe concentrations were greater than reference sites, with one marine sediment sample measuring 31,029 mg/kg. Concentrations of Fe in marine sediment were generally greater in areas with metal debris evident.

had concentrations above TELs for Cu, Pb, and Zn. Cobalt was detected at 449 mg/kg from one sample from Quail Island. Iron concentrations at locations sampled on Kaula and Aviation were similar to intertidal reference sites. Strawn samples had iron concentrations 3 to 4 times

Table 3 Concentrations of selected elements detected in intertidal samples from exclusion areas using a hand-held X-Ray fluorescence spectrometer at Palmyra Atoll, 2008. — Indicates below limit of detection in at least one of three XRF readings for the sampleAll samples were intertidal sand except Strawn 4, which was a blue powder found in a small broken bottle. Sample

Aviation 01 Aviation 02 Aviation 03 Kaula 01 Kaula 02 Kaula 03 Quail 01 Quail 02 Quail 03 Quail 04 Quail 05 Quail 06 Quail 07 Quail 08 Quail 09 Quail 09 Quail 09 Strawn 03 Strawn 04 Strawn 05

Comments

Among concrete slabs and metal Among miscellaneous debris Near 2.5 ft. × 12 ft. tank; edge of area No evidence of reason for exclusion Near-shore perimeter of exclusion zone Off-shore perimeter of exclusion zone In center of impounded dump area At head of flow path from waste area At mouth of drain from waste area On flat adjacent to waste area At mouth of flow path from waste area At same location as SPMD05 At same location as SPMD06 At same location as SPMD07 Mouth of drain from waste area; 0–1 in depth Mouth of drain from waste area; 1–4 in depth Mouth of drain from waste area; 4–8 in depth Dump area; much metallic waste evident Blue chemical powder from 100 mL jar Dump area; metallic waste, burned batteries

Concentration (mg/kg wet weight) As

Co

Cu

Cr

Fe

Mn

Pb

Sb

Zn

– – – – – – – – – – – – – – – – – – – –

– – – – – – – – – – – 449 – – – – – – – –

– – – – – – – – – – – 96 59 64 – – – 39 – –

78 78 60 56 – 80 104 117 73 85 73 106 71 65 – – – 116 – 116

140 363 138 140 143 143 846 549 358 172 301 31,029 2376 2423 239 420 216 400 7667 503

– – – – – – – – – – – – – – – – – – 132 –

– – – – – – – 17 – – – 40 37 76 – – – 25 101 32

– – – – – – – – – – – – – – – – – – 100 –

– – 89 – – – 30 54 – – 26 1050 167 204 22 31 45 – 39 –

101

Marine Pollution Bulletin 128 (2018) 97–105

M.A. Struckhoff et al.

Table 4 Concentrations of selected elements detected on or adjacent to refuge islands, Palmyra Atoll, 2008. — Indicates below limit of detection in at least one of three XRF readings for the sample. Sample

Potential contaminant source

Matrix

Concentration (mg/kg wet weight) As

Co

Cr

Cu

Fe

Mn

Pb

Sb

Zn

Soil Cooper 02 Engineer 01 Engineer 02 Engineer 03 Engineer 04 Paradise 01 Strawn 07 Strawn 08 Strawn 10 Strawn 11 Strawn 12 Strawn 13 Strawn 14

Diesel oil filter house and equipment Power management; medical supplies Oil/fuel equipment/spillage outside hospital Oil/fuel equipment/spillage outside hospital Hospital runoff in low area east of hospital Radar/radio equipment Metals and ordnance magazine Metals and ordnance from magazine Ordnance chemicals, metals from building Creosote, metals, transformer chemicals near utility pole Building decay and munitions in defensive battery Metals and ordnance chemicals; only wires evident Metals and ordnance chemicals; near NW drain

Soil Sludge Soil Soil Soil Soil Coral Coral Soil Soil Soil Soil Soil

– 12 – – – – – – – – – – –

– 352 – – – – – – – – – – –

136 153 – 117 138 76 73 – 90 87 80 99 98

– 37 – – 16 – – – 30 – – – –

708 15,402 177 202 182 648 545 121 2170 140 688 156 416

– – – – – – – – – – – – –

– 96 – 10 – – – – 68 – – – 12

– – – – – – – – – – – – –

87 1198 24 21 15 77 – – 168 – 141 – 36

Marine sediment Cooper 01 Aviation 04 Whippoorwill 01 Whippoorwill 02 Whippoorwill 03 Whippoorwill 04

