Hakonson Report - full version

          Public Comments

More about the Justice Department cover-up of government nuclear crimes.

Review of Sandia National Laboratories/New Mexico
Evapotranspiration Cap Closure Plans for the Mixed Waste Landfill
by Tom Hakonson, Ph.D., Environmental Evaluation Services, LLC

The following report was made possible with a grant from the Monitoring and Technical Assessment Fund (MTA) to assist in performing independent technical studies of the Mixed Waste Landfill (MWL), a hazardous legacy waste site located at Sandia National Laboratories (SNL). The funding, established as a part of a $6.25 million court settlement between the U.S. Department of Energy (DOE) and 39 nonprofit and environmental groups, assists tribes and other non-governmental organizations in conducting their own independent technical studies of sites at DOE facilities.

Citizen Action commissioned Dr. Tom Hakonson, a former environmental scientist with Los Alamos National Laboratory, to perform an independent peer review of the cap closure plans proposed for the MWL. A copy of Dr. Hakonson's curriculum vitae and a list of his published papers are included with this report.

"I am willing to state, unequivocally, that most of the environmental processes discussed in this report will, without doubt, affect the long-term distribution and transport of contaminants in the
Mixed Waste Landfill."

- T.E. Hakonson

REVIEW OF SANDIA NATIONAL LABORATORIES/NEW MEXICO EVAPOTRANSPIRATION CAP CLOSURE PLANS FOR THE
MIXED WASTE LANDFILL

REVIEWED BY T. E. HAKONSON

ENVIRONMENTAL EVALUATION SERVICES, LLC
P. O. Box 315
Daniel, WY 83115

Prepared for Citizen Action
P.O. Box 1133
Sandia Park, NM 87047
2/15/02
TABLE OF CONTENTS

EXECUTIVE SUMMARY 3

A. Objectives Of This Review 10
B. Hakonson's Technical Background And Expertise Relevant
To This Review 11
C. Specific Topics Addressed In This Report 12
D. Technical Concepts And Issues 14
D 1. General Considerations 14
D 2. Regulatory Requirements 15
D 3. Regulatory Basis for Alternative Cap Designs 18
D 4. Relationship of Water Balance to Cap Design 19
D 5. ET Cap Design 20
D 6. Causes Of ET Cap Failure 22
D 6.1 Physical Transport Processes 22
D 6.1.1 Landfill Surface 22
D 6.1.2 Subsurface Processes 26
D 6.2 Biological Processes 27
D 6.2.1 Plant related Processes 27
D 6.2.2 Root Distributions in Soil 29
D 6.2.3 Animal Related Processes 30
D 6.2.3.1 Uptake 30
D 6.2.3.2 Burrowing 31
D 6.2.3.3 Burrow Depths 33
D 6.2.3.4 Rates of soil turnover 33
D 6.2.3.5 Effects on Soil Characteristics 34
D 6.2.3.6 Effects on vegetation cover 36
D 6.2.3.7 General Effects on Erosion 37
D 6.2.3.7.1 Effects on Wind Erosion 38
D 6.2.3.7.2 Effects on Water Erosion 39
D 6.2.3.8 Long term Biological Intrusion 40
D 6.2.3.9 Conclusions About Plant and
Animal Intrusion 42
D 6.2.4 Human Intrusion 43

E. Review Of SNL/NM Proposed ET Caps and Recommendations 44
E 1. General Comments 44
E 2. SNL/NM ET Cap Designs 45
E 3. Post-closure Monitoring Issues 48
E 4. Summary Remarks 51

F. Literature Cited 53
G. Appendix A- Documents Read for this Review 70
H. Appendix B- Hakonson's Resume 72

Disclaimer

The portions of this report relating to the SNL/NM ET cap are based on a review of documents provided to me by Citizen Action as listed in Appendix A of this report. As such, my recommendations in this report are also based upon that review of Appendix A documents. If other SNL/NM documents are available that would relate to my review and recommendations, I was unaware of such documents during completion of this report. This means that my recommendations in this report stand until such time as additional documents, should they exist, are identified and reviewed.

Tom Hakonson, Ph.D.
Environmental Evaluation Services, LLC

Executive Summary

This report presents the results of my technical review of the vegetated soil cap (ET cap) that SNL/NM has proposed for closing the Mixed Waste Landfill (MWL). Conceptually, an ET cap consists of a vegetated soil layer that represents an optimum of soil type, soil depth, vegetation cover, surface slope, and surface management practice in order to control erosion and minimize percolation of soil moisture into the waste. Given that the outcome of the forthcoming Corrective Measures Study and revised risk assessment for the MWL may lead to the use of other remediation alternatives, it is still certain that a final cap will be required as a part of any closure plan selected for the landfill. Furthermore, that final cap will contain all of the functional elements of the proposed ET cap, including soil for moisture storage and vegetation to remove soil moisture via transpiration and to control erosion.

This report presents discussion on 3 topics: 1) a literature review of general concepts concerning landfill closures, capping design alternatives, failure modes, and long term performance, 2) a review of the two evapotranspiration cap design reports offered by SNL/NM for closure of the MWL, and 3) some recommendations concerning the proposed ET cap closure and post-closure monitoring period.

Depending on the reader, there may be concern or delight in the fact that I did not attempt to relate the information I derived from the literature review to specific conditions at the MWL. I have purposely not attempted to link results from the literature review to the MWL because it would require me to speculate about conditions at the MWL that cannot be verified with existing data. For example, there is nothing published on the MWL that quantifies or characterizes the kinds of fauna and flora present or the amount and consequences of biological intrusion, subsidence, or soil erosion on contaminant distribution and transport.

Despite my unwillingness to be specific about the degree to which physical and biological processes will impact the MWL, I am willing to state unequivocally that most of the environmental processes discussed in this report will, without doubt, affect the long term distribution and transport of contaminants in the MWL I would assume that the NMED, EPA, and the owners of the landfill would have a vested interest in documenting just how important these processes are in assuring the long term safety of workers and the public. While SNL/NM may believe that the MWL site has been characterized sufficiently to answer all of the important questions about present and future transport of MWL contaminants, I can assure the reader that important questions remain unanswered

It is clear that applicable EPA regulations permit the use of vegetated soil covers for closure of radioactive and hazardous waste landfills. Criteria permitting their use include the outcome of the health and environmental risk assessments, the climate at the burial site, and demonstration that the alternative cover design provides protection of the waste equivalent to traditional engineered barrier designs that are recommended by EPA. In the absence of site specific performance data, this evaluation of equivalency relies heavily on the use of water balance modeling and relevant site characterization data. Based strictly on this hydrologic analysis, it would appear that the use of an ET cover for closure of the MWL is justified. However, this analysis ignores the potential effects of biological processes in mobilizing buried contaminants and the consequences of this transport on future changes in contaminant concentrations in surface soil. Because SNL/NM proposed to only monitor tritium in the vadose zone for a few years post-closure, changes in contaminant concentrations in surface soils and biota would go undetected.

A modeling study conducted by Pacific Northwest Laboratory addresses the issue of long term consequences of biointrusion into arid site low level waste landfills on dose to man. Results show that estimated dose to man resulting from biological transport of radionuclides at two reference low level waste sites was of the same order (i. e., about 50% less) as dose calculated from a human intrusion scenario.

The potential importance of processes that contribute to contaminant transport by physical and biological processes operating near and on the ground surface stems from the fact that only a couple of meters of cover soil (i. e., about 1 meter of existing soil, 0.5 meters of subgrade, and 1-1.5 meters of ET cap soil) will separate the MWL waste from the ground surface when the ET cap is installed. In contrast, several hundred meters of unsaturated soil in the vadose zone separates the waste from ground water. In arid environments such as the MWL, transport of most if not all of the MWL contaminants through this extensive, dry vadose zone is certain to be low and slow even for tritium which exhibits 2 phase transport. The fact that tritium currently emanates from the surface of the MWL is most certainly related to a large degree to the presence of burrowing animals and vegetation present on the landfill surface and the effects of these organisms have on soil moisture status and soil porosity.

Both plants and animals have the potential to transport buried waste to the ground surface. Plants do so via roots that can penetrate several meters into the landfill. Furthermore, most plant species have the capability to penetrate the relatively thin cover soil layer proposed for the MWL. This means that the term, "shallow rooted" as used by the SNL/NM ET cap designers is inappropriate given that the grass species that they propose to use to revegetate the ET cover all have the capability to send roots several meters into the soil. If soil moisture penetrates beyond the existing rhizosphere, plant root distribution will extend downward to capture moisture at the deeper depths.

Roots in contact with waste can incorporate soluble constituents and transport them to the ground surface. This uptake process is analogous to a one-way valve in that contaminants are pumped upward to above ground vegetation that eventually senesces and deposits associated contaminants on the ground surface. Burrowing by animals and insects also has the potential to access buried waste several meters below the ground surface. This can lead not only to chemical and radiation exposures to the organisms but also to physical transport of the waste upward in the soil profile and to the ground surface.

Should contaminants be transported to the ground surface, several complex but coupled processes involving enrichment of soil fines and associated contaminants begin to operate to transport soil contaminants to biota, across the landfill surface, and to offsite areas. These processes include erosion by wind and water, transport by contaminated animals moving on and off the landfill, deposition of soil particles on biological surfaces from rainsplash and wind resuspension, and wind transport of senescent vegetation to offsite areas.

The importance of erosion of cover soil over long time frames needs to be carefully evaluated in light of potential disturbances by burrowing animals in combination with loss of vegetation cover resulting from catastrophic disturbances such as fire, disease, and drought. Based on recent field studies, fire disturbances can change hydrologic (and presumably wind) erosion rates by as much as a factor of 25 over undisturbed conditions. This means that modeling results based on average erosion rates may under-represent the actual long-term rates of deflation of the cover soils.

The persistence of effects caused by disturbances is not well known. Some studies show that the effects of fire on soil erosion may persist for several years while other studies suggest that these effects are short lived and depend on the rate of recovery (e.g., reseeding) of the disturbed area. The relative importance of these processes in mobilizing MWL contaminants will depend on several factors including vertical distribution of waste contaminants in the landfill trenches and pits, soil type and depth, type of vegetation cover, changes in the vegetation cover, animal and insect species composition, and changes in faunal species composition.

Burrowing by animals also creates extensive disturbances of the soil profile. While it could be assumed that these disturbances would lead to accelerated erosion and percolation of water to the buried waste, the small amount of available research data suggest that burrowing and soil casting to the ground surface has relatively small effects on erosion and percolation as long as a good vegetation cover is present on the soil surface. Increased infiltration of water into the soil from animal and insect burrowing is often followed by an increase in vegetation cover biomass and ET, which combined, reduces the potential for erosion and deep percolation. However, erosion and percolation increase dramatically when the vegetation cover is absent in the presence of burrowing.

