Pen Branch Research Articles

Structural inheritance in Carolinia, external to the excised Rheic Suture: Savannah River Site, South Carolina, USA.

Abstract Phanerozoic reactivations of basement fault zones are documented in 5000 m of basement core recovered from beneath the updip Atlantic Coastal Plain underlying the US Department of Energy Savannah River Site (SRS) in South Carolina. These basement fault zones are adjacent to the excised Rheic Ocean suture. Meta-intrusive rocks from c. 620 and 625 Ma contain a mylonitic fabric and intrude foliated mafic metavolcanic rocks. At c. 305 Ma, granulite facies orthogneisses were thrust over amphibolite facies meta-igneous rocks in the transpressive Tinker Creek Nappe. The overturned limb of the nappe localizes the Triassic Dunbarton Basin Border Fault. The border fault acted as a conduit for fluids in the Mesozoic and Cenozoic. At c. 220 ± 5 Ma, a potassium and silica metasomatic event affected the SRS basement. A propylitic event flushed reducing fluids through rocks as young as the Santonian. The remains of a Triassic sub-basin were identified in the northwesten part of the site. A Cretaceous and younger vein paragenesis overprints the previous events. More than 30 pseudotachylytes are found in the SRS basement and are preferentially localized on metasomatized Alleghanian chloritic fractures. Pseudotachylyte post-dates mineralized fractures. The Pen Branch Fault offsets the basement–Cretaceous unconformity and is present in c. 242 m of core between PBF-7-419 m and PBF-7-660.8 m. The Pen Branch Fault cross-cuts mineralized fractures and must post-date strike-normal zeolites.

Radionuclide distribution in soil and undecayed vegetative litter samples in a riparian system at the Savannah River Site, SC

The aim of this study is to comprehensively investigate radionuclide concentrations in surface soil and un-decayed vegetative litter along four stream systems (i.e. Fourmile Branch, Lower Three Runs, Pen Branch, and Steel Creek) at the Savannah River Site (SRS), Aiken, South Carolina. Soil and litter samples from systematically spaced 12 pairs (contaminated or uncontaminated) of plots along the four streams were analyzed for 16 distinct radionuclide activities. Lower radionuclide concentrations were observed in soil and litter samples collected along Pen Branch compared to the other 3 streams. The anthropogenic radionuclide with the highest activity was 137Cs in soil (10.6–916.9 Bq/kg) and litter (8.0–222.3 Bq/kg), while the naturally occurring radionuclides possessing the highest concentration in the samples were 40K (33.5–153.7 Bq/kg and 23.1–56.0 Bq/kg in soil and litter respectively) and 226Ra (55.6–159.9 Bq/kg and 30.2–101.8 Bq/kg in soil and litter respectively). A significant difference (p < 0.05) of radionuclide concentrations between paired-plots across four streams was observed for 241Am, 137Cs, 238Pu, 239Pu, and 226Ra in both contaminated and non-contaminated samples. 137Cs and uranium isotopes had the highest litter-to-soil correlation in contaminated (rho = 0.70) and uncontaminated plots (rho = 0.31–0.41), respectively. 90Sr was the only radionuclide with higher radioactive concentrations in litter (12.65–37.56 Bq/kg) compared to soil (1.61–4.79 Bq/kg). The result indicates that 1) historical discharges of anthropogenic 137Cs was the most important contributor of radiation contamination in the riparian environment at SRS, 2) 90Sr was the only radionuclide with higher concentration in litter than in soil, and 3) no apparent pattern in deposition density in soil or litter along downstream was observed for the radionuclides measured in this study.

Use of fish community data to evaluate restoration success of a riparian stream

From 1985 to 1988, stream and riparian habitats in Pen branch and Four Mile branch began recovering from deforestation caused by the previous release of hot water from nuclear reactors. The Pen branch corridor was replanted with wetland trees in 1995 to expedite recovery and restore the Pen branch ecosystem. Pen branch, Four Mile branch, and two relatively undisturbed streams were electrofished in 1995/1996 to determine how fish assemblages differed between the previously disturbed and undisturbed streams and whether such difference could be used to measure restoration success in Pen branch. Fish assemblages were analyzed using nonparametric multivariate statistical methods and the index of biotic integrity (IBI), a bioassessment method based on measurement of ecologically sensitive characteristics of fish assemblages. Many aspects of fish assemblage structure (e.g. species richness, disease incidence, taxonomic composition at the family level) did not differ between disturbed and undisturbed streams; however, the disturbed streams were characterized by higher densities of a number of species. These differences were successfully detected with the multivariate statistical methods; whereas, the IBI did not differ between most recovering and undisturbed sampling sites. Because fish assemblages are strongly influenced by instream habitat, and because instream habitat is strongly influenced by the riparian zone, fish assemblages can be used to measure restoration success. Nonparametric ordination methods may provide the most sensitive measure of progress towards restoration goals, although the IBI can be used during early stages of recovery to indicate when certain ecologically important aspects of structure and function in recovering streams have reached levels typical of undisturbed streams.