Rusting mooring dolphin, oil/fuel Ordnance chemicals, metals from structure Unidentified debris 4 m × 3 m Pipe with copper coil and other debris 2, 15″ × 20′ iron pipes; debris over 30 m × 5 m 4 pipes over a 2 m × 2 m area

Sand Sand Sand Sand Sand Sand

– – – – – –

– – – – – –

64 45 – – 62 84

– – – – – –

669 142 141 796 842 146

– – – – – –

– – – – – –

– – – – – –

– – – – 82 –

In nearly all samples where Cu was detected, concentrations exceeded the Eco-SSL for avian wildlife (28 mg/kg). Six of these samples were nearly an order of magnitude or more above the Eco-SSL. A sample of soil collected near an unprotected stack of pipes and valves had a wet-soil Cu concentration of 1155 mg/kg. Manganese was detected in 8 samples. Of these, 4 exceeded the EcoSSL for plants (220 mg/kg); 2 were from the floor of the former power station at the west end of Menge Island, with one sample having a measured concentration of 1310 mg/kg. The other 2 samples above the manganese Eco-SSL were collected adjacent to deteriorating electrical transformers. Other sites with detectable Mn were associated with vehicle maintenance, storage, or with radio communications. Lead concentrations exceeded the Eco-SSL for avian wildlife (11 mg/kg) in all 46 samples where it was detected. Lead concentrations in 21 measurements were 1 to 2 orders of magnitude above the Eco-SSL; 11 measurements exceeded the Pb Eco-SSL by more than 2 orders of magnitude. Zinc was the most frequently detected element on Cooper and Menge Islands, occurring in all but 11 samples. Zinc was measured above the Eco-SSL for avian wildlife (46 mg/kg) in 51 unique samples. Of these, 21 were 1 to 2 orders of magnitude above the Eco-SSL and 5 were more than 2 orders of magnitude greater than the Eco-SSL. Many of the highest Zn concentrations were measured in soils at sites associated with power generation or transmission, or with radio and radar communications. There was a broad range of Fe concentrations among the soil samples from Cooper and Menge Islands. A third of the samples were between 1 and 3 orders of magnitude greater than the reference soil concentration. A soil sample from Cooper Island taken near a deteriorating transformer had a Fe concentration of 118,315 mg/kg soil and one from Menge Island near a former power generator had Fe at 305,953 mg/kg. The locations of the highest concentrations of specific elements are best explored using the geospatial data on which this report is based [dataset] (Struckhoff et al., 2017). To facilitate remediation planning on Cooper and Menge, a map was generated that displays sample locations based on the number of metals detected above their respective Eco-SSL (Fig. 4). A second map identifies the most hazardous sites by presenting sample sites by the number of metals detected at concentrations more than an order of magnitude greater than their Eco-

Sampling from intertidal areas and islands owned by USFWS included two samples, Cooper 01 and Cooper 02, on or adjacent to Cooper Island but within the USFWS ownership boundary. The most frequently detected elements on refuge islands were Cr, Fe, and Zn (Table 4). Chromium concentrations ranged from 45 mg/kg to 153 mg/ kg, up to six times greater than the Eco-SSL for soil invertebrates (26 mg/kg), but were generally similar to reference sites. Zinc concentrations exceeded the Eco-SSL for avian wildlife (46 mg/kg) in 5 of the 9 terrestrial locations where it was detected. In the few instances where they were detected on refuge islands, Pb, Co, and Cu tended to exceed their respective Eco-SSLs. Iron concentrations in 7 soil samples exceeded reference concentrations, ranging from 416 mg/kg to 15,402 mg/kg. The highest value was from a sludge sample collected from a drainage trench within the hospital on Engineer Island; concentrations of Co, Cu, Pb, and Zn from the same sample all exceeded their Eco-SSLs. The samples of sandy sediments from intertidal areas near Aviation and Whippoorwill Islands had concentrations of elements similar to background levels found in reference areas, except for one sample with Zn above the TEL. Two additional samples from Whippoorwill had Fe concentrations of 796 mg/kg and 842 mg/kg. The sediment sample from near the mooring dolphins at the west end of the runway (Cooper 01) had elevated Fe (669 mg/kg). For the 41 marine sediments collected and XRF-analyzed in 2010, none of the elements were above background concentrations, except for Fe which was detected in two sediment samples (447 mg/kg and 637 mg/kg) collected near the docks. From areas owned and managed by TNC on or adjacent to Cooper and Menge Islands, seventy-three unique samples were collected. During WWII, these islands were sites for numerous facilities for powergeneration, radio and radar communication, fuel storage, and other industrial activities. Antimony was not detected on Cooper or Menge Islands. Nine soil samples had Cr concentrations an order of magnitude greater that the Eco-SSL for avian wildlife (26 mg/kg); otherwise concentrations were comparable to reference values. Arsenic levels exceeded the Eco-SSL for plants (18 mg/kg) in all 11 samples where it was detected; concentrations ranged from 30 mg/kg to 560 mg/kg. Cobalt exceeded the EcoSSL for plants (13 mg/kg) by more than an order of magnitude in all 6 samples where it was detected. 102