The magnitude of subsurface transport is intimately related to processes that operate on or near the landfill surface. Aqueous phase transport of contaminants is dependent on the ability of the cover to control the amount of moisture that penetrates into the waste. Key among those processes is evapotranspiration, which has the potential under most arid site conditions to remove virtually 100% of the moisture that infiltrates into the soil.

However, it is ironic that a cover that is effective in minimizing soil moisture in the landfill can also contribute to an increase in vapor phase transport of volatiles such as tritium. The relative importance of aqueous versus vapor phase transport of tritium at the MWL will be difficult to determine but will depend on a host of physical, chemical, and biological processes that are complex and coupled. The fact that tritium moves in more than one phase ensures that it will be relatively widely dispersed from the initial burial location. Therefore, I am certain that monitoring data from the MWL will show that tritium is currently present in fauna and flora.

A further complication is that if moisture does penetrate through the landfill cover, plants have the ability to send roots downward in pursuit of that moisture. This means that the concept of shallow rooted plants versus deep rooted plants is misleading in that most "shallow rooted" plant species have the capability to send roots much deeper than the couple of meters of cover proposed for the MWL. The good news is that this plasticity of plant roots to penetrate downward in search of moisture helps ensure that very little moisture will escape into the vadose zone. The bad news is that deeper penetrating roots can also contact buried waste and transport plant available contaminants to the ground surface.

Because burrowing organisms can come into much closer contact with buried waste, it is also possible that they can be exposed to relatively high chemical and radiation doses. Radiation doses to free ranging burrowing animals that live on the MWL would be relatively easy to measure. A technique that was developed in the 1970's uses thermoluminescent dosimeters (TLD) that can be implanted or attached to free ranging animals. When retrieved, the TLD's provide a good measure of radiation exposures to the organisms over time.

Data on the concentrations of contaminants in plants, animals, and cast soil at the MWL apparently are not available but would be instructive about the potential of biological intrusion for mobilizing MWL waste. Because the site was opened in 1959 and closed in 1988, there may be some portions of the landfill site that have been undisturbed for several decades. Directed sampling in those areas would provide a measure of the importance of biological transport.

I recognize that SNL/NM conducted grid sampling of soils for the Phase 2 RCRA Facility Investigation of the MWL in 1990. However, the RFI soil sampling was coarse in resolution, non-random in space, involved at least some sampling areas that were recently disturbed (e. g., Trench F backfilled just prior to soil sampling), and did not purposely include disturbances created by burrowing animals.

Concerning human intrusion, a conservative approach would be to assume that institutional control is lost and that humans come to occupy the landfill surface for a home site, growing crops, industrial activities, or other uses that are intimately associated with the landfill. There would seem to be no easy answers on how to prevent this intrusion, but I would start by considering the use of marker systems placed judiciously at the site during closure activities. For example, ceramic or glass tiles, ala Anasazi clay pottery with embossed warning messages, could be scattered beneath the cover as it is constructed so that any future excavation on the landfill would encounter the warning tiles. Surface markers could also be constructed but one would have to assume that such tiles or surface markers do not become an attractive nuisance, i. e., become collector's items.

The argument could be made that the use of a marker system in the early phases of the closure increases the possibility that the landfill owners will forget about the landfill. Consequently, it is prudent to get a binding administrative and financial commitment from the landfill owners and appropriate regulatory authorities to fulfill all obligations during the period of institutional control. I am not sure how this would be accomplished from a legal view but I would presume it might involve an escrow account that would cover any reasonable projections of future problems.

Specific to the SNL/NM ET cover designs for the MWL, my review led me to three conclusions. First, both reports do a credible job of analyzing the ET cover for the MWL given the guidance provided by EPA. They also adequately discussed the regulatory and technical basis for the ET cap and used the results from several modeling studies to evaluate design variables. Construction details in both reports were sufficient to convince me that the ET cap could be built to specifications.

One could quibble about technical details relative to the respective designs. However, in my opinion deficiencies in the areas described below far outweigh the relatively minor problems with cap design. Let me say that I believe that an ET cap closure for the MWL, as described in both SNL/NM reports, will provide adequate protection against percolation of site contaminants to ground water. However, this assumes that the site is diligently monitored and maintained throughout the post closure period. It also assumes that the surface pathway involving biota proves to be unimportant in contributing dose to man.

This leads to what I believe is one of the more important deficiencies in the proposed MWL closure, namely the assumption that vertical and horizontal transport of site contaminants resulting from biological processes is not an important contributor to exposure pathways. My review suggests that relevant data from the MWL on contaminants in vegetation, animals, and soil cast to the surface by burrowing animals apparently do not exist. The reason biointrusion may be important is that it represents the major mechanism leading to vertical transport of contaminants to the ground surface and through the drying effect of plant transpiration on cover soils, plays a major role in the evolution of volatile contaminants from the ground surface. While vertical transport by biota may be small on a short time scale, over many decades these processes may become dominant in mobilizing buried waste.

It is my opinion that the soil sampling done by SNL/NM in 1990 as a part of the Phase 2 RFI provides little information that can be used to answer questions about the effects of biointrusion in transporting MWL contaminants to the soil surface. The RFI soil sampling grid resulted in evenly spaced samples (i. e., that were non-randomly distributed), that provided coarse spatial resolution of contaminant concentrations, and that involved sampling locations that were recently disturbed such as Trench F where backfill was added just months before the soil samples were taken. Furthermore, those samples that were taken in 1990 represent a single snap shot in time and depending on the degree of past mechanical disturbances that occurred within the MWL boundaries, they may represent a snap shot with little elapsed time between soil surface disturbance and when the soil samples were taken.

Technical evaluation of biointrusion at the MWL would involve a careful survey of contaminants in surface soils and biota and identification of species presently occupying the site. This could be done now on trench/es closed early in the landfill operation and assuming that the ground surface has not been disturbed since the trench/es were closed. If repeated mechanical disturbances of the ground surface have occurred at the MWL throughout it's use history, then the alternative method for evaluating the importance of biointrusion would be to initiate a long term sampling program after the site is closed.

I would add that the addition of less than two meters of clean soil during ET cap construction does not assure that problems with biointrusion go away. Most plants and many animals have the potential to penetrate deeper than the proposed thickness of the ET cover.

The third conclusion I drew from the review is that little or no planning has been done on the post-closure phase of the MWL closure. It is possible that the prevailing philosophy behind this lack of planning and guidance is to address problems as they are recognized. Obviously the ability to recognize problems with the containment system is dependent on diligent monitoring of all relevant pathways.

Long-term data are not available to demonstrate how well vegetated soil covers will work in preventing transport of buried contaminants over extended time frames. Furthermore, it is likely certain that the proposed ET cover will not be 100% effective in isolating MWL contaminants from the biosphere over long periods of time. Therefore, a comprehensive monitoring program during the period of institutional control will be important for verifying the validity of initial closure assumptions and calculations and for identifying potential problems.

The general consensus of the scientific community is that some problems with the containment of waste in the landfill will occur over the time frames involved. This places a special burden on the owner of the MWL to identify and resolve problems in real time. Without careful monitoring of the site during the post-closure phase, little problems can become bigger problems that may be defy remedy and will certainly elevate costs for solutions. This aspect of the MWL closure would appear to be especially important since statements are made that the ET cap closure is intended to lead the way for other DOE landfill closures using alternative caps.

In my opinion, the post closure monitoring plan should provide for measurements on all possible migration pathways including movement through the vadose zone, surface contamination, and biological transport. Those measurements will be invaluable for validating or invalidating some of the assumptions and data used in the risk modeling. While some or all of these pathways may eventually prove to be unimportant from a risk perspective, the lack of consideration of plant and animal intrusion into the MWL and the consequences relative to effects on concentration in surface soil, soil erosion and other means of contaminant transport detract from the proposed closure.

A further problem is the lack of a well-defined plan of action should the cap not perform as predicted. This means that there are no decision criteria or action plans for mitigating the various failure modes should one or more occur. The ET cap reports do not discuss these issues and in my opinion, they are vital to the credibility of the proposed closure.

Whether the MWL requires removal or not is not for me to judge. I do believe that a well designed cap, a financially secured, quality post-closure monitoring plan, and plan of action in the event of a containment problem/s, will likely work for the MWL, at least until re-evaluation of the site is made at some point in the future. However, based on documents I reviewed, SNL/NM has done little or nothing of substance on evaluating the long term effect of biointrusion on the surface pathway, developing a post-closure monitoring plan, or establishing decision criteria for possible corrective actions for the MWL closure.

I recognize that the costs of any additional sampling before and after MWL site closure and including possible future corrective actions will fall directly on the taxpayers of this country, not the State of New Mexico, not DOE, nor SNL/NM. Setting aside large amounts of public funding for some unspecified corrective action to remedy problem/s with a low probability of occurring would not seem to be the best use of financial resources for protecting public health given that the taxpayer will be expected to shoulder the financial burden of any needed future fix for the site. Latest cost estimates for excavation, transport and disposal of mixed waste, based on the Area P clean closure at Los Alamos, are about $10,000/ yd3.

In my opinion, I believe secure funding would be better spent on the post-closure monitoring phase for the MWL because a well designed monitoring program with assured continuity will provide early identification of potential problems with the MWL containment system. I believe that early identification of problems with the closure will greatly reduce potential costs of corrective actions that might be required to fix unspecified future problems. Furthermore, it will provide a defensible basis for recommending for or against further closures using the same technology.

A. Objectives Of This Review

This report to Citizen Action represents my attempt to evaluate the technical merits of Sandia National Laboratories, New Mexico (SNL/NM) plan to use a vegetated soil cap for closure of the Mixed Waste Landfill (MWL). The MWL, located on Kirtland Air Force Base in Albuquerque New Mexico, is about 2.6 acres in size and was used for disposal of radioactive and hazardous waste. Details concerning what are known about the use history of the MWL including the kinds and amounts of waste are presented in several existing documents and will not be repeated here (see documents reviewed in Appendix A). Suffice it to say that the MWL use history began pre-1976 (pre-RCRA). This means that the landfill is unlined, used a tip and dump disposal method, contains a variety of radioactive and non-radioactive contaminants in various waste forms, and that precise inventories of all of the radionuclides, metals, and chemicals that went into the landfill are not known.

The MWL has been the subject of several special studies to characterize waste contaminant concentrations, inventories, distribution, transport, and associated risks. The SNL/NM closure of the MWL proposes to use an evapotranspiration (ET) cap and to implement a post-closure monitoring and maintenance program based on the ground water pathway. I think their intent is to reassess the need for additional closure measures a few decades into the future, when relatively short-lived radionuclides have decayed. Details concerning if and when this reassessment will occur are vague at best.

My review is primarily focused on the technical aspects of the proposed evapotranspiration cover closure of the MWL to include some post-closure monitoring issues. I do not intend to address topics concerning accuracy or completeness of the waste inventory, methodology and estimates of risks, the accuracy and representativeness of the monitoring data, or accuracy and completeness of published information by the various stakeholders.