Influence of riparian alteration on canopy coverage and macrophyte abundance in Southeastern USA blackwater streams

Two tributary streams (Fourmile branch and Pen branch) located on the US Department of Energy's Savannah river site in west-central South Carolina, USA received thermal discharges from nuclear production reactors for over 30 years. Effluent releases produced stream water temperatures of over 50°C and stream flows of ten times above their base level. Consequently, existing plant and animal communities within the stream channels were killed and riparian zones largely destroyed. We compared canopy coverage and macrophyte abundance in these disturbed streams after 7–13 years of ambient flows and compared them to two similar, undisturbed streams. We also examined the effects of a more recent woody canopy removal associated with a restoration effort in ‘treated’ sections of Pen branch. We collected data in Spring and Fall from May 1995 through May 1997. A gradient in canopy cover existed, ranging from a fully open herbaceous canopy in the treated sections of Pen branch, through a moderately closed canopy of herbaceous plants, shrubs, and willows in the post-thermal ‘control’ streams, to a nearly closed hardwood tree canopy in the undisturbed streams. Even though Fourmile branch had 3 more years of growth, Pen branch had a more closed upper canopy. Total aquatic macrophyte abundance was negatively related to canopy cover producing nearly the reverse gradient among streams as did the canopy cover. However, control sections of Pen branch had a more closed canopy than Fourmile branch, but total macrophyte abundance was higher in Pen branch. This can be attributed to the presence of submergent macrophytes in Pen branch that were absent in Fourmile branch. Different structural types of macrophytes varied in their degree of limitation by canopy coverage and in their seasonal patterns of growth. Stream habitats remain severely altered due to destruction of the riparian vegetation by past thermal effluents. The full range of effects of the alteration of canopy coverage and the resultant macrophyte abundances on these streams should be the focus of future analyses.

Operational restoration of the Pen Branch bottomland hardwood and swamp wetlands — the research setting

The Savannah River swamp, a 3020 ha forested wetland on the floodplain of the Savannah River, USA is located on the Department of Energy's Savannah River site (SRS) near Aiken, SC. Historically, the swamp consisted of approximately 50% bald cypress–water tupelo ( Taxodium distichum– Nyssa aquatica) stands, 40% mixed bottomland hardwood stands, and 10% shrub, marsh, and open water. Creek corridors were typical of southeastern bottomland hardwood forests. Hydrology was controlled by flooding of the Savannah River and by flow from four creeks that drain into the swamp prior to flow into the Savannah River. Upstream dams have caused some alteration of the water levels and timing of flooding within the floodplain. Major impacts to the swamp hydrology occurred with the completion of the production reactors and one coal-fired powerhouse at the SRS in the early 1950s. Flow in one of the tributaries, Pen Branch, was typically 0.3 m 3 s −1 (10–20 cfs) prior to reactor pumping and 11.0 m 3 s −1 (400 cfs) during pumping from 1954 to 1988. Sustained increases in water volume resulted in overflow of the original stream banks, the creation of additional floodplains, considerable erosion of the original stream corridor, and deposition of a deep silt layer on the newly formed delta. Heated water was discharged directly into Pen Branch and water temperature in the stream often exceeded 65°C. The nearly continuous flooding of the swamp, the thermal load of the water, and the heavy silting resulted in complete mortality of the original vegetation in large areas of the floodplain. In the years since pumping was reduced, no volunteer seedlings of heavy-seeded hardwoods or cypress have been found in the floodplain corridor. Research was conducted to determine methods to reintroduce tree species characteristic of more mature forested wetlands. Species composition and selection were altered based on the current and expected hydrological regimes that the reforestation areas will be experiencing.