Marine Pollution Bulletin 128 (2018) 97–105

M.A. Struckhoff et al.

Fig. 4. Map of sample locations on Cooper Island and Menge Island showing the number of elements detected at concentrations greater their Eco-SSLs at Palmyra Atoll, 2008.

Fig. 5. Map of sample locations on Cooper Island and Menge Island showing the number of elements detected at concentrations more than an order of magnitude greater than their EcoSSL at Palmyra Atoll, 2008.

SSLs (Fig. 5). Many samples that switch from “1” in the former map to “0” in the latter map had chromium above the Eco-SSL but comparable to reference values. Wars leave behind a legacy of environmental disturbance and contamination (Lawrence et al., 2015). The data collected in this survey add to existing information indicating that elemental contamination remains at levels of ecological concern at several locations at Palmyra Atoll, despite initial remediation actions (U.S. Army Corps of Engineers (by Environmental Chemical Corporation [ECC]), 1998; U.S. Fish and Wildlife Service, 2000, 2001). The specific locations of contaminants documented here are consistent with the function of the former naval facilities. Samples from sites associated with power generation and transmission consistently exhibited elevated concentrations of elements, as did vehicle repair shops and other locations where industrial

activities were performed. Compared to the TNC islands and the exclusion areas, relatively little contamination was detected on USFWS refuge islands; this may reflect either the limited activity by the Navy on these sites, differences in thoroughness of remediation, or unintentional bias introduced by the targeted sampling employed in this survey. The elements most frequently occurring above terrestrial Eco-SSLs, or sediment TELs, were arsenic, cobalt, copper, chromium, lead, and zinc. Numerous laboratory studies have documented adverse effects of lead and zinc on the physiological function of commonly tested agricultural plant species (Andersson, 1988; Das et al., 1997; Fargašová, 2001; Kabata-Pendias and Pendias, 1992; Påhlsson, 1989). Only a few studies have documented deleterious physiological effects of these metals on native flora, usually expressed as reduced growth and 103

Marine Pollution Bulletin 128 (2018) 97–105

M.A. Struckhoff et al.

2017) is a resource to guide remediation and restoration of habitat on Palmyra Atoll. The data also provide a foundation for critical additional sampling necessary to fully characterize the spatial extent and magnitude of surface and subsurface contamination on the Atoll. The acquired information on the historic naval operations can facilitate additional screening for locations suspected to be contaminated, but which were not sampled during previous scoping efforts or decontaminated during prior remediation efforts (U.S. Army Corps of Engineers (by Environmental Chemical Corporation [ECC]), 1998; U.S. Army Corps of Engineers (by R.M. Towill Corporation), 1993; U.S. Fish and Wildlife Service, 2000, 2001). As other elements on Palmyra tended to co-occur with lead, zinc, and/or copper, these elements would likely be useful targets when screening for contamination generally, or during remediation planning and implementation. The utility of the data could be significantly improved if augmented with data from reports from previous scoping and remediation efforts, as well as other historic documents. This could include integrating existing geospatial data (if it exists) or mining documents to generate new data that can be integrated into the dataset described here [dataset] (Struckhoff et al., 2017).

