The DOE has not monitored radioactivity on Amchitka Island or in biota around the site since the underground nuclear tests in 1965-1971, but they wished to close the site, turning it over to DOE’s long-term Legacy Management (for periodic monitoring of radionuclides) and ultimately transferring ownership to the State of Alaska or USFWS. However, the State of Alaska, led by the Alaska Department of Environmental Conservation (ADEC) did not want to assume responsibility for Amchitka, fearing possible monumental cleanup costs. Transfer would have to wait until they were sure the environment around Amchitka was not contaminated with radionuclides, and food was safe both for the Aleuts living on nearby islands and for commercial fishery. The US F&WS was also concerned about contamination in the food web overall. These groups were not convinced by DOE’s geophysical and human health risk models, nor were the other stakeholders, particularly the Aleuts who would gather sea food from the shore and offshore water [[33], [47]]. It was an issue of trust and needing to have current data on radionuclide levels in different nodes on the food chain. The dispute was resolved by the Governor of Alaska asking the Secretary of the DOE to have the Consortium for Risk Evaluation with Stakeholder Participation (CRESP) develop and implement a program to examine radionuclide levels and risks in the Amchitka Ecosystem [[51]]. CRESP is a multidisciplinary, multi-university group of scientists that has worked with DOE-EM on a variety of scientific and technical issues related to the legacy waste cleanup.

The CRESP approach was to write a Science Plan for Amchitka that included not only biological sampling and radionuclide analysis, but also geophysical characterization of the test sites. The biological Science Plan was aimed at providing the data needed to understand the risk from the viewpoint of consumers of marine organisms (Aleuts, other Alaskan citizens, commercial fisheries), and biota living in the environment on Amchitka Island and surrounding waters [[34], [35]]. From the beginning to the end of the Amchitka project, a broad range of stakeholders were involved (Figure 2). CRESP sought input from different agencies, as well as from Aleuts. Specifically, the Aleuts suggested adding some species to the protocol to reflect their interests and needs. Before the expedition (2004), we visited the Aleut villages to exchange ideas about the expedition with Tribal Aleut Elders and other Aleuts in their communities, as well as holding meetings with all the major stakeholder federal and state agencies [[52]]. Monitoring and long-term stewardship will continue on the island for the foreseeable future [[53]].

Figure 2.The initial Amchitka Science Plan developed and implemented to address both human health and ecological health issues and concerns (after Burgeret al.[[32],[33]]).APIA = the Aleutian/Pribilof Islands Association; US F&WS—the United States Fish and Wildlife Service.

The CRESP Science Plan laid out a marine and land-based plan to collect samples for radionuclide analysis from invertebrates, fish, and birds that represented different nodes on the food web, species of ecological concern, and species specifically requested by Aleuts, NOAA, and US F&WS [[31]]. We rented a 55-m fishing ship and retrofitted it for the month-long expedition; we had only limited (once/day or less) contact with anyone on shore. Without shore contact, we had limited options for rescue or evacuations if something happened to the land-based or diving crews. Because DOE detonated three underground nuclear test shots in different parts of Amchitka Island, samples were taken around each of these sites, and on Kiska Island (Figure 1) as a control. The sampling plan was designed and modified to include commercial species (algae, fish, and crabs), Aleut subsistence species (e.g., many species not commercially taken), and key species on the biota food web. At each sampling site, the land-based biological and diving teams collected samples along transects, while the Aleuts collected subsistence foods they normally eat and foods they would eat when stranded. Personnel included a team of four biologists from Rutgers University (New Jersey), four Aleuts from local villages (present on the vessel), and four divers from the University of Alaska.

Species collected were processed in the on-board laboratory at night to reduce the volume of material. All appropriate sampling procedures and chain of custody were followed and samples were placed in freezers. Large enough samples were taken to allow for radionuclide analyses and later mercury analyses, as well as for archiving [[54], [55]]. In the end, we collected samples from 4 algae species, 12 invertebrate species, 17 fish species, and 5 bird species (including eggs from two of these species). Composites consisted of samples from many different individuals of the same species. That is, for algae, fish, and birds, sample material from five individuals was combined to form one composite sample for analysis. For small invertebrates, material from many more individuals was required to make a composite sample. Samples were further processed at Rutgers University, and radionuclide analyses were completed at Vanderbilt University and Idaho National Laboratory.