I refer repeatedly in this report to the SNL/NM risk assessment as described in the SNL/NM Phase 2, RCRA Facility Investigation report published in 1990. It has been pointed out that this risk assessment may no longer be valid just as the proposed ET cap closure may no longer be valid depending on the outcome of the proposed Corrective Measures Study being imposed upon SNL/NM. Because my charge was to evaluate the ET cap proposed for the MWL, it seems logical to use the Phase2 risk assessment that was used in part to justify the selection of the ET cap. Furthermore, since my goal was to evaluate the ET cap, I believe it is appropriate to refer to this risk assessment with the caveat that it (and the proposed ET cap) may no longer apply to what eventually happens with the MWL.

It is obvious from reading the various documents supplied to me as a part of my assigned task, that some issues related to the MWL are contentious, including the choice of closure alternatives. I do not intend to become embroiled in those issues in this report. Whether the ET cap is the right choice or not for the MWL closure is not for me to judge given that the final decision must weigh non-technical (i. e., social and political), regulatory, as well as technical issues.

The task at hand in this review is whether a cover closure plan proposed by SNL/NM for the MWL is technically defensible. My approach will be to review what is known and not known about environmental processes that will likely affect the long-term ability of the cover to isolate the underlying waste.

A recent decision by the State of New Mexico Environment Department directs SNL/NM to conduct a Corrective Measures Study (CMS) to reconsider several alternatives, in addition to the ET cap, for closure of the MWL. This means that the final closure plan for the MWL may or may not involve the use of an evapotranspiration cover as proposed by SNL/NM. I expect, however, that SNL/NM will continue to favor the use of a cover as the primary means of closure for the MWL because their monitoring data and risk assessments to this point in time indicate that the potential for migration to ground water and exposure of receptors is low. However, this may change if new monitoring data, risk assessments, and corrective measures studies do not support the use of a cover closure as the sole remedy for the landfill.

Even if the ET cap is not selected as the central feature of the MWL closure, I am certain that under almost any cleanup scenario chosen for the site, including waste removal, that a final cap of some kind will be required. Furthermore, the waste removal option is certain to be less than 100% efficient in removing MWL waste contaminants such as tritium. Depending on the amount and kinds of residual contamination, post-closure maintenance and monitoring may be required for the removal option.

There may be concern or delight, depending on the reader, that I did not attempt to relate the information I derived from the literature review to specific conditions at the MWL. I have purposely not attempted to link results from the literature review to the MWL because, first of all, I do not have any certain answers about linkages, and secondly, it would require me to speculate about conditions at the MWL that cannot be verified with existing data. For example, there is nothing published that quantifies or characterizes the kinds of fauna and flora at the MWL or the amount and consequences of biological intrusion, subsidence, or soil erosion on contaminant distribution and transport.

Despite my unwillingness to be speculate about how physical and biological processes relate specifically to the MWL, I am willing to state unequivocally that all of the processes discussed in this report will, without question, effect the long term distribution and transport of MWL contaminants. I would assume that the NMED, EPA, and/or the owners of the landfill would have a vested interest in discovering how important these processes are in assuring the long term safety of workers and the public.

B. Hakonson's Technical Background And Expertise Relevant To This Review

Before proceeding, the reader should know something about my technical background and experience in order to place my review comments in context. I have 3 graduate degrees with science emphasis in Wildlife Biology (MS, 1964), Radiation Health Physics (MS, 1967) and Radioecology (PhD, 1972) from Colorado State University. Most of my professional life was spent as a research scientist and Group Leader in the Environmental Science Group at Los Alamos National Laboratory (1972-1993). My research at Los Alamos focused on two areas including the distribution and transport of plutonium and other radionuclides in liquid waste disposal areas at Los Alamos, in the fallout zone from the Trinity Site atomic bomb test in south-central New Mexico, and in nuclear safety shot areas in Plutonium Valley at Nevada Test Site. My radionuclide transport studies focused on hydrologic processes and especially the role of runoff and erosion in mobilizing Pu and other soil radionuclides in these arid/semiarid ecosystems.

My second area of expertise is on landfill cover alternatives. Beginning in 1980, my colleagues and I used fully instrumented field plots to measure hydrologic processes (i.e. water balance relationships, erosion, and contaminant transport) in RCRA and alternative cover profiles including evapotranspiration cap designs. We designed, constructed, monitored, and published results on landfill cover field demonstrations at Los Alamos, NM; Hill Air Force Base, Utah; and Marine Corp Base Hawaii, at Kaneohe, HI. These studies were predecessors to SNL/NM's current Alternative Landfill Cover Demonstration. I was nominated as a Laboratory Fellow in 1981 and received Los Alamos National Laboratory's Distinguished Performance Award for my work on landfill covers in 1982. Most of my publications since 1980 are on landfill cover research.

I retired from Los Alamos in 1993 and was appointed to the faculty at Colorado State University to develop an academic, training and research program for the University's Center for Ecological Risk Assessment and Management. I served as Director of the ERAM program from 1993-1996 and participated in a variety of human health and ecological risk assessments for industry, citizen groups, and DOE. From 1997-2001, I was an Associate Professor in the Radiological Health Sciences Department where I continued my research on landfill covers and actinide transport.

I served as a technical expert on various aspects of plutonium distribution and transport as a part of downwinder litigation at Hanford and to the International Atomic Energy Agency on closure of low and intermediate level radioactive waste sites. My most recent research investigated runoff and erosion as Pu transport mechanisms at the Waste Isolation Pilot Plant, Rocky Flats Environmental Technology Site, and Hanford as a component of a comprehensive, multi-organization study of the various modes of actinide transport in the environment. I am currently retired but consult with various government and public groups on a range of problems in radioecology, hydrology, ecology, and environmental restoration. I have over 120 publications on my research activities

C. Specific Topics Addressed In This Report

Citizen Action, a citizen activist group located in Albuquerque NM, commissioned me to perform a technical review of the proposed MWL evapotranspiration cap closure that addressed the following topics:

1. Potential migration of contaminants via abiotic/biotic pathways that may occur under context of an evapotranspiration soil cap combined with long-term stewardship to include:

a. Potential modes of contaminant transport from radiological/ non- radiological waste items buried in the landfill (physical, aqueous, vapor phase), and

b. Evaluation of plant, animal and human "intrusion potential" for long-term isolation of waste, which will consider the following: contaminant pathways via vegetation;
" Location of landfill to city and future development of area;
" Current and future intrusion by burrowing animals; and
" Potential for risk associated with abiotic/biotic means of transport.

2. Review of the evapotranspiration soil cap design submitted for approval by the NMED for final closure of the Mixed Waste Landfill (Environmental Restoration Project), and of a second cap design (Dwyer, Stormont, and Anderson, 1999) not selected by the NMED.

" Selection of type and depth of cover materials.
" Surface slope, runoff, and erosion control.
" Vegetation component planned for evapotranspiration cover and potential for future intrusion by other plant species.
" Post closure monitoring systems designed to detect air, soil, water, moisture content, and vadose zone contamination.
" Financial assurance plan for supporting post closure activities.
" Uncertainties associated with projected performance of an experimental cap.
" Engineering QA/QC to demonstrate cover is constructed as specified.

3. The study will also give the following recommendations regarding long-term
performance:

" Soil cap design, sediment mix, implementation of a biointrusion barrier, back up systems in case of failure;
" Establishment of an air monitoring program at the MWL;
" Sampling program to detect contaminants in vegetation;
" Erosion control.

This report provides discussion in three areas as follows:

(1) A section reviewing general concepts concerning landfill closures, capping design alternatives, failure modes, and long term performance (topic A items listed above). I want to focus heavily on this first section of the report to present the results of published research on landfill caps. My intent is to identify what is known and unknown about abiotic and biotic processes that effect cap design and function and to provide a basis for judging the merits of the SNL/NM proposed ET cap closure and post closure monitoring plan,

(2) A review of the two evapotranspiration cap design reports offered by SNL/NM for closure of the MWL (item B above), recognizing that the ER group closure plan has been selected by SNL/NM and has undergone scrutiny by NMED, and

(3) Recommendations as listed in item C above.

D. Technical Concepts And Issues

D1. General Considerations

Selecting a cleanup action for the MWL that adequately protects human health and the environment requires risk managers to synthesize and evaluate a large amount of technical, regulatory, and socio-economic information to arrive at an optimum, or best decision for closing the site. Applicable regulations require an assessment of the human health and environmental risks of candidate management alternatives prior to any corrective action (EPA, 1989a; Harwell, 1989; Bartell et al. 1992; Suter 1993). These risk assessments provide the fulcrum between science and policy. They are the interface at which predictive capability about ecological processes and contaminant kinetics can be applied to aid in resolving environmental problems and managing risk at sites such as the MWL.

While I am not going to discuss the MWL risk assessments conducted by SNL/NM and the various critiques of that assessment, I will say that, in my opinion, the risk assessments are the most important part of the decision process for selecting a closure option for the MWL as they are the only means for estimating future ability of the site to contain waste. These risk assessments must be based on reliable, quality data for conducting the assessment and for evaluating risk model output.

Depending on the level of risk to both humans and ecosystems from MWL contaminants, the alternatives for remediating the landfill can range from a no further action, in-situ / ex-situ soil treatment processes to remove selected contaminants, or removal of most of the contaminant inventory by excavation of the landfill. It is possible that all of these options could be candidates for the forthcoming Corrective Measures Study of the MWL.

The option finally selected for remediating the MWL should have a strong technical basis as derived from the health and environmental risk assessments. Properly done, the risk assessments should represent the best science available on the distribution and transport of the contaminants at the landfill under different closure alternatives in both the near and long term.

Whatever the eventual choice for closure of the MWL, it seems reasonable to demand that any use of public funds for remediation of the MWL landfill must be tied closely to the level of current and potential future risk. Cost will always be important criteria for selecting options for remediating sites such as the MWL. In general, the objective is to reduce costs to a minimum while convincing decision-makers that the potential risks from the site are acceptable under the assumed land use scenario. A comparison of unit costs for construction (i.e., O&M costs not included) of several capping alternatives at Los Alamos is compared to the cost of excavating the waste in Table 1. The message in Table 1 is that cost of remediation can vary by orders of magnitude depending on the alternative chosen and that none of the options are inexpensive.

The cost of the excavation alternative is estimated for a low level radioactive waste landfill at Los Alamos and for mixed waste based on the recently completed clean closure of the Area P landfill at Los Alamos (Bostick, pers. comm.). Cost for excavation and disposal of mixed waste is estimated to be about a factor of 20 higher ($10,000/yd3 vs. $500/yd3) than for low-level waste. Costs for in-situ and ex-situ soil treatment alternatives, which are not shown in Table 1, are also of the same order as the excavation options. Actual costs of any of these options will depend on local conditions, waste types and amounts, and re-disposal options.