biomass (Beyer et al., 2013; Stroh et al., 2015). Negative effects of elevated lead and zinc have been shown on vegetation communities in England (Clark and Clark, 1981; Thompson and Proctor, 1983), Montana (LeJeune et al., 1996), Pennsylvania (Beyer et al., 2011), and Missouri (Struckhoff et al., 2013). There is little research to indicate that tropical plants are adversely affected by the metals above; however excessive Fe as documented in many terrestrial locations at Palmyra Atoll can reduce nodulation and plant growth in tropical legumes (Paudyal et al., 2007). The extent to which Palmyra's terrestrial, avian, and aquatic animals may be adversely affected by the elements we found on the atoll is generally unknown. Only two reports from Palmyra Atoll directly consider the impact on biota of the elements we measured. McFadden et al. (2014) reported low metal concentrations in green sea turtles that forage around Palmyra. Work et al. (2008) described that increased iron concentrations indirectly affected coral health and distribution by stimulating growth of a competing corallimorph. Although the iron likely came from a ship wrecked in a lagoon reef (since removed), iron and other elements measured in the terrestrial environment could migrate off-shore, similar to that occurring on Saipan (Denton et al., 2016). The extreme rainfall that Palmyra experiences may contribute to mobilization of contaminants to the surrounding waters. Despite legacy military contamination and disturbance, populations of aquatic and avian species appear to be healthy on Palmyra Atoll (Braun et al., 2008; Mundy et al., 2010; Vargas-Angel and Wheeler, 2008). However, as has been done for similar Pacific Ocean atolls, assessment of baseline tissue concentrations of the contaminants found on and around the islands is warranted (Burger et al., 2001; Lobel and Lobel, 2008; Miao et al., 2001; Morgan et al., 2010). Because most of the XRF measurements in this study were made on freshly collected wet samples, concentrations reported here are likely biased low due to moisture-induced impedance of spectral signatures when using XRF technology (U.S. Environmental Protection Agency, 2015). For the seven samples in this study for which direct comparison can be made, measured concentrations of individual elements in dry readings were on average twice the concentration found in wet readings. This limitation was a known bias accepted during this study given the logistics of sampling at the remote atoll. As such, values reported here probably underestimate the hazard posed by elements on the atoll. This bias may also have been a contributing factor in some elements (e.g., mercury, nickel, selenium, silver, and vanadium) being consistently below the limits of detection; these elements may yet be present at concentrations that pose risks to the flora and fauna of Palmyra. Future screening protocols would benefit from XRF readings on dry substrates whenever possible, consistent with EPA sampling protocols (U.S. Environmental Protection Agency, 2015). Where not possible, uniform wet measurement protocols can be used with correction factors to estimate dry concentrations, if such correction factors are available. Collection and laboratory analysis of samples by inductively coupled plasma-mass spectrometry (ICP-MS) is another alternative that may be appropriate depending on data quality objectives, and would be particularly useful for determinations of the bioavailable fractions of metals detected by XRF. We compared marine sediment concentrations to Threshold Effects Levels (Macdonald et al., 1996). Further analyses of marine sediment samples could be evaluated against additional marine sediment toxicity reference values provided in the National Oceanic and Atmospheric Administration Screening Quick Reference Tables (SQuiRTs), which aggregate toxicity screening levels (including TELs) from numerous marine sediment studies (Buchman, 2008). This latter reference provides a broad range of potential toxicity values, and may prove particularly informative for additional screening and remediation of the 8 near-shore exclusion areas. Human health screening criteria were not used in this report because such considerations are beyond the scope of this report. The dataset referenced in this report [dataset] (Struckhoff et al.,