The amount of sample material in composites was an important aspect of analysis for ¹³⁷Cs and actinides. Rather than present the values for each species, we grouped the species into different tropic levels: 1) primary producers (plants), 2) low tropic level grazers and filter feeders, 3) lower-level predators, 4) medium level predators, and 5) top trophic level predators. Because the radionuclide levels were so low or non-detectable, we present two types of data: percent of composites above the MDA (minimum detectable activity), and levels of radionuclides in samples that were above the MDA. Data on percentage above the MDA were provided. The percentage that was below the MDA indicates that for those species, there was likely no or little risk.

Table 1 provides the number of composites that were above the MDA, listing the number of composites analyzed and the number of species in each trophic level. That is, there were a total of 84 composites analyzed for primary producers for each radionuclide; primary producers included five algae species (Table 1). Some trophic levels had no samples above the MDA for a particular radionuclide, and others were 100% above the MDA. This information accomplishes two things: it shows which groups were not accumulating radionuclides and identifies species to use for biomonitoring. Several conclusions can be drawn from this

Table 1.Percentage of composites from Amchitka that were above the MDA as a function of species and isotopes examined in the Amchitka Study (1000 gr samples for¹³⁷Cs, 100 gr samples for actinides, after Burgeret al.[[34],[38],[41]]).Given is the mean percentage for each group, as well as the number of analyses and species in each group.

Note: There were no composite samples with values above the MDA for any species for ¹²⁹I, ⁶⁰Co, ¹⁵²Eu, ⁹⁰Sr, and ⁹⁰Tc.

table regarding detection: 1) ¹³⁷Cs composites were below MDA for organisms low on the food web and were above MDA for organisms that were high on the food chain, 2) ²³⁴U and ²³⁸U were mainly above the MDA for organisms low on the food chain (i.e., 100%) and low for medium and top-level predators, 3) ²³⁶U had a low percent above the MDA for species low on the food chain (and none for top level predators), and 4) a low percentage of organisms had values above the MDAs for ²⁴¹Am and ^239-249Pu (Table 1).

Examining both the percentage of samples above the MDA (shown in Table 1) and the mean values is useful in determining potential bioindicators for future monitoring. Algae (primary producers) and filter feeders had some actinide (U, Pu, Am) levels above the MDA, and the mean values for three of the species with the highest levels are shown in Table 2. Table 2 indicates that Fucus generally had the highest levels of actinides, meaning it would be a good bioindicator for future monitoring.

For high trophic level fish and birds, all the composites with values above the MDA for ¹³⁷Cs levels are shown in Table 3. The species shown in Table 3 are those within each trophic level that had the highest

Table 2.Comparison of mean actinide specific activity levels (Bq/kg) among lower trophic level organisms collected on the Amchitka Expedition. Values in parentheses mean thatonly one or two of the composites were above the MDAs. Values were significantly different (P > 0.01) for all but²⁴¹Am[[34],[40]].Such information is essential to assess human health risk but also to design a future biomonitoring program.The highest value for each actinide is highlighted.

Table 3.Levels of¹³⁷Cs in fish and birds from the Amchitka Expedition that were above the MDA[[41]]. All the other fish that were analyzed did not have any composites above the MDA.

^aThese values are what one might get if they ate a fish or bird above the MDA.

detectable values. The three high values for dolly varden were from composites collected from Lake Cannikin on Amchitka Island that formed right above the underground test blast. Pacific Cod is high because they are top level predators. Figure 3 is a summary figure illustrating the percentage of composites above the MDA, the mean values for ¹³⁷Cs, the number of composites and number of individuals, and the maximum age of each species. In general, the higher the trophic level, the higher the specific activity of ¹³⁷CS (in Bq/kg). The values in Table 3 are those a consumer might expect to get if they consumed that species at the higher range. For fish, larger fish generally have higher specific activities.