Arguments about O&M costs of one closure option versus another are problematic in my opinion since even the excavation option will very likely require some long term inspection, maintenance, and/or monitoring. As mentioned, attempts to remove the waste from the MWL will not be 100% efficient (e. g., H3). The O & M costs for a particular closure option will depend on several factors including the type of waste in the landfill, landfill location and climate, and required measurement frequency for inspections and monitoring.

None of the options in Table 1 are risk free including actions to remove the waste from the landfill. Consequently, risk managers must decide whether near term risks associated with waste removal are more or less acceptable than potential future risks resulting from other closure options. Further complicating the issue is that near term risks likely have less uncertainty associated with them compared to more uncertain predicted future risks.

D 2. Regulatory Requirements

Cover closures of sanitary (there are about 226,000 in the U.S.), and radioactive and hazardous waste landfills (there are a few thousand) are pervasive in the U.S. for a variety of reasons, including some that are valid and some not so valid. There are cases where old (pre-RCRA) sanitary and radioactive waste landfills have leaked for a variety of reasons including poor siting of the landfill, high local precipitation, and/or inadequate attention to cover design (Duguid, 1977; Jacobs et al., 1980; EPA, 1988). EPA (1988) details some of the containment problems in a survey of 163 randomly chosen sanitary landfills. Containment problems of various degrees had cropped up at most of these sites and about 25% of those required near term corrective actions.

Many of these old pre-RCRA landfills used similar disposal methods in that a trench or pit was dug into the soil, waste was placed into the excavation w/ or w/o backfill, and when full, the landfill was covered with soil (Duguid, 1977; Jacobs et al., 1980). Sometimes the landfill was reseeded and sometimes it was not. This general approach to waste disposal dates back thousands of years (Langer, 1968). Requirements of current regulations such as RCRA and CERCLA now stipulate methods for remediating old landfills in considerable detail and for constructing, operating, and closing new landfills.

Table 1. Estimated costs of remediation alternatives for landfills.
Alternative1 Cost/Unit2
Excavate LLW Landfill $80M/Ha
Excavate MW Landfill $ 1,600M/Ha
RCRA Cap $4.9M/Ha
Soil/Capillary Barrier Cover $1.5-3.7M/Ha
Infiltration Control w/ Vegetation $0.24-0.5 M/Ha
ET Cap with Erosion Control $0.12-0.8M/Ha

1Technical basis for selecting remediation alternatives should be based on human and ecological risk assessments.
2Costs for the LLW excavation option (assume excavation to depth of 8 m) adapted from DOE's Environmental Restoration program 5-year plan. Costs of excavation of mixed waste based on actual costs for recently completed Area P clean closure at Los Alamos. The RCRA cap costs are from the Maxey Flats Kentucky Corrective Measures Study. Costs for the remaining capping options estimated by author based on installation costs of field demonstrations including SNL/NM Advanced Landfill Cover Demonstration (Dwyer, 1998). Exact costs for a particular option will be site specific.

Based on many years of research and observations on landfill operations and closures, I am of the opinion that most of the "problems" that crop up at old landfills resulted from the period when active waste disposal was occurring. During this period, disposal trenches were open and lacked any kind of mechanism for removing precipitation that fell on the landfill. Because these old landfills were usually active for many years, the potential existed for very large amounts of precipitation to enter the landfill, move downward through the waste, and, depending on the permeability of the surrounding soil, to move out of the bottom of the landfill. Furthermore, the amount of precipitation entering the disposal units each year of operation could approach the annual precipitation. I further believe that observed problems with contaminant migration from landfills in arid sites occurred prior to closure of the landfill with a vegetated cover. Failure to vegetate soil covers after landfill closures was an especially bad practice because this major soil moisture removal mechanism, (i. e. transpiration) was lacking.

The regulatory requirements for closure of landfills with caps are detailed in several EPA guidance documents (EPA, 1979; 1982, 1985, 1989b and c, 2002). The regulations basically require the owner/operator of a landfill to perform landfill closures. The primary closure requirement is that the owner/operator must design and construct a low-permeability cover over the landfill to minimize migration of liquids into the waste and to provide for post-closure monitoring and maintenance in order to prevent unacceptable future waste migration into the environment.

EPA guidance to permit applicants also recommends that an analysis of the final cover design be presented in the closure plan. The analysis of the final cover design must describe how the design meets the following performance criteria:

1. Minimizes liquid migration,
2. Promotes drainage while controlling erosion,
3. Minimizes maintenance,
4. Has a permeability equal to or less than the permeability of natural subsoil,
5. Accounts for freeze/thaw effects, and
6. Accommodates settling and subsidence so that the covers integrity is maintained.

It should be emphasized that these general performance standards allow flexibility in the cover design proposed for the site (EPA, 1989b, c). This flexibility has now being formalized in a new guidance document soon to be released by EPA (EPA, 2002).

In order to demonstrate that a proposed final cover design complies with the regulatory performance standards, EPA states that it is necessary to model water balance and erosion on the proposed cover. EPA also suggests that the water balance model, HELP, be used for the water balance calculations (User's Guide for the Hydrologic Evaluation of Landfill Performance (HELP), version 3, model (EPA, 1994).

The numbers of landfill cover designs that can be evaluated with HELP are limited to a couple of versions of the RCRA cap design and monolithic soil cap designs. The HELP model does not give accurate predictions of water balance in a particular cap design (EPA, 1994). It is intended as a screening tool to provide comparative response between design alternatives rather than realistic numbers.

Because the HELP model cannot handle unsaturated soil moisture conditions, cap designers sometimes use other models to supplement the HELP modeling results. Both SNL/NM ET cover design groups did this. The use of alternative models is an exercise in futility in my opinion because seldom is there enough data to initialize these models and nor is there enough relevant data to compare with model predictions. Consequently, without such data, it is a far stretch to say that one of the "better" models predicts water balance any better than HELP.

EPA also recommends the use of an empirical formula called the Universal Soil Loss Equation (USLE) to calculate annual average soil loss from the proposed cover design (EPA, 1989b,and c). The average annual soil loss is predicted based upon a number of factors including the geographical location, the length and steepness of slopes, the texture of the cover soil, and the vegetation cover.

More recent erosion prediction models on erosion developed by the U.S. Department of Agriculture, Agricultural Research Service (USDA-ARS) are sometimes used to evaluate erosion on landfill caps. These models, which were not used by SNL/NM, include the Revised Universal Soil Loss Equation (RUSLE) (Renard, 1985) and Chemicals, Runoff, and Erosion in Agricultural Management Systems (CREAMS) (Knisel, 1980). Predictions with erosion models suffer from the same problems as water balance models. Good quality data are needed to initialize the models and validate predictions. All of these models come with default parameters and default databases for use in setting up the model runs. EPA (1994) states that site-specific data for initializing the model are much preferred.

As long as these models are used as screening tools, they will be useful in comparing cap designs. However, if the model output is used to specify exact numbers (and they are believed), then there must be an independent means of confirming model predictions. Very seldom do such data exist for a site like the MWL.

D 3. Regulatory Basis for Alternative Cap Designs

A range of cover designs representing various complexities and costs have been proposed for landfill closures (Hakonson, 1997a). Those designs range from multi-layered engineered barriers, such as EPA's clay cap designs (EPA, 1989b, c), to simple vegetated soil caps that rely primarily on evapotranspiration to manage site water balance (Anderson et al. 1997, Hauser, 1994; Hakonson et. al., 1997a, b, Dwyer et. al., 2001).

An alternative cap design is acceptable for radioactive and hazardous waste landfill closure (EPA 1989a, b, 2002) as long as it's performance can be demonstrated to be equivalent to the RCRA clay cap designs EPA (1989a, b). Because experimental data demonstrating equivalency are generally unavailable, site operators and regulators in the past have been reluctant to use or approve alternative designs. However, based upon the results of field demonstrations over the last few years, EPA now recognizes that alternative cap designs are sometimes preferable to RCRA clay barrier designs especially in arid sites where clay barriers are subject to desiccation cracking (EPA, 1989 a, b).

As a consequence of this now formalized latitude in selecting cover designs for landfills, there has been renewed interest in vegetated soil caps as an alternative for landfill closures. Modeling studies (Hauser et. al., 1994; Khire et al, 1995) and limited experimental data (Anderson, 1993, Dwyer, 1997, Dwyer et. al., 2001) suggest that vegetated soil covers, or ET caps, can be effective in controlling site water balance, particularly in arid and semiarid locations.

There are several possible cover designs that could be considered for the MWL besides the ET cap and it is possible that one of these options may eventually be agreed upon for closure of the site. Designs options include those that incorporate engineered barriers such as EPA's RCRA C cap with it's clay barrier or RCRA-GCL (geosynthetic clay liner) designs (EPA, 1989). Other barrier design possibilities include capillary barriers (Nyhan et. al., 1990b Nyhan et. al., 1997a, b; Stormont and Morris, 1998) and combinations of several barrier technologies (UMTRA-DOE. 1989; Buckmaster 1993).

EPA's technical guidance for final covers describes what has been called the RCRA C cover that they believe will meet the final cover performance standards for hazardous and mixed waste sites. The RCRA C cover design is composed of three layers which can be configured for soil water management:

1. An uppermost-vegetated layer to prevent erosion and promote evapotranspiration,
2. Underlying drainage layer/s to convey percolation out of the cover, and
3. A moisture barrier/s to prevent percolation through the cover.

The functional requirements of the topsoil layer are to promote runoff from major storms but restrict erosion rates to acceptable limits. While runoff can promote erosion and degradation of the cover system, vegetation and/or other surface treatments impede erosion and support the long-term integrity of the cover. The topsoil must also be thick enough to provide storage capacity for soil moisture for removal by evapotranspiration and to protect the hydraulic barrier layer from freezing and desiccation. EPA states that typically, a thickness of 60 to 90 cm (2 to 3 ft) is sufficient, but the actual requirement is climate, soil, and design specific.

EPA also states that drainage layers are not necessary at all sites since some sites may not have sufficient rainfall and infiltration to produce a standing head of water on the hydraulic barrier for long periods. If used, the drainage layer should be designed to reduce leakage through the hydraulic barrier layer by lowering the hydraulic head.

Capillary barriers have been investigated extensively as components of landfill cap designs since 1980 (Nyhan et al. 1990). Conceptually, they consist of two layers with widely differing hydraulic conductivities. Generally they are comprised of a fine-grained soil or sand overlying a coarse-grained gravel or rock.

Capillary barriers function as a moisture barrier by impeding flow at the interface between the two layers. Water moving downward through the soil encounters the coarse gravel layer, where capillary forces in the overlying soil prevent breakthrough of the water into the gravel at soil moisture conditions less than saturation. When saturation of the soil occurs, water breaks through the capillary barrier. By placing an angle on the interface between the soil and gravel layers, lateral flow can occur, reducing the potential for saturation and barrier failure by preventing build up of hydraulic head. Selection of materials for the capillary barrier system is critical in that rapid lateral flow is essential to prevent the buildup of water as it drains laterally downslope.