Acknowledgements Our gratitude is extended to Mandy Annis for 2010 XRF data collection and to Thomas Cecere for map development supporting early mission planning. Thanks also to Gareth Williams and Ingrid Knapp for SCUBA-based sediment collection. Additional thanks to acting Palmyra Refuge Manager Andrew Guda, Station Manager Ned Brown, and the rest of the staff (Kydd, Bobby, Martha, Jeff, Robert, and Linda) that made our research possible and our time on Palmyra such a great experience. This report is contribution number PARC-0137 to the Palmyra Atoll Research Consortium, to which we also extend thanks. This project was funded by the U.S. Geological Survey Environmental Health Mission Area Contaminant Biology Program and by in-kind support from U.S. Fish and Wildlife Service and The Nature Conservancy. Any use of trade, firm, or product names is for descriptive purposes only and does not imply endorsement by the U.S. Government. Appendix A. Supplementary data Supplementary data can be accessed on ScienceBase at https://doi. org/10.5066/F74F1P00. References Andersson, M., 1988. Toxicity and tolerance of aluminium in vascular plants. Water Air Soil Pollut. 39, 439–462. Batianoff, G.N., Naylor, G.C., Olds, J.A., Fechner, N.A., John Neldner, V., 2010. Climate and vegetation changes at Coringa-Herald National Nature Reserve, Coral Sea Islands, Australia. Pac. Sci. 64, 73–92. Beyer, W., Krafft, C., Klassen, S., Green, C., Chaney, R., 2011. Relating injury to the forest ecosystem near Palmerton, PA, to zinc contamination from smelting. Arch. Environ. Contam. Toxicol. 61, 376–388. Beyer, W.N., Green, C.E., Beyer, M., Chaney, R.L., 2013. Phytotoxicity of zinc and manganese to seedlings grown in soil contaminated by zinc smelting. Environ. Pollut. 179, 167–176. Braun, C.L., Smith, J.E., Vroom, P.S., 2008. Examination of Algal Diversity and Benthic Community Structure at Palmyra Atoll, U.S. Line Islands, 11th International Coral Reef Symposium. ReefBase, Ft. Lauderdale, Florida, pp. 865–869. Buchman, M.F., 2008. NOAA Screening Quick Reference Tables. National Oceanic and Atmmospheric Administration, Office of Response and Restoration Division, Seattle, WA, pp. 34. Burger, J., Shukla, T., Dixon, C., Shukla, S., McMahon, M.J., Ramos, R., Gochfeld, M., 2001. Metals in feathers of sooty tern, white tern, gray-backed tern, and Brown Noddy from islands in the North Pacific. Environ. Monit. Assess. 71, 71–89. Clark, R.K., Clark, S.C., 1981. Floristic diversity in relation to soil characteristics in a lead mining complex in the Pennines, England. New Phytol. 87, 799–815. Das, P., Samantaray, S., Rout, G.R., 1997. Studies on cadmium toxicity in plants: a review. Environ. Pollut. 98, 29–36. Dell, J., 2014. Harnessing geospatial data to enhance ERW clearance in Pacific Islands. J. ERW Mine Action 18, 14–17. Denton, G.R.W., Emborski, C.A., Hachero, A.A.B., Masga, R.S., Starmer, J.A., 2016.

104

Marine Pollution Bulletin 128 (2018) 97–105

M.A. Struckhoff et al.