Figure 3.Percentage of samples above the MDA and the mean specific activity (Bq/kg) for key species collected at Amchitka Island.Also provided are the maximum age and number of composites analyzed (and individuals that made up those composites).

For most of the organisms sampled, the levels of radionuclides were below the MDA’s, the levels were lower than established health standards, and they were lower than recorded in other oceanic locations in the northern hemisphere [[38], [40], [41]]. Some of the high trophic level fish had levels of mercury that are of concern, especially for pregnant women [[54]-[58]]. The information we provided allowed individual Aleuts to determine which invertebrates or fish to eat or to avoid eating because of contamination. Further, the source of mercury was atmospheric deposition, and not from DOE activities. The Aleuts, however, appreciated the information on both radionuclides and mercury [[38], [54], [56], [58]].

Results for the analyses were presented to all the stakeholder groups in open, public meetings, including DOE, NOAA, US F&WS, ADEC, APIA, and the public. In addition, we re-visited Aleut villages. Communication and collaboration at every phase in the Amchitka Project led to openness, transparency, and acceptance of the data [[34], [35], [37]]. Two additions to the original Science Plan were very important to success: 1) including four Aleut hunters and fishers on the expedition itself and 2) the addition of a fisheries biologist on the NOAA research trawler. These two additions ensured that appropriate commercial and subsistence species were collected, as well as engendering trust in the data. Attitudes and perceptions related to the health risk from radionuclides among Aleuts, for example, indicated a significant decline in concern about radionuclide exposure [[37]].

Data from the expedition were used to propose a biomonitoring plan that included representative species at all levels on the food web and included key commercial and subsistence species [[34]]. The plan was to be implemented every 5 years, but an earthquake in the region should trigger an immediate census of a small set of biomonitoring species. If any radionuclides were detected, an immediate assessment of the species originally examined would be triggered. An examination of the radionuclide tables by DOE, Tribes, regulators, and different stakeholders in public meetings suggested that top-trophic level organisms were better indicators of ¹³⁷Cs exposure, while primary producers were generally the best for actinides overall. The results suggest that the primary producers should be examined with respect to the algae species of most concern for subsistence people, and fish and birds should be examined for ¹³⁷Cs. Once Amchitka Island was transferred from DOE-EM to long-term stewardship, our biomonitoring plan was implemented on a 5-year review cycle [[53]].

When we discussed the Science Plan with Aleuts on different Aleutian Islands, they were also quite concerned about mercury exposure, and we later analyzed mercury levels using non-DOE funds [[54]-[57]]. These results were reported back to the Aleuts when we returned to report on radionuclide exposure. Subsequently, Hardell et al. [[58]] examined polychlorinated biphenyls (PCBs) and three organochlorine pesticides in some of the fish from the Amchitka Study. These three studies (radionuclides, mercury, PCBs) provided the Aleuts and other consumers of fish and crabs from the Bering Sea data on potential risk from consumption of foods from these waters. These contaminant studies were important for the Aleuts in making informed decisions about eating fish and crabs gathered from around their islands.

In this study there were two major points at which trust and consensus were needed to move forward toward the solution: 1) agreement on the Science Plan, and 2) agreement that radionuclides did not pose a human or ecological health problem allowing DOE to turn over the land to the State of Alaska and U.S. Fish and Wildlife Service (US F&WS). The original agreement between the parties (DOE, State of Alaska, EPA, and APIA) required an approved Science Plan before implementation that would answer the main questions of all parties regarding radionuclide exposure. This was accomplished through in-person meeting and approval of all signatories. The second point involved meeting with all parties to present results, and their final agreement that DOE could proceed with closure of the Island. APIA and the Aleut community were the most concerned about radionuclides because most of their diet came from subsistence foods gathered in the region. While APIA itself agreed with the findings, we demonstrated trust in the data and consensus on moving forward by collecting data on Aleut concerns prior to the expedition, and following the expedition as expressed in public meetings conducted in their Aleutian villages. There were significantly fewer concerns about radionuclides and the safety of their food following the expedition [[37]].