D 4. Relationship of Water Balance to Cap Design

The fate of meteoric water falling on the surface of a landfill is often referred to as the water balance of the site. A simplified representation of water balance describes surface runoff and one-dimensional movement of water in the soil profile to the plant rooting depth. For net rates and amounts, the water balance equation is:

dS/dt = (P - Q - ET - L)/dt (Equation 1)

Where dS/dt is the time rate of change in soil moisture, P is the precipitation per unit area, Q is the runoff per unit area, ET is the evapotranspiration per unit area, L is the percolation below the root zone per unit area, and t is the unit of time used in solving the equation.

Application of the concept of water balance to design of landfill caps, including the ET cap design, takes advantage of the fact that there are strong interactions between the various components of water balance (Hakonson et. al., 1982a). For example, a reduction or elimination of the runoff term (Q) must be accompanied by increased infiltration of water into the soil. Increased infiltration results in increased soil moisture storage, which is then followed, by an increase in ET and/or percolation. Conversely, reducing percolation necessitates that more of the precipitation be partitioned between soil moisture storage, ET, and/or runoff.

The coupled nature of the processes comprising the water balance can be used to advantage in designing a caps that minimize or eliminate hydrologic processes that contribute to contaminant migration (i.e. percolation (L) in Equation 1) while enhancing other terms (i.e. ET) that reduce the potential for aqueous phase contaminant transport (Nyhan et al., 1990b; Hakonson, 1997a, b; Hakonson et al., 1993; Lane, 1984a; Lane and Nyhan, 1984b; Dwyer, 1997, 2001; Anderson et. al., 1993).

D 5. ET Cap Design

A common misconception about ET caps is that they are unique and different from other cap designs. In fact, virtually any cap design that contains a vegetated soil layer functions primarily as an ET cap regardless of the number, type, and placement of engineered barriers within it. This means that most of the precipitation that infiltrates in the cap soil is removed by the vegetation cover via evapotranspiration.

For example, from about 90-100% of the precipitation that infiltrates into the vegetated top soil in environments receiving <50 cm, is removed by evapotranspiration (Anderson et. al., 1993; Hakonson, 1998; Hakonson et. al., 1990; Lane, 1984a,b; Nyhan et. al., 1990b, Nyhan et. al., 1997b; Dwyer, 2001). This also applies to the RCRA cap where most of the function in controlling percolation downward into the soil is due to the ET component of the design and not to the clay barrier (Dwyer, 2001; Warren et al, 1996 a, b).

In the absence of engineered barriers, a well-designed ET cap consists of a single, vegetated soil layer constructed to represent an optimum combination of soil hydraulic properties, soil thickness, surface slope, and vegetation cover (Hakonson, 1997a; Figure 1). The conceptual framework for the ET landfill cap design was developed from several short-term studies (i. e. studies of a few years duration) in the 1980's (Hakonson et. al., 1982a; Hakonson et. al., 1986a; Nyhan and Lane, 1986, 1987; Nyhan et. al., 1990b). Those studies, which used instrumented field plots in arid and semiarid environments, demonstrated that simple but well designed soil covers were very effective (i.e. as effective as EPA's RCRA designs) in preventing excessive runoff, erosion, and percolation of water through the cover. Those studies also demonstrated that the vegetation component of the cover served as the principal mechanism inhibiting soil moisture movement through the cover and into the waste.

Figure 1. Conceptual ET cap representing optimal combination of soil type, soil depth, surface slope, vegetation type and density, and surface management practices such as gravel mulching to control erosion.

Ideally, the vegetation cover should consist of an optimum mixture of native species that represent late successional stages, including cool and warm season grasses and possibly species with different growth forms (such as grasses, shrubs, or trees). The intent of the vegetation cover is to provide long-term protection against soil erosion, resistance to catastrophic events such as fire and drought, and to fully utilize precipitation that falls on the site by spreading ET over as much of the year as possible.

For a site such as the MWL, the vegetation growing on undisturbed Upper Sonoran rangelands in the Albuquerque area would best represent a species mix that would be stable over the long-term. Shrubs should not be discounted in the species mix selected for the MWL. As will be discussed later, concerns about shrub root penetration into the MWL miss the point in that any species or mix of species planted on the MWL has the potential to penetrate into the waste environment. Conceptually, the goal is to capture all of the moisture before it gets to the waste. If that can be achieved under actual precipitation rates at the MWL then plant roots will remain in the cover soil.

Successional processes will most likely occur when the vegetation consists of early successional species or non-native species. Therefore, the optimal cover for maximizing ET over the long term is likely to be the one maximizing both Leaf Area Index (LAI) and diversity among life-forms, and species that have evolved in the area.

D 6. Causes Of ET Cap Failure

D 6.1 Physical Transport Processes

D 6.1.1 Landfill Surface- Under ideal conditions, the primary functions of any cap design are to isolate the buried waste from the surface and ground water environment by controlling abiotic and biotic processes that can contribute to contaminant migration from the site (Hakonson et al., 1992a). Hydrologic and erosion processes account for most of the performance-related problems at LLW and sanitary landfills (Duguid, 1977; Jacobs et al., 1980; EPA, 1988).

For example, erosion associated with runoff can breach the cap and expose waste to the biosphere (Nyhan and Lane, 1982; Nyhan et al., 1984). Erosion rates must be within tolerances to leave the cap intact over the mandated lifetime of the facility. EPA recommends a soil Tolerance Level (TL) of 2 t/ac/yr (4.4T/ha/yr) in order not to deflate the cover surface over the lifetime of the site. Tolerance Level is an arbitrary amount of erosion (i. e., specific to site conditions) that is about equal to the rate of soil formation. The goal is to ensure that erosion rates are at or below the level where the soil surface begins to deflate with time.

Changes in erosion rates on a landfill cover can occur and may be associated with changes in vegetation biomass, animal species, plant and animal species composition, and disturbances such as fire or drought. Cover design features that are used to prevent erosion include the establishment of vegetation, the use of mulching techniques, synthetic mats, control of the slope and slope length, and the construction of terraces or benches.

Preventing buried waste from reaching the ground surface is important because once contamination on the soil surface, it is subject to transport by wind and water erosion processes (Hakonson et. al., 1981, Hakonson and Lane, 1992b; Lane and Hakonson, 1982). Erosion of contaminated soil can lead to transport of contaminants off the landfill by both physical and biological processes.

The importance of wind and water in transporting soil contaminants results from the fact that many contaminants are nearly quantitatively deposited in soil and are tightly bound to soil particles (Hanson, 1975; Ritchie and McHenry, 1990; Whicker and Schultz, 1982). Furthermore, concentrations of most contaminants in soils are a strong function of soil particle size. Generally, the silt and clay size fractions (< 50µm diameter) contain up to a factor of 10 higher concentrations than the bulk soil. This is especially true of most radionuclides such as cesium, plutonium and strontium (Watters et. al., 1980). Consequently, processes which transport soil also transport soil-associated contaminants.

In many cases, both wind and water preferentially detach and transport the finer size fractions that often contain the highest concentration of radionuclides and other contaminants. Moreover, the finer soil fractions are carried farther (and deposited later) than coarser fractions of eroding soil (Lane and Hakonson, 1986).

Recent studies were conducted at the Waste Isolation Pilot Plant, Rocky Flats Environmental Technology Site, and Los Alamos National Laboratory to evaluate the effects of fire on transport of fallout Cs-137 by wind and water erosion (Johansen, et. al., 2001a; Johansen et. al., 2001b). Those studies, which used a large rainfall simulator to generate runoff and applied controlled or natural fires as study treatments, found that soil erosion and soil associated fallout cesium-137 transport on burned plots increased by a factor of 4-25 over unburned plots depending on the study area.
Results demonstrated that erosion and soil radionuclide transport rates determined under undisturbed conditions do not reflect possible changes in rates due to a catastrophic disturbance such as fire. Without intense post fire management, such as reseeding, the effects of fire on water erosion can persist for years (Simanton and Emmerich, 1994).

The combined phases of runoff, soil erosion, sediment transport, and deposition of sediment on upland areas and in stream channels usually result in enrichment of smaller soil particles and organic matter in transported sediment (Graf, 1971) including concentrations of sediment associated contaminants (Massey and Jackson, 1952; Menzel, 1980; Lane and Hakonson, 1982). This enrichment is often expressed as an enrichment ratio, defined as the concentration of contaminant in the transported sediment divided by its concentration in the un-eroded soil.

Enrichment ratios have been related to sediment concentration, sediment discharge rate, and sediment yield (Massey and Jackson, 1952; Menzel, 1980). Lane and Hakonson (1982) analyzed sediment transport rates by particle size classes in alluvial channels and derived the following expression:

*[Cs (di) x Qs (di)] (1)
ER = -------
Cs *[Qs (di)]

Where:
ER = alluvial channel enrichment ratio,
Cs (di) = Concentration of contaminant in sediment particles of size
class i, with representative diameter, di, in millimeters.
Qs (di) = Sediment transport (mass/time) for particles in size class i,
with representative diameter, di, in millimeters.
Cs = Mean concentration of contaminant in soil over all particle
size classes.

Equation 1 supports the empirical observation that enrichment ratio increases with decreasing sediment discharge rates. For example, at very low sediment discharge rates (those associated with low runoff velocities) the bed load discharge rate for coarse sediment particles is low and most of the transported sediment is in the smaller particle size classes. Under such conditions, ER would approach the ratio of concentrations in the finest size classes (Cs (di)) to the mean concentration over all size classes (Cs).

At high sediment discharge rates (those associated with high runoff velocities) more of the bed sediments are in transport. At the extreme, if all of the bed sediments were in transport in the same proportion, as they exist in the bed material, ER in Equation 1 would be unity.

Field measurements of enrichment ratios for nutrients and plutonium at several locations in the United States are listed in Table 2. The first four entries are for soil nutrients in runoff from small agricultural areas; mean values vary from 2.6 to 7.1. The next three entries represent enrichment of fallout plutonium in runoff from agricultural watersheds; mean values range from about 1.6 - 2.5. The last entry represents enrichment of plutonium in runoff in ephemeral stream channels at Los Alamos, New Mexico.

Enrichment ratios based on measurements in runoff in canyons at Los Alamos ranged from 1.4 to 13.3 with a mean of 5.5. Predicted enrichment ratios for Los Alamos stream channels [Eq. (1)] ranged from 2.9 to 7.0 with a mean of 5.2 (Lane and Hakonson, 1982). The close agreement between observed and predicted enrichment ratios suggests that particle sorting alone can account for ratios observed at Los Alamos.

In spite of wide differences in watershed size, hydrologic regime and chemical characteristics, the enrichment ratios resulting from sediment transport given in Table 2 are quite similar for several sediment-associated chemicals. Particle sorting is clearly one of the important factors involved in transport of soil and sediment associated contaminants.
Enrichment processes can lead to higher concentrations of radionuclides in sediment deposition areas. For example, fallout cesium and plutonium concentrations in reservoir sediments have been shown to be higher than concentrations in soils from upland areas that contributed the sediments to the reservoirs (Sprugel and Bartelt, 1978; Muller et. al., 1978). This is because the small soil particle sizes that usually contain highest concentrations of radionuclides are easily detached and transported to downstream areas and they are last to settle out of the water column as flow velocities decrease.

