Our Work
These research projects span diverse ecological and environmental themes, integrating fieldwork, laboratory analysis, and statistical modeling. Drs. Belk and Tuckfield focused on fish ecology and conservation, studying recruitment and habitat in endangered species like the June sucker and Leatherside chub. In collaboration with NOAA, Dr. McArthur led research on the link between metal pollution and antibiotic resistance in estuarine bacteria. Internationally, conservation initiatives in Latin America and the Caribbean assessed industrial and agricultural impacts on biodiversity, including fish and bird populations. Additional work includes groundwater contamination analysis under federal RCRA regulations, statistical evaluation of global temperature trends in relation to CO₂ levels, and development of genomic tools for distinguishing bacterial strains. A forthcoming book emphasizes statistical thinking in microbial biology for more effective conservation decision-making.
Biobio River Study: Chile, South America
This research stems from a decade of collaboration on life history research in fish. The aim is to understand the relationships between life history characteristics such as age (i.e. the standard length) and clutch mass and other features of fish ecology that include recruitment, predator escape velocity, and the effects of survival refugia.
The scope of this research with Dr. Mark Belk included the development of experimental designs, and the creation and placement of June sucker (Chasmistes liorus) fish enclosures in Utah to assess density dependence effects on growth and survival. This research is part of the statewide effort to reestablish and preserve this endangered species (http://www.junesuckerrecovery.org/).
Other research included the field collection of Mosquito fish (Gambusia affinis) from South Carolina swamps to study the relation between body morphology older gravid females and younger pregnant females with regard to predator avoidance and the associated selection pressure on life history characteristics.
In addition, Chilean biologist Dr. Evelyn Habit, while on sabbatical leave from the University of Concepcion, was primarily engaged in ecological research among fishes of her native South American country.
Dr's. Habit, Belk, and Tuckfield collaborated to review, analyze and publish evidence for fish community impacts on the Biobio
River in Chile downstream from industrial facilities. Data anlaysis and experimental design consulting support was provided by Dr. Tuckfield to the lead ecologist, Dr. Belk.
The intent of this research was to characterize the status of fish populations in the Biobio river system after more than a decade of introduced industrial activity.
This study found that both the catch per unit effort (CPUE) and the biomass per unit effort (BPUE) in the sample collections of fish along the major channels of the Biobio watershed were significantly reduced in Zone 3 (i.e. lower mid reaches of the river) located downstream from sewage treatment plants and pulp production plants compared to such measures in the prior decade.
NOAA Oceans and Human Health Initiative
In 2003 our Research Team submitted a proposal in response to a call from the NOAA Oceans and Human Health Initiative. The aim was to study the potential direct
selection effects of metal pollution on the apparent acquisition of antibiotic resistance traits by indirect selection among estuarine bacteria.
Prior research in a metal contaminated stream at the US Department of Energy Savannah River Site (SRS)
had shown a positive correlation between two aminoglycosides (kanamycin and streptomycin) and mercury concentration in the sediments.
Three estuarine study sites were chosen, two on the South Carolina coast and one on the Georgia coast. One (Ace Basin, SC) of the three was selected as a control because it is widely considered a near pristine wetland. The other two, Ship Yard Creek near Charleston, SC and an EPA Superfund (LCP) site near Brunswick, GA were highly and historically contaminated with heavy metals.
Of the three species of bacteria identified in the brackish sediments, only E. coli showed showed the predicted prevalence (proportion of isolates) patterns of one or more antibiotic resistance traits among study sites, that is, higher prevalence in contaminated sites compared to the control. Three peer reviewed publications came from this research effort.
Antibiotic Resistance Trait Complexity
The strong positive association between metal stress an the acquisition of antibiotic resistance traits among microbes in laboratory microcosms was not observed with two species of Vibrio bacteria collected from the estuarine study sites. However. in E. coli bacteria the complexity of antibiotic resistance traits metal contaminated sites was higher compared to the ACE basin control study site. Further research initiative are suggested by this evidence that would establish a species dependent metals bioavailability effect.
Global Surface Temperature Analysis
Recently the Fourth Assessment Report of the International Panel on Climate Change (IPCC) reported an "accelerating" trend in global surface temperatures attributed to the increasing accumulation of carbon dioxide (CO2) in the earth's atmosphere.
Dr. Craig Loehle, widely published ecologist and Principal Scientist with the National Council for Air and Stream Improvement (NCASI), invited Dr. Tuckfield to corroborate
This map produced by NASA shows an example of the cumulative surface temperature changes in during a warming period from 1951-1980.
the analyses performed by the IPCC on the global surface temperature trends. These data were obtained from the NASA Goddard Institute for Space Studies (GISS). We compared trend slope of the most recent 30, 40, and 50 year sequences of annual mean surface temperatures to a corresponding sequence in the 1880-present data collection with the most positive trend.