Sheet 2012–3002, 4 p. pp. 4. The Nature Conservancy, 2017. Palmyra Atoll: a center for scientific study. http://www. nature.org/ourinitiatives/regions/northamerica/unitedstates/hawaii/palmyraatoll/, Accessed date: 20 January 2017. Thompson, J., Proctor, J., 1983. Vegetation and soil factors on a heavy metal mine spoil heap. New Phytol. 94, 297–308. U.S. Army Corps of Engineers, 1987. Inventory project report: Palmyra Naval Air Station (Palmyra Atoll); appendix A. In: RM Towill Corporation, 1993. Palmyra Atoll Remediation Plan Vol. 2. U.S. Army Corps of Engineers, Defense Environmental Restporation Program (DERP) for Formerly Used Sites (FUDS), pp. 11 (appendices). U.S. Army Corps of Engineers (by Environmental Chemical Corporation [ECC]), 1998. Draft Closure Report: Remedial Action Palmyra Atoll. Palmyra USA, US Army Corp of Engineers, Honolulu, pp. 406. U.S. Army Corps of Engineers (by R.M. Towill Corporation), 1993. Palmyra Atoll Remediation Plan: Vol I (Report) and Vol II (Appendices). US Army Corps of Engineers, Pacific Division, US Army Corps of Engineers, Honolulu. U.S. Environmental Protection Agency, 2005a. Ecological Soil Screening Levels for Antimony: Interim Final. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington D.C., pp. 29. U.S. Environmental Protection Agency, 2005b. Ecological Soil Screening Levels for Arsenic: Interim Final. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington D.C., pp. 128. U.S. Environmental Protection Agency, 2005c. Ecological Soil Screening Levels for Cobalt: Interim Final. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington D.C., pp. 75. U.S. Environmental Protection Agency, 2005d. Ecological Soil Screening Levels for Lead: Interim Final. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington D.C., pp. 211. U.S. Environmental Protection Agency, 2007a. Ecological Soil Screening Levels for Copper: Interim Final. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington D.C., pp. 313. U.S. Environmental Protection Agency, 2007b. Ecological Soil Screening Levels for Manganese: Interim Final. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington D.C., pp. 311. U.S. Environmental Protection Agency, 2007c. Ecological Soil Screening Levels for Nickel: Interim Final. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington D.C., pp. 133. U.S. Environmental Protection Agency, 2007d. Ecological Soil Screening Levels for Zinc: Interim Final. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington D.C., pp. 781. U.S. Environmental Protection Agency, 2008. Ecological Soil Screening Levels for Chromium: Interim Final. U.S. Environmental Protection Agency, Office of Solid Waste and Emergency Response, Washington D.C., pp. 106. U.S. Environmental Protection Agency, 2015. Field X-Ray Fluorescence Measurement. U.S. Environmental Protection Agency, Science and Ecosystem Services Division, F.S.B., Athens, Georgia, pp. 11. U.S. Fish and Wildlife Service, 2000. Level I Pre-Acquisition Contamination Survey for the Proposed Palmyra Atoll National Wildlife Refuge. US Fish and Wildlife Service, US Fish and Wildlife Service, P.R., Honolulu, pp. 16. U.S. Fish and Wildlife Service, 2001. Level 3 Pre-Acquisition Contamination Survey for the Proposed Palmyra Atoll National Wildlife Refuge. US Fish and Wildlife Service, US Fish and Wildlife Service, P.R., Honolulu, pp. 39. Vargas-Angel, B., Wheeler, B.D., 2008. Coral Health and Disease Assessment in the U.S. Pacific Territories and Affiliated States, 11th International Coral Reef Symposium. ReefBase, Ft. Lauderdale, Florida, pp. 175–179. Wells, M.L., Price, N.M., Bruland, K.W., 1995. Iron chemistry in seawater and its relationship to phytoplankton: a workshop report. Mar. Chem. 48, 157–182. Wheeler, B.D., Al-Farraj, M.M., Cook, R.E.D., 1985. Ion toxicity to plants in base-rich wetlands: comparative effects on the distribution and growth of epilobium hirsutum (l.) and Juncus subnodulosus (Schrank). New Phytol. 100, 653–669. Work, T.M., Aeby, G.S., Maragos, J.E., 2008. Phase shift from a coral to a corallimorphdominated reef associated with a shipwreck on Palmyra Atoll. PLoS ONE 3, e2989.