Spatial and temporal patterns of carbon storage and species richness in three South Carolina coastal plain riparian forests

The distribution of organic matter within a floodplain is a controlling factor affecting water quality, habitat, and food webs. Accordingly, development of vegetation in the riparian zone can be expected to influence ecosystem functions, and organic matter storage patterns are believed to be indicators of functional recovery in disturbed riparian zones. Our objective was to compare the distribution and allocation of organic matter among microsites within the floodplain and with temporal changes (successional status) associated with community development. Three third order streams in the upper coastal plain of South Carolina were selected. Measurement transects were established across three floodplains of varying successional status, Meyer's branch; a mature riparian hardwood forest; Fourmile branch; a mid-successional riparian forest; and Pen Branch, an early successional riparian forest. Overall, measurements of aboveground biomass, soil carbon, and stand structure indicate that the early and mid successional stands are becoming more similar to the mature stand and that microsite differences within the braided, riparian stream systems are small.

Small mammal populations in a restored stream corridor

An opportunity to study the response of a small mammal community to restoration of a riparian wetland was provided by the Pen branch project at the Savannah river site (SRS). Live trapping of small mammals was conducted on six transects at Pen branch in 1996 and 1998 and at three transects at Meyers branch, an unimpacted stream at SRS, in 1997 and 1998. Distributions of rates of capture of the four most common species were both spatially and temporally uneven. Kruskal–Wallis one-way analysis of variance revealed no significant differences in the relationship of capture rates between species and between treatment and both the within-stream control and Meyers branch. Habitat use and movement within stream corridors appears to be dependent primarily on species, with age and sex perhaps contributing to habitat preference and distance moved. The lack of differences in capture rates related to transect or treatment may be due to the close proximity of sample transects relative to the movement potential of the species sampled.

Herpetofaunal and vegetational characterization of a thermally-impacted stream at the beginning of restoration

Pen Branch, a third order stream on the Savannah River Site (SRS), located near Aiken, SC, USA, received thermal effluents from the cooling system of a nuclear production reactor from 1954 to 1988. The thermal effluent and increased flow destroyed vegetation in the stream corridor (i.e. impacted portion of the floodplain), and subsequent erosion created a braided stream system with a greatly expanded delta. Restoration of the area began with planting of bottomland hardwood species in 1993. Occurrence of amphibians and reptiles was monitored by daily sampling from 1 January 1995 through 30 September 1996 to characterize the course of the restoration. Vegetation was sampled in the summer of 1996 to characterize the habitats in the unimpacted riparian zone and the impacted stream corridor. A total of 12 580 individuals representing 72 species of herpetofauna were captured. There were no significant differences in relative abundance or diversity of herpetofauna in unplanted versus planted zones within the impacted corridor 3 years after planting. Likewise, there were no significant differences in abundance or diversity of herpetofauna in the upper and lower corridor areas, which differed in site preparation before planting, or in riparian zones of different widths. However, species diversity of amphibians and reptiles in the unimpacted riparian zone was significantly higher than on vegetated islands located between stream braids within the impacted floodplain corridor. There were also significantly more species and individuals within the riparian zone than in the corridor, and the species assemblage within the riparian zone differed from that of the corridor. Woody vegetation within the unimpacted riparian zone was significantly higher in basal area than on islands within the corridor.

Effects of restoration techniques on breeding birds in a thermally-impacted bottomland hardwood forest

We evaluated the effects of revegetation techniques on breeding bird communities in a bottomland hardwood forest impacted by thermal effluent. In 1993, sections of the Pen Branch bottomland on the Savannah River Site, South Carolina, were herbicide-treated (glyphosate), burned, and planted; other sections were planted only while others were unaltered and served as controls. Few differences in the avian community occurred at 1 and 2 years post-treatment among treatments. Plots that were herbicide-treated, burned, and planted had greater species richness in 1994 and abundance in 1995 than sections that were planted only ( P<0.05). Bird species composition differed slightly among treatments and White-eyed Vireos ( Vireo griseus), Common Yellowthroats ( Geothlypis trichas), Indigo Buntings ( Passerina cyanea), and Red-winged Blackbirds ( Agelaius phoeniceus) were the most abundant species in the corridor. Revegetation techniques used to restore this thermally-impacted bottomland had little effect on the avian communities 1 and 2 years post-treatment.