DOE has sites still requiring remediation in many states and it is important to be able to identify which sites have unique and rare ecosystems that are likely to include threatened and endangered species [[23]-[26]]. One of the goals of DOE-EM is to conduct remediation, restoration, and long-term protection of stored nuclear and other wastes while maintaining protection of human health and the environment [[59]]. Some DOE sites have uncontaminated buffer areas as well as cleaned-up areas suitable for a variety of recreational or cultural activities. These areas are important to a range of stakeholders from Tribal, cultural, and subsistence users to commercial and recreational hunters and fishers, as well as for neighbors and others worried about human health and safety. DOE-EM sites have some level of human health and environmental monitoring, although in some cases it is only soil and water. For many sites the protection and conservation of ecological resources is part of future land use plans, even at the beginning of DOE-EM program [[60]]. Thus, sites usually have on-site ecological monitoring for species’ health (i.e., changes in population sizes) and for contaminant levels. In the latter case, assessing contaminant levels and trends provides not only information to determine the potential risk to the plants or animals themselves, but as surrogates for exposure to people and as direct indicators of exposure to people consuming them (e.g., deer, fish) or using them for medicinal or cultural purposes. DOE-EM has many sites and limited resources. DOE, other federal and state agencies, Tribal agencies, and community members have expressed interest in protection of ecological lands on DOE-EM sites, particularly in relation to regional resources. There are three metrics that will be evaluated here: 1) ecological resources on a DOE site, 2) DOE site ecological resources compared to regional resources, and 3) ecological resources on one DOE site compared to another. The results for these three metrics can also inform DOE and state agencies about the allocation of funds to site-specific protection of funds.

One of CRESP’s on-going projects is to examine and develop indicators that can be used across DOE sites. After discussions with DOE, relevant DOE National Laboratories, DOE contractors, state regulators, resource trustees, and the public over many years, it was obvious a screening indicator was needed that could quickly identify whether there were potential high quality ecological communities on-site, off-site, and in regions surrounding a given site. No individual indicator (e.g., a species) could be used at all sites because the DOE sites are in many different eco-regions across the U.S. from forests to deserts [[5]]. For any non-urban area, there is a characteristic vegetation type influenced by water, temperature, and soil conditions. Even a denuded area would be predicted to undergo gradual changes in vegetation, until it becomes the characteristic vegetation. This succession may take a century or more. It is called the “climax” vegetation because once reached, it is stable and characteristic of an area (at least until climate change alters conditions).

5.1.1. Example: Idaho National Laboratory (INL)

INL is located in shrub-steppe and grassland eco-region in the western United States (Figure 1). Cleanup involved dealing with special nuclear materials, spent nuclear fuel, transuranic waste, tanks waste, and soil and ground water contamination [[59]]. In such a large site (231,105 ha), the NLCD [[42], [43]] land cover maps can capture and portray the overall land types (Figure 4). Figure 4 illustrates that there is little development in the region; shrub/scrub dominates the habitats in the region. There is no agricultural cultivation on site while agriculture (for example, potatoes, barley) is important in the regions surrounding the site. Grassland on site extends out into the surrounding region. Livestock grazing occurs on- and off-site but the NLCD does not identify grazed versus ungrazed grassland. However, although the maps visually display these differences, the actual data illustrate the relative proportion of habitats on- and off-INL (Table 4). Several observations can be made from this table: 1) shrub-scrub (also called shrub-steppe) dominated the site and the region, 2) INL has significantly more grassland that the surrounding region, and 3) the surrounding region has significantly more agriculture cultivation. The actual area is important for DOE, regulators, and the public for the purpose of future land use planning and understanding whether key habitats on site are being protected. This understanding is particularly important at INL because some of the surrounding regions may be vulnerable to future agriculture development, making on-site habitats even more rare and valuable.