Impact of WWII dumpsites on Saipan (CNMI): heavy metal status of soils and sediments. Environ. Sci. Pollut. Res. 23, 11339–11348. Fargašová, A., 2001. Phytotoxic effects of Cd, Zn, Pb, Cu and Fe on Sinapis alba L. seedlings and their accumulation in roots and shoots. Biol. Plant. 44, 471–473. Guasch, H., Acosta, X.G., Urrea, G., Bañeras, L., 2012. Changes in the microbial communities along the environmental gradient created by a small Fe spring. Freshwater. Science 31, 599–609. Handler, A.T., Gruner, D.S., Haines, W.P., Lange, M.W., Kaneshiro, K.Y., 2007. Arthropod surveys on Palmyra Atoll, Line Islands, and insights into the decline of the native tree Pisonia grandis (Nyctaginaceae). Pac. Sci. 61, 485–502. Kabata-Pendias, A., Pendias, H., 1992. Trace Elements in Soils and Plants, 2nd ed. CRC Press, Boca Raton, FL. Lawrence, M.J., Stemburger, H.L.J., Zolderdo, A.J., Structhers, D.P., Cooke, S.J., 2015. The effects of modern war and military activities on biodiversity and the environment. Environ. Rev. 23, 443–460. LeJeune, K., Galbraith, H., Lipton, J., Kapustka, L.A., 1996. Effects of metals and arsenic on riparian communities in southwest Montana. Ecotoxicology 5, 297–312. Lobel, L.K., Lobel, P.S., 2008. Contaminants in Fishes From Johnston Atoll, 11th International Coral Reef Symposium. ReefBase, Ft. Lauderdale, Florida. Lucassen, E.C.H.E.T., Smolders, A.J.P., Roelofs, J.G.M., 2000. Increased groundwater levels cause iron toxicity in Glyceria fluitans (L.). Aquat. Bot. 66, 321–327. Macdonald, D.D., Carr, R.S., Calder, F.D., Long, E.R., Ingersoll, C.G., 1996. Development and evaluation of sediment quality guidelines for Florida coastal waters. Ecotoxicology 5, 253–278. McFadden, K.W., Gómez, A., Sterling, E.J., Naro-Maciel, E., 2014. Potential impacts of historical disturbance on green turtle health in the unique & protected marine ecosystem of Palmyra Atoll (Central Pacific). Mar. Pollut. Bull. 89, 160–167. Miao, X.-S., Balazs, G.H., Murakawa, S.K.K., Li, Q.X., 2001. Congener-specific profile and toxicity assessment of PCBs in green turtles (Chelonia mydas) from the Hawaiian Islands. Sci. Total Environ. 281, 247–253. Morgan, L., Chandler, W., Douse, E., Brooke, S., Guinotte, J., Myhre, S., 2010. Research Priorities for the Pacific Remote Islands Marine National Monument. Marine Conservation Biology Institute, Bellevue, Washington, pp. 39. Mundy, B.C., Wass, R., DeMartini, E., Greene, B., Zgliczynski, B., Schroeder, R.F., Musberger, C., 2010. Inshore Fishes of Howland Island, Baker Island, Jarvis Island, Palmyra Atoll, and Kingman Reef, Smithsoniann Institution, History. N.M.o.n, Washington, D.C., pp. 133. Nautilus Institute, 2017. Toxic bases in the Pacific. Nautilus Institutehttp://nautilus.org/ apsnet/toxic-bases-in-the-pacific/, Accessed date: 20 January 2017. Påhlsson, A.-M.B., 1989. Toxicity of heavy metals (Zn, Cu, Cd, Pb) to vascular plants. Water Air Soil Pollut. 47, 287–319. Paudyal, S.P., Aryal, R., Chauhan, S.V.S., Maheshwari, D.K., 2007. Effect of heavy metals on growth of rhizobium strains and symbiotic efficiency of two species of tropical legumes. Sci. World 5, 27–32. Saaltink, R.M., Dekker, S.C., Eppinga, M.B., Griffioen, J., Wassen, M.J., 2017. Plantspecific effects of iron-toxicity in wetlands. Plant Soil 1–14. Smith, J., 2014. ERW contamination in the Pacific Islands. J. ERW Mine Action 18, 10–13. Snowden, R.E.D., Wheeler, B.D., 1993. Iron toxicity to fen plant species. J. Ecol. 81, 35–46. Stevenson, C., Katz, L.S., Micheli, F., Block, B., Heiman, K.W., Perle, C., Weng, K., Dunbar, R., Witting, J., 2007. High apex predator biomass on remote Pacific islands. Coral Reefs 26, 47–51. Stroh, E.D., Struckhoff, M.A., Grabner, K.W., 2015. Soil Lead and Zinc Effects on Growth and Survival of Native Plant Species: Administrative Report to the Missouri Trustee Council. U.S. Geological Survey, Reston, VA, pp. 52. Struckhoff, M.A., Stroh, E.D., Grabner, K.W., 2013. Effects of mining-associated lead and zinc soil contamination on native floristic quality. J. Environ. Manag. 119, 20–28. Struckhoff, M.A., Orazio, C., Shaver, D., Papoulias, D., Annis, M.L., Tillitt, D.E., 2017. Geospatial Data of Elemental Contamination at Palmyra Atoll, 2008 and 2010, Columbia Environmental Research Center. U.S. Geological Survey, Reston, Virginia. https://doi.org/10.5066/F74F1P00. Suchanek, T., 2012. The Palmyra Atoll Research Consortium: U.S. Geological Survey Fact

105