The use of tree shelters in restoring forest species to a floodplain delta: 5-year results

Without herbivory control, natural seed sources, and seasonal flood events, recovery of the Pen branch delta in South Carolina to former conditions (prior to thermal discharge) may take many years. To assess the recovery process, seedlings of baldcypress ( Taxodium distichum), water tupelo ( Nyssa aquatica), swamp blackgum ( Nyssa sylvatica var. biflora), and green ash ( Fraxinus pennsylvanica) were planted in four areas of the delta in 1994. One-half of the seedlings were protected using tree shelters 1.5 m tall. Heights of seedlings were taken after planting and at the end of each growing season from 1994 to 1998. Survival at the end of 5 years ranged from 67 to 100% for seedlings in tree shelters and 2–90% for those not in tree shelters. Survival of seedlings without tree shelters was generally low, and mortality was attributed mainly to beaver damage. Although water tupelo, swamp blackgum, and green ash seedlings tended to die once clipped by beaver, 85% of the clipped baldcypress resprouted after clipping, and new sprouts grew vigorously. During year 1, height growth of tree shelter seedlings was significantly greater than non-tree shelter seedlings for all species, but once the seedling emerged from the top of the shelter, growth differences declined dramatically. Differences in height growth among areas was highly variable from year to year, and no one species tended to grow better in one area over another throughout the period. Restoration of the Pen branch delta to a baldcypress–water tupelo forest similar to the surrounding forest is possible. Baldcypress and water tupelo seem ideally suited to growing in all areas of the delta equally well, but it may take 10–20 years before the seedlings are of sufficient size to not be affected by herbivory and old enough to produce sufficient quantities of seed to maintain the forest.

Seismotectonic structures along the Savannah River Corridor, South Carolina, U.S.A.

The Savannah River Corridor is defined herein to extend from the Brevard zone in South Carolina to just northwest of Charleston, South Carolina. The length of the Corridor is 400 km, the width 300 km, and the strike is S 40°E. The Corridor includes (1) the Savannah River Site, which is the locus of the most detailed seismic reflection data set in the eastern United States, and is the keystone for the extrapolation of regional reflection seismic data away from the Site, and (2) the South Carolina–Georgia seismic zone. The principal data used for the conclusions of this study are reflection seismic data acquired by the Consortium for Continental Reflection Profiling (COCORP), the Appalachian Ultra-Deep Core Hole (ADCOH) project, and Conoco, Inc. (SRS). Relatively higher seismicity appears to be associated with a crustal scale antiform that has been imaged in the Savannah River Corridor as well as to the northeast along strike in Virginia. An increase in earthquake activity occurs in that part of the crystalline allochthon that overlies and conceals a foreland-dipping imbricate stack that has developed in early Paleozoic rocks beneath the overthrust Piedmont crystalline allochthon. To the southeast, hypocenters of two SRS earthquakes (1985 and 1988), projected along the strike of the Pen Branch northwest-bounding fault of the Dunbarton Triassic rift basin plot on a highly reflective southeastward-dipping shear zone that underlies the basin. Here, the proximity of the hypocenters to the border fault of the Dunbarton basin suggests that syntectonic structures associated with the Pen Branch border fault might be involved. Some of the seismicity is probably associated with reactivation of older Paleozoic and Mesozoic structures.

The Pen Branch Project

The Pen Branch Project

Notes on the use of wood refugia by Physella heterostropha (say) in a thermally disturbed swamp

Physella heterostropha size distributions and densities were measured in Pen Branch delta, a thermally altered swamp of the Savannah River, South Carolina. During cessation of thermal input, measurements were taken in a highly impacted area and more natural area. Of the three substrates sampled, logs, small woody material and benthos, snail density was highest on small pieces of submerged wood and < 20% of the snails on logs were out of the water. Following resumption of thermal input, snail density was four times higher on logs than previously measured. Snails responded to elevated water temperatures by congregating in a narrow band slightly above the water. Snail size did not appear to affect their response. However, the thermal history of the snails influenced their behaviour and survival rates. Logs may be a potential refuge for snails when water temperature are greater greater than 38°C. However, our results indicate that long-term survival in this manner may be impossible.

Thermal modeling of heat dissipation in the pen branch delta using thermal infrared imagery

Abstract Thermal infrared imagery of the Pen Branch delta on the Savannah River Plant near Aiken, South Carolina, USA, were obtained on 12 March 1983 and 6 April 1984 during flood and non‐flood conditions. These data were registered to one another and then analyzed using digital image processing techniques to: (1) compute the rate of temperature change as the thermal effluent progressed downstream using transect techniques, and (2) mathematically model the spatial distribution of the thermal effluent using fourth and seventh‐order trend surfaces. These two mathematical expressions of the thermal data allowed more effective understanding of how the temperature of the thermal effluent changed as it progressed through the Pen Branch delta hydrologic system.