A statistical method was devised via a general linear model which provides a simultaneous inference test between the linear slope parameters of each sequence pair, for each of the three sequence lengths. The results indicate no significant difference in the slope values. In other words, these statistical tests do not support the IPCC conclusion of acceleration in global surface temperatures during the last century.
Dr. Tuckfield compared these data also to CO2 measurements from the Mauna Loa, HI observatory and found that there is in fact a positive and statistically significant correlation between increasing CO2 and increasing temperature in the last 30 years.
However, there is absolutely no correlation between these two measurements during the period of 1954 (the 1st year of CO2 measurement) and 1977. In fact, the GISS average mean temperature data suggest successive periods during the last 120+ years of sequences of temperature increase and sequence of temperature "plateaus" on the order of 30 years each.
The second and fourth sequences (1917-1946, 1977-2006) in the GISS data show statistically equivalent and positive slopes. The first and third sequences (1880-1916, 1947-1976) are not significantly different from zero or from each other. The slope of the first sequence in the Hadley data is significant and negative and the third significant and positive, though significantly smaller than either the second or fourth, which are again statistically equivalent. Why isn't this a consistently increasing temperature trend? Annual atmospheric CO2 measurements have been increasing continuously since 1959 at the Mauna Loa, HI weather station, but show a positive correlation with annual mean global surface temperature only since 1977 and no correlation before.
What are we missing here? Global energy demand will surely increase with temperature and perhaps will not be offset by reduced energy expenditures in higher latitudes. It will only be exacerbated by population and economic growth. Therefore, adaptation may be as essential as mitigation if science cannot effectively model, simulate, and learn fast enough. Since the first holistic natural science is ecosystem science, it is in a strong position to champion efforts to seek a deeper understanding of planetary homeostasis. The study of globosystem science modeling and measurement will be the next higher scale of ecosystem research, and the ultimate logos of this oikos.
RCRA Groundwater
Statistical Consulting Groundwater (GW) Monitoring
According to the Congressional Federal Register (CFR) 40 regulations part 264 subpart F, all facilities that have the potential to impact groundwater quality with facility
leachate must provide in the post-closure care plan that will analyze groundwater contamination in each monitoring well that is downgradient to the facility and compare them to the contamination (if any) in the
designated reference wells upgradient to that facility.
At the US Department of Energy, Savannah River Site (SRS), these plans are the responsibility of the Environmental Restoration Division and later the Soil and Ground Contamination Protection division of the principal contractor, viz., the Washington Savannah River Company. Each plan required a state regulator approved statistical analysis method for the requisite upgradient vs. downgradient monitoring well comparisons.
As a statisticians with the Savannah River National Laboratory (SRNL), Drs. Harris and Tuckfield provided and supervised the development of such statistical methodology and data analysis of the SRS facility groundwater monitoring data as required by the
federal regulations. These were often team consulting efforts with participation by other statisticians in the Statistical Consulting Section (SGS) and other groundwater and hydrogeology professionals (see attached example groundwater contamination reports).
Example Upgradient vs Downgradient
Well Comparisons
A comparison of each POC well versus the combined data from five Background wells is shown for two example constituents of concern, viz., gross alpha, and Ra226. Note also that data were transformed to a common log scale since measurement values crossed 1 or more orders of magnitude. Well IDs (names) in black are significantly
different (p <.05) than BACKGROUND.
CVOC Historical Case Analysis
A nationwide survey of chlorinated volatile organic compound (CVOC) plumes was conducted across a spectrum of sites from diverse hydrogeologic environments and contaminant release scenarios. The goal was to evaluate significant trends in the data that relate plume behavior to site variables (e.g., source strength, mean groundwater
velocity, reductive dehalogenation regime) through correlation and population analyses. Data from 65 sites (government facilities, dry cleaners, landfills) were analyzed, yielding 247 individual CVOC plumes by compound. Data analyses revealed several trends, notably correlations between plume length and maximum observed concentration (presumably reflecting the source term) and mean groundwater velocities.
This work was a collaborative effort between the US Department of Energy (DOE) and the EPA and included hydrogeologists at the Lawrence Livermore National Laboratory and statisticians from the Savannah River National Laboratory. It led to a technical report prepared by the collaborators and to a professional journal publication given in the reference below.
CVOC Plume Length Modeling
REFERENCE
McNab, W. W. Jr., Rice, D. W., and Tuckfield, R. C. 2000. Evaluating chlorinated hydrocarbon plume behavior using historical case population analyses. Bioremed. J. 4:311-335.