Soil resuspension by wind, rain splash, and mechanical processes also undergoes particle sorting. For example, rain splash and/or wind resuspends soil particles and deposits them on vegetation surfaces (Dreicer et al., 1984; Foster et al., 1985), and animals (Romney and Wallace, 1977; Hakonson and Nyhan, 1980).

Table 2. Approximate Enrichment Ratios for Nutrients and Plutonium Associated with Locations in the United States.

Land use Approximate Comments
and location enrichment ratios
Mean Range
a
Cropland, USA 4.5 2.5 - 7.4 Nitrogen
3.6 2.6 - 6.0 Phosphorus

a
Rangeland, USA 2.6 1.1 - 6.7 Nitrogen
7.1 2.7 - 17 Phosphorus

b
Cropland, USA 1.6 1.1 - 2.5 Fallout Plutonium

b
Pasture, USA 2.3 0.8 - 4.0 Fallout Plutonium

c
Mixed Cropland, USA 2.5 1.2 - 4.0 Fallout Plutonium,
Transport in Perennial River
d
Semiarid, USA 5.5 1.4 - 13.3 Waste Effluent Plutonium Transport in Ephemeral Streams
___________________________________________________________

aSmall agricultural watersheds (5.2 - 18 ha) at Chickasha, Oklahoma.

bSmall agricultural watersheds (2.6 - 2.9 ha) near Lebanon, Ohio.

cGreat Miami River (Drainage area = 1401 km2 ) at Sidney, Ohio.

dLos Alamos Watersheds (176 - 15,000 ha) near Los Alamos, New Mexico.

Field studies with plutonium (Nyhan, 1980; Watters et al., 1980; Romney and Wallace, 1977) show that physical deposition of contaminated soil particles on vegetation surfaces is 100-1000 times more important than root uptake as a means of contaminating vegetation with this radionuclide.
Likewise, most of the body burdens of plutonium are associated with the GI tract and pelt of small mammals living in contaminated areas. The association of plutonium with those tissues implies that soil ingestion and grooming activities are important mechanisms for contaminating animals.

The importance of physical transport of MWL contaminants will ultimately depend on the effectiveness of the ET cap in preventing transport of buried contaminants to the ground surface. Should transport of buried waste to the landfill surface occur, then transport by wind and water erosion will also occur. Ultimately, the fate of soil contaminants once they are present on the ground surface will depend on the fate of the soils and sediments themselves.
In an attempt to predict the amounts of soil lost under different regimes of climate, soil, topography, and land management, a number of approaches have been taken. Wischmeier and Smith (1960) developed the Universal Soil Loss Equation (USLE), a tool for predicting soil loss. The USLE combines known factor values for rainfall erosivity, soil erodibility, slope length and steepness, cover/management, and supporting practices to estimate soil loss values for use in environmental planning and assessment. Others have gone a step further by combining equations for fundamental hydrologic and erosion processes into models to predict soil losses and losses of pollutants including agricultural chemicals. One of these efforts includes that of the U.S. Department of Agriculture's development of CREAMS (Chemicals, Runoff, and Erosion from Agricultural Management Systems), a model capable of evaluating runoff, erosion, and chemical transport from field-sized areas (Knisel, 1980).

D 6.1.2 Subsurface Processes- Depending on climate, geology and soil conditions, water that infiltrates into and through the cap on old landfills can accumulate in the trench (bathtub effect) and/or percolate with solutes into groundwater. Percolation can also increase subsidence of the cap as a result of enhanced decomposition of bulky waste in the trench. Subsidence may occur some variable time after closure of the land disposal unit and after final placement of the cover. Although subsidence has the potential to seriously damage a landfill cover, predicting subsidence and subsidence effects is very difficult because of the heterogeneous nature of the waste forms, backfill materials, and variation in local climatic conditions.

In landfills that contain volatile contaminants, movement might also involve both aqueous and vapor phase transport such as has been observed for tritium at the MWL. Controlling aqueous transport of volatile contaminants does not necessarily control vapor phase transport. In fact, maintaining low soil moisture content of cover and backfill soils to reduce aqueous phase transport may be associated with increases in vapor phase transport of volatile contaminants (Jury, 1987). This is likely what occurs at the MWL regarding tritium. Vapor phase transport may also be more pronounced near the ground surface where changes in soil barometric pressure, rapid wetting and drying of soil, and plant root biomass and animal burrowing leading to macropore formation are greatest.

In climates that experience freezing temperatures or drought, the cover, including the soil layer and man-made construction materials, may be effected by the freezing or drying of the soil. Frost heaving and desiccation cracking is common in soils with high clay content. However, these soil surface disturbances are reduced when vegetation cover is present.

Freezing temperatures and drought can disrupt the integrity of barrier layers and freezing can also increase the amount of runoff when frozen ground limits infiltration of moisture into the soil. Desiccation of clay soils, such as hydraulic barriers, is a potentially important problem in arid sites (Suter et al., 1993). Because of potential problems with desiccation, EPA (1989) notes that alternatives to clay hydraulic barriers should be considered for sites with a high risk of barrier desiccation.

D 6.2 Biological Processes

Biological processes associated with the cap include plant root and burrowing animal intrusion into cover soil and potentially into the underlying waste. Penetration of plant roots and animals to the waste may contribute to migration of contaminants from the burial environment by both biological and physical/mechanical processes (Klepper et al., 1979; O'Farrell and Gilbert, 1975; Winsor and Whicker, 1980; and Arthur and Markham, 1983; Arthur and Markham, 1982, 1987; Arthur and Janke, 1986). Both plants and animals also affect physical processes such as water balance and erosion. Plant and animal related processes in landfill covers are closely coupled so that a response in one elicits a response in the other.

D 6.2.1 Plant Related Processes- Although vegetation is very important in controlling erosion and percolation in landfill covers (Nyhan et al., 1984), deeply penetrating plant roots have the potential to access buried waste and bring plant available constituents including landfill contaminants to the surface of the site (Klepper et al., 1979; Foxx et al., 1984; Tierney et al., 1987).

Contaminants such as tritium can be incorporated within plant tissue and enter the food web of herbivorous or nectivorous organisms. For example, at Los Alamos National Laboratory tritium transport away from a controlled low-level waste site occurred via the soil moisture/plant nectar/honey bee/ honey pathway (Hakonson and Bostick, 1976).

As another example, deep-rooted Russian Thistle (Salsola kali) growing over the waste burial cribs at Hanford penetrated into the waste, mobilized 90Sr, and then transferred it to the ground surface. The contaminated surface foliage was transferred away from the cribs when the matured Thistle (tumbleweeds) blew away from the site (Klepper et al., 1979).

Two mechanisms for soil contaminant transport to terrestrial plants are absorption by roots and deposition of contaminated soil particles on foliage surfaces. Field studies suggest that deposition of soil particles on foliage surfaces is a major transport mechanism for soil associated contaminants under many arid site and contaminant source conditions (Romney and Wallace, 1976; Romney et al., 1987; White et al., 1981; Arthur and Alldredge, 1982).

Table 3. Comparison of plutonium concentration ratios for field and glasshouse conditions (Romney and Wallace, 1976).
______________________________________________ ___
Soil Source Field Glasshouse
a -2 -1 -4
NTS Area 11B 1.3 x 10 to 1.6 x 10 1.5 x 10

-2 -1 -4
NTS Area 11C 4.5 x 10 to 3.4 x 10 1.8 x 10

-2 -1 -4
NTS Area 13 7.8 x 10 to 4.4 x 10 1.1 x 10

aNTS (Nevada Test Site)

Comparative studies of plant uptake of plutonium under both field and laboratory conditions generally yield results of the type shown in Table 3. The results of laboratory studies represent root uptake of plutonium from soils and yield concentration ratios that are at least one order of magnitude (and often 2-3 orders of magnitude) lower than ratios observed under comparable conditions at field sites.

The differences in concentration ratios between laboratory and field studies implies that a mechanisms exists in arid environments for delivering at least 10 times more plutonium to vegetation than would be predicted based upon root uptake as measured in greenhouse studies.

The higher ratios observed at field sites have been attributed to the presence of surficial contamination on vegetation (Romney et al., 1987; Hakonson and Nyhan, 1980; Little et al., 1980). That conclusion is supported by the obvious presence of soil on foliage surfaces in the field and by the ability to remove up to 90% of the plutonium contamination from vegetation by washing (White et al., 1981; Arthur and Alldredge, 1982).

Studies at Los Alamos demonstrated that rain-splash of soil particles with subsequent deposition on foliage surfaces can easily contribute all of the plutonium measured in field-site vegetation (Dreicer et al., 1984). More importantly, those studies, which employed a labeled-soil particle technique and the scanning electron microscope, have shown that relationships that govern lateral movement of plutonium by soil erosion processes also govern transport of plutonium to foliage surfaces.

For example, the energy of impacting raindrops caused an enrichment of the smaller soil particles on foliage surfaces. The amount of soil deposited on the plants was also related to height from the ground surface and characteristics of the rainstorms. Calculations based on the mass and plutonium content of soil measured on the plants demonstrated that the rainsplash mechanism could easily account for the high concentration ratios observed in field samples (White et al., 1981; Foster et al., 1985).

While absorption of soluble forms of plutonium through leaf surfaces has been demonstrated (Cataldo and Vaughn, 1980) it is likely to be of limited importance in arid field sites because environmental plutonium exists as an oxide and is very tightly bound to soil. Studies on the uptake of plutonium by vegetable crops grown in field sites at Los Alamos show that as much as 50% of the plutonium in edible crop samples was surficial contamination that could be removed by washing (White et al., 1981) or peeling.

The remainder that could not be removed was associated with very fine soil particles adhered to the vegetable surfaces as determined by the electron microscope. Cataldo and Vaughn (1980) and White et al. (1981) showed that submicron particles on foliage surfaces are difficult to remove by simulated wind, rain, or household food washing procedures. While a large number of the published studies on environmental fate and effects deal with plutonium, physical processes will also control the environmental behavior of many other radionuclides.

D 6.2.2 Root Distribution in Soil- Root distribution in the soil profile is strongly related to the depth of water penetration into the soil (Canadell et al, 1996; Jackson et al. 1996). Although average and maximal reported rooting depths vary with species and life form, there is a great deal of plasticity within most species to respond to variation in soil water availability. Hence, if water is available at deeper depths, roots of a species viewed as "shallow rooted" may occur there. For example, in a semiarid ecosystem in New Mexico, plant roots of a number of species have been observed to depths of at least a few meters in the pursuit of soil moisture (Foxx et al., 1984; Tierney et al., 1987). Alfalfa roots have been found over 40 m below the ground surface (Foxx et al., 1984).