5.1.2. Example: Rocky Flats (RF)

In contrast to INL, Rocky Flats (RF) is a small site of 21 km² that is located 26 km northwest of Denver [[60]]. From 1952 to 1989, its mission was to fabricate plutonium pits, the trigger initiating the fission of an atomic bomb explosion [[61], [62]]. For security, the industrial facilities were in the center of the site. RF is an example of a DOE-EM site that was remediated as of 2005, leaving some residual hazardous material contained in a central exclusion area. RF is now in long-term stewardship whereby the DOE retains responsibility for the developed/nuclear area in the center, but the surrounding land has reverted to ecological lands as a national wildlife refuge under the auspices of the U.S. Fish & Wildlife Service. Recreational

Figure 4.Map of Idaho National Laboratory (INL) illustrating different land use types on site and in the 10-km and 30-km band around INL (after Burgeret al.[[46]]).Note that while the map illustrates different types of habitat, it does not indicate the amount of each.

Table 4.Comparison of land use cover (and habitats) on the Idaho National Laboratory with the surrounding 10-km and 30-km bands[[43],[46]]. Area given is ha and (percent).100 ha = 1km². (significant difference between INL and region of P < 0.05, Tukey’s Test).

visitation on the refuge is confined to the surrounding buffer area which is primarily dry grassland similar to what once covered thousands of hectares in the U.S. plains [[63]]. The NLCD [[42], [43]] can also capture the land use currently on the site. Because of the smaller size of RF, a better communication tool is to show the site itself and the surrounding regions in separate maps; otherwise, the land cover on RF is not visible (Figure 5). Several conclusions are obvious from Figure 5: the DOE facilities on RF are in the center, surrounded by shrub/grassland, technically a xeric tall grass prairie. The development in the surrounding region is to the west of RF; there is a band of shrub/grassland that runs north and south through RF and the 10-km buffer, and forests run north and south to the west of the site.

Figure 5, however, does not provide data that can be used to directly compare the amount of each habitat type on- and off-site, nor does it allow for a discussion of the importance of RF relative to the surrounding landscape. This evaluation requires a comparison of the data. Using the NLCD we computed the amount and percentage of each land use type (Table 5).

Table 5.Comparison of land use on Rocky Flats (RF) compared to the 10-km and 30-km buffer zones around RF (after Burgeret al.[[48]] and unpub. data). Area given is ha and (percent) (Significant difference between INL and region of P < 0.01, Tukey’s Test).

^aWhere two types are present, the percentage does not distinguish how much is of each type (e.g. what percent is crops versus pasture).

Rocky Flats have significantly more grassland/herbaceous vegetation and less development than the surrounding region (Table 5). Figure 5 shows that the development on Rocky Flats was concentrated in the middle of the site surrounded by grasslands that are continuous with those in the 10-km band around the site. This xeric tall grass prairie is a rare habitat. The data in the NLCD, however, do not describe the kind of grassland, although the NLCD [[42], [43]] does describe different types of development. The grassland/herbaceous

Figure 5.Comparison of land cover on Rocky Flats with the surrounding 10-km and 30-km buffer bands based on the National Land Cover Database[[42],[43],[48]].

habitat on site is important because it is continuous with the adjacent land and has been relatively undisturbed for decades. Because the land was allowed to undergo succession to the climax vegetation, it is in better condition than habitat in the surrounding area. Moreover, some of this habitat on site is tall grass prairie, one of the most threatened vegetation types in the United States [[63]]. Rocky Flats is unique because it has tall grass prairie, and it is protected; the undeveloped land is now a U.S. Fish and Wildlife Refuge, with tourists and recreationists coming from Dever and other regions [[64]]. Considerable radionuclides and other hazardous materials continue to reside on site [[62]]. The developed, core area will remain under DOE-Legacy Management in perpetuity [[60]].

The screening bioindicator of climax vegetation developed by CRESP has been tested at other DOE sites [[44]-[48]] but requires more in-depth analysis to determine its usefulness by DOE, regulators, and the public. This analysis will involve further discussions and refinements after public input.