Cumulative distributions of residuals from the ANCOVA model. The small minimum value for the strong reductive dehalogenation set is indicative of shorter-than-expected plume lengths after taking source concentration and groundwater velocity covariates into account. GW Monitoring Well Network Efficiency. We at SGS also developed methods in collaboration with other SRNL and Lawrence Livermore National Laboratory (LLNL) professionals to improve the efficiency of groundwater sampling in groundwater monitoring well networks. We developed a method for a reliable, databased groundwater sampling frequency and a method for identifying redundant groundwater monitoring wells within a well network (see attached example monitoring network efficiency articles).
GW Sampling Frequency Estimation
Using a technique from geostatistics known as a variogram, an estimate of the sampling frequency for a given groundwater monitoring well can be obtained as the empirical variogram range parameter. This is calculated from among all possible pairs of constituent concentration values in a given temporal sequence of measurements separated by h units of time. The sum of the squared differences between all possible pair (γ h ) separated by h time units is plotted for increasing values of h. The typical behavior of such a plot usually shows and increase in γ h with increasing h to an asymptote. The corresponding value of h at this point is called the range and
represents the an estimated amount of time between groundwater sampling events such that the correlation between any pair of constituent concentration measurements separated by that much time is equivalently zero.
In the paper by Tuckfield (1994) two simulated time series (X and Y) of constituent concentration data were used to calculate the temporal variograms for each. Results showed that indeed a range parameter could be calculated using gaussian models of the variogram plot wherein the range parameter value in units of lag-time between measurements becomes and reasonable estimate of groundwater sampling frequency.
Research Findings - Background
The genome of any organism is "its hereditary information encoded in DNA (or, for retroviruses, RNA)." Each of these molecules of inheritance is composed of a double stranded helix of pairs of nucleotide bases, adenine (A) - thymine (T), and guanine (G) - cytosine (C). If we were to unwind the DNA helix we would see a long sequence of such bases on each single strand of DNA, sequences of A, T, G, or C the corresponding single letter codes for each nucleotide base.
Three species of bacteria (prokaryotes) were selected and their genomic DNA sequences were analyzed. Of these species, Escherichia coli, Synechococus elongata, and Deinococus radiodurans, they were thought to be of different evolutionary ages.
The proportion of As, Ts, Gs, and Cs on a single strand of DNA for each species genome was calculated using the SAS statistical computing software and displayed using the JMP statistical computing software, both data analysis tools of the SAS
Institute. The proportion was calculated as the number of As, Ts, Gs, or Cs in a sequence of 1,000 codons (i.e., 3,000 bases) along the 5' - 3' length of a single DNA strand in the genome. In prokaryote microorganisms, the entire genome (with a few exceptions) consists of single often circular chromosome which consists on the order of some multiple 10 6 nucleotide pairs or bases denoted as Mb in the unit of measure.
Patterns in these proportions of nucleotide bases may be suggestive not only of the kinds of environments in which inhabit (e.g., E. coli is an enterobacter), but their relative phylogeny. For instance, microorganisms that are known as extremopiles such as those found in geothermal vents, Antarctic ice cores, or hyperbaric environments are frequently referred to as "G-C rich". That is, these organisms show much higher proportions of guanine and cytosine within their genomes.
Hypothesis
The hypothesis was considered wherein S. elongata was considered to be relatively older phylogenetically than, E. coli, and E. coli than D. radiodurans since the latter was only recently isolated from the walls of waste tanks containing highly radioactive waste (on the order of 10M rads). Therefore, the prediction was that former two
species should be more & "G-C rich" than D. radiodurans.
Genomic Diversity
Results (see sections below) showed just the opposite. D. radiodurans was even more G-C rich that either S. elongata or E. coli. S. elongata is a representative of a the cyanobacteria or blue-green algae. Such can be found among the most ancient of fossils called stromatolites which are approximately 3.5 billion years old. But these results suggest that G-C richness is related more to biological function in extreme environments than relative phylogeny.
These data also showed that E. coli, which is likely to live in the least extreme and most stable environment among these three species of bacteria, i.e. your gut, has the
least disparity between G-C and A-T base pairs as reflected in the highest values for the Shannon-Weaver diversity index, a measure of the evenness of information
content.
In addition, a geostatistical variogram and corresponding correlogram was produced for the genome of each of these three species of bacteria. This graphic (see below)
shows the degree of spatial correlation among pairs coding units (CUs) separated by h CUs. These graphics show that for each species that CUs that are contiguous or nearly so (i.e., within 3 CUs of each other) are more like each other than those CUs that are farther apart in genomic sequence.
Spatial Correlation of Codons