If the root structure of certain species is confined to the upper few centimeters of the soil profile, it is largely because that is where most of the soil moisture is captured by the plants and removed from the soil. If moisture becomes available at deeper depths, most species have the potential to exploit this moisture by sending roots downward to capture available moisture, often to depths greater than previously recognized (Canadell et al. 1996). In normal situations where multiple species co-exist on a site,one species may exploit moisture near the ground surface while another exploits moisture deeper in the soil profile (Evans and Ehleringer, 1994, Golluscio et al. 1998, Breshears and Barnes, 1999).

Canadell et al. (1996) summarized what was known about the maximum rooting depth of species belonging to the major terrestrial biomes. They found 290 observations of maximum rooting depth in the literature that covered 253 woody and herbaceous species. Maximum rooting depth ranged from 0.3 m for some tundra species to 68 m for Boscia albitrunca in the central Kalahari; 194 species had roots at least 2 m deep, 50 species had roots at a depth of 5 m or more, and 22 species had roots as deep as 10 m or more. The average for the globe was 4.6+/-0.5 m.

Maximum rooting depth by biome was 2.0+/-0.3 m for boreal forest, 2.1+/-0.2 m for cropland, 9.5+/-2.4 m for desert, 5.2+/-0.8 m for sclerophyllous shrubland and forest, 3.9+/-0.4 m for temperate coniferous forest, 2.9+/-0.2 m for temperate deciduous forest, 2.6+/-0.2 m for temperate grassland, 3.7+/-0.5 m for tropical deciduous forest, 7.3+/-2.8 m for tropical evergreen forest, 15.0+/-5.4 m for tropical grassland/savanna, and 0.5+/-0.1 m for tundra.

Grouping all the species across biomes (except croplands) by three basic functional groups: trees, shrubs, and herbaceous plants, the maximum rooting depth was 7.0+/-1.2 m for trees, 5.1+/-0.8 m for shrubs, and 2.6+/-0.1 m for herbaceous plants. The mixture of grasses that SNL/NM intends to use in reseeding the MWL is lumped within the herbaceous plant category.

These data show that deep root habits are quite common in woody and herbaceous species across most of the terrestrial biomes, far deeper than the traditional view has held up to now. The implications for the MWL are that no matter what vegetation is planted on the landfill, if moisture penetrates beneath the ET cover, roots can be expected to follow.

D 6.2.3 Animal Related Processes

D 6.2.3.1 Uptake- As with vegetation, the resuspension of soil particles can be a major source of contaminants to animals living in arid ecosystems. Soil particles can be transported to animals in association with exterior surfaces of food and by direct transfer of soil to the animal via inhalation, ingestion and contamination of the pelt (Hakonson and Lane, 1992b).

Plutonium is the best example of a radionuclide whose transport to animals in arid ecosystems is dominated by physical processes. Data from many field sites and source conditions show that gut availability of plutonium and other contaminants bound to soil in a variety of animals including rodents, deer and cattle is very low (gut to blood transfer <10-5) leading to very low concentrations of contaminant in internal tissues and organs (Smith, 1977; Moore et al., 1977; Hakonson and Nyhan, 1980; Arthur et al., 1987; Romney et al., 1970).

Highest concentrations of most soil contaminants in dry, dusty environments are usually found in tissues exposed to the external environment. Those tissues include the pelt, gastro-intestinal tract, and lungs. At Los Alamos, about 96% of the plutonium body burden in rodents from the canyon liquid waste disposal areas was in the pelt and gastro-intestinal tract (Hakonson and Nyhan, 1980).

Because soil passes through the gastro-intestinal tract of free-ranging animals on a daily basis, there is a potential to redistribute soil radionuclides across the landscape. Studies at Nevada Test Site with cattle (Moore et al., 1977), at Rocky Flats Plant with mule deer and small mammals (Little, 1980; Arthur, 1979), and at Idaho National Engineering Laboratory with small mammals and coyotes (Arthur and Markham, 1983; Arthur et al., 1980) demonstrate that horizontal (and vertical in the case of burrowing animals) redistribution of soil plutonium does occur as animals move within and outside contaminated areas. However, the magnitude of this transport was shown to be very small over the short-term (Arthur, 1979); Arthur and Markham, 1983; Arthur et al., 1980).

There are circumstances where animal transport of soil contaminants can assume more importance. For example, fission product sludge containing Sr90 and Cs137 in salt form was released to unlined cribs at Hanford and the cribs were backfilled with clean soil. A large animal, probably a coyote or badger then burrowed down to the sludge and created direct access for other animals seeking the salts including jackrabbits (O'Farrell and Gilbert, 1975). Jackrabbits ingested the radioactive salts, became contaminated and then excreted 90Sr on the ground surface. Levels of 90Sr in excreta were found over 15 km2 (O'Farrell and Gilbert, 1975). I would emphasize that this incident with 90Sr and jackrabbits was a special case that involved liquid waste sludge disposal trenches that were not adequately covered.

Potentially more soluble strontium and cesium transport to animals in arid ecosystems involves a combination of physical and physiological processes. The more tightly bound these radionuclides are to soil (related to clay content of soil and local climate), the more their transport will be governed by soil particle transport. Data on Sr90 and Cs137 in small mammals from Nevada Test Site (Romney et al., 1983) and at a burial ground at Idaho National Engineering Laboratory (Arthur et al., 1987) show relatively high concentrations of these radionuclides in lung, pelt and gastro-intestinal tract similar to plutonium. This suggests that physical transport of these more "soluble" radionuclides is also important as with plutonium. The bioavailability of radionuclides such as cesium and strontium will depend on chemical form, local environmental conditions, and the structure and function of the relevant food webs.

Tritium would be one of the few exceptions to the general observation that physical transport mechanisms dominate in the transport of soil surface contaminants to biota. Uptake by roots or sorption through the leaf surface would dominate in tritium transport to vegetation. Levels of tritium in animals would reflect levels in the source (i. e., concentration ratios are 1 or less) since tritium is not concentrated as it moves through abiotic and biotic pathways.

As mentioned, tritium in vegetation is available to nectivorous organisms such as honeybees as well as herbivores. While tritium is readily transported through ecosystems, it is rapidly turned over in biological systems at rates corresponding to water turnover in these systems. In humans, body water turnover is about 3 days (RHH, 1970).

D 6.2.3.2 Burrowing- The subject of animal intrusion into landfills is seldom explored in any detail. The Dwyer et. al. ET cap proposal includes the use of an animal intrusion barrier. Their basic assumption in proposing the use of the barrier is that animal burrowing is bad for landfills. This is an assumption that by no means has been demonstrated conclusively. Therefore, I want to explore the subject of animal burrowing effects in more detail in this section.

The role of burrowing animals in mobilizing buried waste is not well known because very few relevant studies have ever been conducted. Some field studies deal specifically with animal burrowing in contaminated sites (O'Farrell and Gilbert, 1975; Winsor and Whicker, 1980; and Arthur and Markham, 1983; Hakonson et. al., 1982) and those studies show that burrowing animals may, in some cases, alter the vertical distribution of soil radionuclides that are present near the ground surface and in the process can become contaminated themselves.

Other studies show that animal burrowing can influence water balance, erosion, and vegetation species composition and biomass on landfill caps by changing the physical and hydrologic characteristic of cap soil (Sejkora, 1989; Hakonson et. al., 1982b; Gonzales et. al., 1993; Hakonson, 1998). Burrowing activity loosens the soil, creates surface roughness, increases infiltration, and increases soil moisture at least temporarily (Hakonson, 1998).

It could be assumed that increased soil moisture could lead to increased moisture movement into the waste. However, controlled studies on this potential problem show that increased soil moisture does not lead to increased percolation of moisture into the waste when a vegetation cover is present on the cap (Sejkora, 1981, Hakonson, 1998, Gonzales et. al., 1993).

The increased soil moisture resulting from burrowing effects on infiltration stimulates plant growth and plant transpiration (Hakonson, 2000; Gonzales et. al., 1993). Consequently, these studies show that the net effect of the animal burrowing is lower, not higher potential for percolation of water into the waste as long as avegetation cover is present (Gee and Ward, 1997; Hakonson, 2000).

The small number of studies actually conducted on contaminated sites all suggest that the effects of animal burrowing on erosion, infiltration of water into the cap, and contaminant redistribution are second order in importance or that they actually promote isolation of buried contaminants via feedback between soil moisture status and the vegetation cover. However, these studies were conducted on sites that were vegetated. Results may be altogether different if the soil surface is highly disturbed such as from overgrazing, fire, or mechanical actions such as mowing of the vegetation cover.

While the numbers of studies that are specific to waste sites are limited, there are substantially more ecological studies of animal burrowing effects on soil, vegetation, and water balance on forests and rangelands. The vast majority of published information on the effects of animal burrowing is specific to pocket gophers (Geomyidae). Furthermore, much of the research was carried out between 1930 and 1960 and appears to have been initiated primarily in response to potential impacts of gopher burrowing and herbivory on the health of rangelands and forests.

Animal burrowing into the soil on the MWL has been conclusively documented (see Dwyer et al ET cap proposal) although I could not find any publications that document the species of animal involved, their numbers, or the consequences relative to waste transport to the ground surface or contamination of the animals. Kangaroo rats, a communal animal that creates an extensive den, are probably present on the landfill. In the absence of monitoring data or special studies on the effects of burrowing by Kangaroo rats or any other species on the MWL, little can be said about the consequences of this burrowing.

D 6.2.3.3 Burrow Depths- Fossorial animals spend a major part of their life underground in tunnel systems created for resting, breeding, feeding, and excreting of waste products. Assumptions for ecological risk assessments usually use tunnel depths of about 60 cm. However, there is ample evidence in the literature that fossorial mammals can excavate burrows to much greater depths.

For example, pocketgophers develop very extensive tunnel systems in the soil although most of the tunnel system is concentrated in the upper rhizosphere. Gopher tunnel systems can extend to depths of 2 meters (Miller, 1957).

Prairie dogs excavate tunnels to over 4 m while ground squirrels, depending on species, can burrow to depths of 30-120 cm (Reynolds and Wakkinen, 1987; Linsdale, 1946). Larger species such as the badger may create burrows to at least 150 cm and 15-20 cm in diameter (Table 4). Estimates of burrowing depths for other species are given in Table 4.

Insects also have the ability to tunnel deeply into a landfill cap. For example, some ant species develop tunnel systems to 6 m (Table 5) below the ground surface (Cole, 1966; Cowan, et. al., 1985; Cline et. al., 1976) although most species tunnel to depths of 1-4 m (Hölldobler and Wilson, 1990). Studies in Idaho show that infiltration of water in areas disturbed by ants is higher than in non-disturbed areas (Blom et. al., 1994.) but that ant mound soil moisture dries out quicker than non-mound soil.

D 6.2.3.4 Rates of soil turnover- There is little question that pocket gophers, and likely other small burrowing mammals, have the potential to displace large amounts of soil as a consequence of burrowing. Maximum pocket gopher densities have been reported to range from 54-120 animals ha-1 (Hansen 1965). Actual amounts of soil moved to the surface by pocket gophers have ranged from 16-103 T ha-1 yr-1 (Mielke 1977, Spencer et al. 1985).

Estimates of 12-20 T acre-1 yr-1 have been reported for pocket gopher densities on the order of 10 per acre (Grinnell, 1923; Grinnell and Storer, 1924; Ingles, 1952; Shelford, 1929; Ellison, 1946). However, much of the displaced soil is not pushed to the surface, but is re-deposited in other parts of the burrow system. For example, Andersen (1987) found that 41-87% of excavated soil was deposited as backfill in tunnels below ground.

Hakonson et. al. (1982) conducted a study of soil excavation rates by pocket gopher on a low level waste site at Los Alamos. They found that over a 401-day period, pocket gophers on the 0.95 created 1998 separate mounds ha study area for an average mound production of about 5 day-1 ha-1. The total mass of the soil in these mounds over the 401 day study period was 11 T ha-1 yr-1, for an average excavation rate of about 30 kg ha-1day-1. Mound building activity was greatest in late summer and fall when a total of about 60 kg ha-1 of soil was brought to the surface of the landfill each day.

Hakonson et. al., (1982) also found that the digging activity of pocket gophers on the LLW site at Los Alamos turned over less than 1/10% of the cap soil during the 401-day observation period. However, the 11,255 kg of material brought to the soil surface over the 14-month period represented a volume of about 8.3 m3 so presumably about 8.3 m3 of void space was created within the cover profile. Based on an average tunnel cross sectional area of 30 cm2, as measured in the field, 8.3 m3 of void space within the cover profile represents about 2800 m of pocket gopher tunnel system per hectare.

D 6.2.3.5 Effects on Soil Characteristics- Aubertin (1971) in a study of macropores in forest soils attributed differences in hydraulic conductivity to void spaces left by decomposing roots and animal passages. These macropores provided direct conduits for water movement into the soil profile. Lysikov (1982) reported hydraulic conductivities of 6.7 mm min-l on non-mound soil in an area disturbed by moles (Talpa europaea) compared to 96.4 mm min-1 on mounds less than 1 year old. Grant et al. (1980) reported a 2-fold increase in hydraulic conductivity on pocket gopher (Thomomys talpoides) mounds compared to that of adjacent, undisturbed prairie soil.

Table 4. Burrowing Depths of Some Representative Burrowing Animals (from Cline et. al., 1982)
Species Recorded Tunneling depth (cm)
Marmota monax (marmot) 40 -50
Cynomys Ludovicianus (Black tailed prairie dog) 91-427
Spermophilus townsendi (ground squirrel) 50-80
Thomomys bottae (Botts pocket gopher) 5-35
Thomomys talpoides (pocket gopher) 10-30
Geomys bursarius (Plains pocket gopher) 23
Perognathus longimembris (pocket mouse) 52-62
Perognathus parvus (Great Basin pocket mouse) 35-193
Dipodomys spectabilis (kangaroo rat) 40-50
Dipodomys microps (Banner-tailed kangaroo rat) 25-45
Dipodomys merriami (Merriam's kangaroo rat) 26-175+
Taxidea taxus (badger) 150+

Figure 2. Burrowing depths of some representative ant species (Jensen, 2000).

Salem and Hole (1968) reported 20% of the volume of ant (Formica exsectoides) mounds being occupied by voids 2-23 mm in diameter. By applying Darcy's Law describing movement of fluid through a porous medium, the intrinsic permeability of the soil is proportional to the squared radius of the soil pores (Marshall and Holmes 1979). The range of void dimensions in the above case would result in a 100-fold difference in hydraulic conductivity.

Lockaby and Adams (1985) found a significant (P<0.0001) reduction in bulk density from 1.07 Mg m-3 to 40.85 Mg m-3 on non-mound and mound soils, respectively, in the vicinity of fire ant (Solenopsis invicta) activity in a forest soil. Similar findings were reported by Baxter and Hole (1967) on ant (F. cinerea) mounds in a prairie soil. Decreases in bulk density imply a higher fraction of pore space in the soil.

Lower bulk densities on mound vs. non-mound soils have also been reported for pocket gopher mounds (Laycock and Richardson 1975, Ross et al. 1968). This increase in pore space undoubtedly has a large influence on hydraulic conductivity of the soil.

Mielke (1977) found that soil moisture content increased from 2.6 to 7.7% on non-mound vs. mound soils in an area disturbed by pocket gophers. Although not statistically significant, the findings of Grant et al. (1980) indicated a tendency for higher moisture content on gopher mounds. Conversely, Skoczen et al. (1976) documented the drying effect brought about by mole tunnels. This drying effect was attributed to airflow through the open tunnels.

Ross et al. (1968) found that other animals more frequently disturb the soil present on and near mima-type mounds. Ground squirrels (Citellus spp.), badgers (Taxidae taxus), and toads (Bufo hemiophzts) were among the species found at these sites. The increase in animal activity in the vicinity of these mounds is thought to perpetuate the effects of the mound in modifying bulk density, soil chemistry, and vegetation distribution.

Movement of soil material by animal activity can influence the distribution of primary particles (sand, silt, clay) in the soil. Baxter and Hole (1967), Salem and Hole (1968), Alvarado et al. (1981) and Levan and Stone (1983) reported that soil material in ant mounds has a higher proportion of clay than adjacent non-mound soil. The findings of Laycock and Richardson (1975) also indicate a tendency for enrichment of soil fines in mounds resulting from pocket gopher burrowing.

In addition to affecting the compaction, porosity, and particle size distribution of the soil, animal activity has been shown to influence the amount and distribution of chemicals in the soil. Many of the studies on the influence of ant activity have indicated significant increases in levels of P, K, Ca, Mg, and Fe in mound vs. non-mound soils (Baxter and Hole 1967, Culver and Beattie 1983, Czerwinski et al. 1971, Levan and Stone 1983, Lockaby and Adams 1985, Salem and Hole 1968).

Increases in plant nutrients have also been shown to occur in mounds created by burrowing mammals (Abaturov 1972, Grant and McBrayer 1981, Mielke 1977). Laycock and Richardson (1975) also showed a slight increase in nitrogen on gopher mounds. However, Hirsch et al. (1984) and Spencer et al. (1985) reported lower levels of some nutrients in mound soils.

These discrepancies may be due to specific site characteristics and time since disturbance (Turner et al. 1973). Since clay content of soil has a direct influence on the cation exchange capacity, the differences in clay content of mound vs. non-mound soils noted earlier may contribute to the observed differences in soil chemistry. Clay also is important to soil structure and the stability of aggregates, factors which affect the detachment of soil by rainfall and runoff (Alberts et al. 1980).

D 6.2.3.6 Effects on vegetation cover- Large differences in the abundance and type of vegetation on and near areas disturbed by animal activity have been reported. Ward and Keith (1962), Luce et al. (1980), and Bandoli (1981) related differences in plant species composition to the food preference of pocket gophers. Ellison and Aldous (1952) suggested that pocket gopher foraging was responsible for a decline in dandelion (Taraxacum officinale) and other annual plant species. Ross et al. (1968) reported that mima mounds had a much higher abundance of shrubs than adjacent non-mound soils. Similar results regarding vegetation abundance and species diversity have been shown for areas disturbed by ants (Culver and Beattie 1983).

Ellison and Aldous (1952) reported increases in rhizomatous species, grasses, and sedges in central Utah. Laycock and Richardson (1975) also measured increases in total standing crops of grasses and rhizomatous forbs where gophers were present. Grant et al. (1980) found that gopher mounds resulted in a net increase of about 5-6% in overall primary productivity in shortgrass prairie.

Presumably, these increases were due to increased aeration from turnover of the soil and increased infiltration of surface water resulting from increased infiltration rates and increased surface roughness (due to soil mounding) leading to reduced runoff velocities. Mielke (1977) reported enhanced plant growth in a semi-arid environment and attributed this to gopher activity and consequent alteration of soil texture, humus content, mineral availability, and change in surface roughness.

Foster and Stubbendieck (1980), Hirsch et al. (1984), and Spencer et al. (1985) found higher proportions of bare ground in areas disturbed by gophers when compared to exclosures free of gopher activity. There was also less organic matter in the soils from the disturbed areas.

In contrast, Turner et al. (1973) and Mielke (1977) found greater aboveground biomass production near mounds. Laycock (1958), Bookman (1983), and Tilman (1983) related such differences to the environmental gradients established by animal activity. Differences in moisture and chemical content, soil compaction, and species selection by animals were listed as possible explanations for these changes in vegetation abundance, biomass, and diversity.

A study at Los Alamos (Hakonson, 1998; Gonzalez, et. al., 1997) on ET cover plots showed that pocket gopher burrowing in the presence of vegetation resulted in large decreases in runoff, erosion, and contaminant loss (tracer Cs133) via erosion but increased migration of the surface applied tracer into the subsurface soil due to increased infiltration. Vegetation slightly decreased runoff but greatly decreased erosion and contaminant loss by erosion. As with gophers, vegetation enhanced movement of contaminant into the soil. Gophers alone had an effect similar to vegetation alone in that they decreased runoff and erosion and only slightly decreased contaminant losses due to erosion.

The study concluded that the effects of pocket gopher burrowing in degrading an ET cover plots were minimal when vegetation was a component of the cover. Burrowing decreased erosion of the cover but did so at the expense of increasing water and surface contaminant migration into the soil. Those effects, however, were mitigated by soil moisture removal by the vegetation.

D 6.2.3.7 General Effects on Erosion- There is some disagreement as to the role of pocket gophers as causative agents of soil erosion. However, a preponderance of the information in the literature indicates that their role in soil erosion processes is rather small and that they may actually be beneficial in soil retention. There seems to be little or no evidence to suggest that rodents or other animals, under natural conditions, promote soil erosion (Taylor, 1935). The few authors who do state that pocket gophers contribute significantly to soil erosion (Day, 1931 and Gabrielson, 1938) provide anecdotal observations and no data specific to the question.

Gopher burrowing impact on soil structure includes increased porosity as a result of mechanical loosening, which aids in water infiltration and prevents heavy surface runoff (Grinnel, 1933; Ellison, 1946; Ingles, 1952; Ursic and Esher, 1988; Laundre, 1993). Ellison (1946) and Buechner, (1942) studied different species of pocket gophers in Utah and Texas and concluded that they had little appreciable effect on soil erosion. Numerous other studies also concluded that animals in a natural environment do little to promote erosion and may even help prevent soil loss (Taylor, 1935; Lowdermilk, 1934; Hansen and Morris, 1968).

Lowdermilk (1934) points out that the primary cause of accelerated erosion in rangelands is destruction of the native mantle of vegetation and that gophers, in the absence of heavy grazing by other herbivores, are gen