Slide 1

Ridge to reef assessment of metal concentrations and mineralogy in rocks and
*R.J. Harrington , S.C. Gray , C.E. Ramos-Scharrón , B. O’Shea
University of San Diego, Department of Marine and Environmental Studies, San Diego, CA 92110
2 University of Texas at Austin, Department of Geography and the Environment, Austin, TX 78712
*[email protected]
HCA Dendrogram
Land development on high relief islands, particularly in the form of unpaved roads, has the potential to increase surface
erosion and enhance the contribution of land-based sediments and associated metals (i.e., Fe, Al, Ti, Ba, Cu, Zr, Zn, K) to
the marine environment. High concentrations of metals in sediment are of concern because they may bioaccumulate in
some benthic invertebrates (Williams et al. 2010). The island of St. John in the US Virgin Islands (Fig. 1 at right) is a steep
volcanic island surrounded by fringing coral reefs. St. John is an ideal location to study how watershed development may
be impacting metal accumulation in sensitive coral reef habitats because: a) elevated concentrations of metals have been
found in St. John’s hydrothermally-altered bedrock and soils (relative to surrounding unaltered bedrock); b) land protected
under the Virgin Islands National Park makes the comparison of undeveloped and developed areas viable; and c) previous
research on St. John has demonstrated that human activities (especially the building of unpaved roads) have increased
watershed sediment yields (Ramos-Scharrón and MacDonald 2007), terrigenous sediment accumulation in the marine
environment (Gray et al., 2012), and sedimentation rates by 3-4 times the long-term historic rates (Brooks et al. 2007).
Figure 2. (Left) Field sampling of a bedrock outcrop in Coral Bay Watershed. (Center) Stream sediments were
collected with silt-fence sediment trap. (Right) Marine samples were collected using a sediment trap consisting of
PVC tubes set 60 cm above the ocean floor.
In order to better understand the linkages between terrigenous erosion and marine deposition of metals on St. John, our
study objectives were to:
1) identify which major and trace elements originate from the watershed and are deposited in the marine environment
around St. John;
2) describe how the mineralogy and concentration of major and trace elements vary spatially from ridge to reef;
3) quantify any differences between developed and undeveloped watershed-bay pairings in terms of a) major and trace
elemental concentrations and b) metal flux; and
4) describe whether developed vs. undeveloped differences in major and trace element concentrations and flux in the
marine environment were consistent with watershed-scale modeled sediment yield rates.
Petrography & mineralogy. Petrographic and XRD analyses
of plagiorhyolite and basalt (Fig. 3) demonstrate that the
bedrock units (Water Island Fm.), stream bed sediments,
and the terrigenous component of the marine sediments
were generally similar in composition and composed of
mostly albite, quartz, chlorite (plagiorhyolite), plagioclase,
clinopyroxene, epidote, chlorite, and opaque minerals
Watershed-derived major and trace elements. In order to
determine which of the 29 metals measured were most
likely to have been conservatively transported from the
watershed to the marine environment, the major and trace
element concentrations were compared to the proportion
of terrigenous sediment in 12 marine samples collected in
marine sediment traps. Five major and trace elements
(FeO, Al2O3, TiO2, Cu, Zr) were strongly correlated (R2 = 0.80
- 0.94) and three (Ba, Zn, K2O) were moderately correlated
(R2= 0.72 - 0.79) with percent terrigenous sediment (Fig.4).
This suggests that the major and trace elements TiO2, FeO,
Al2O3, Cu, and Zr may be indicators for terrigenous source
A total of 63 geologic (hydrothermally altered [n=4] and non-hydrothermally altered bedrock [n=14], soil [n=20], and
stream bed [n=11]) samples, were collected by hand or in silt-fence sediment traps from within the developed (Coral
Bay) and undeveloped (Little Lameshur) watersheds (Fig. 2). Shore and reef sediments (n=14) were collected in marine
sediment traps over a 26 day period (Fig. 2). The mineralogy of all samples was determined by petrography & XRD, and
10 major and 19 trace element concentrations were measured by XRF. Principle component analysis (PCA) correlated
samples that were then plotted into 2 dimensional space. Hierarchical cluster analysis (HCA) grouped samples with
similar major and trace element concentrations. The percent terrigenous sediment for each marine sample was analyzed
by LOI (Loss on Ignition), and metal flux rates to the marine environment were calculated.
Sediment production rates from unpaved roads were measured with silt-fence sediment traps in 1998-2000 (n =76
measurements) and in 2010-2011 (n = 25 measurements). Multiple-linear regression analyses were used to develop a
road-segment scale erosion model that was incorporated into a GIS system to estimate the contribution of roads into
watershed- and sub-watershed-scale sediment yields.
Figure 1. St. John, U.S. Virgin Islands (Top Left). Eastern St. John, the study area (Bottom) showing sample locations
within and below developed (Coral Bay) (shaded in light green) and undeveloped (Little Lameshur) (shaded light
brown) watersheds. Sample types are depicted as different shaped symbols. The sample shapes are colored according
to their grouping by hierarchical cluster analysis (HCA) as shown in the dendrogram (middle right). Orange shading
shows the general location of hydrothermally altered bedrock.
Figure 4.Correlations
between percent
terrigenous material and
major element (above)
and trace element
(below) concentrations in
marine sediments. Zn, Ba,
and K2O were also
correlated with R2 of 0.80,
0.79, and 0.72, and Pvalues of < 0.001
Undeveloped (Lameshur Bay)
Developed (Coral Bay)
Ridge to reef variation. The ranges of major
and trace element concentrations measured
in samples collected in the watersheds
(bedrock, soil, stream) generally do not
appear to differ from each other. Median
major and trace element concentrations
were always higher in the rock and sediment
samples collected in the watershed sites
compared to those collected at the reef sites
and the shore sites below undeveloped
watersheds (Fig. 5). However, median major
and trace element concentrations in
sediment collected at the shore sites below
developed watersheds were similar to
concentrations at watershed sites. Major
and trace element concentrations of all
sediment types were usually higher (and
more variable) in the developed compared to
the undeveloped locations (Fig. 5).
Figure 6. PCA results for different sample types using major element percentages (left
graphs) and trace element concentrations (right graphs). The upper graphs compare PCA
results between sample types and the lower graphs compare PCA results between samples
collected in developed and undeveloped areas. For the major elements, a positive PC1 =
higher CaO and lower SiO2 and Al2O3 and a positive PC2 = higher FeO and MnO and lower
K2O concentrations. For the trace elements, a positive PC1 = higher Ni, Cr, and V and lower
Rb and Zr and a positive PC2 = higher Ba and Zr and lower Sr. Two separate groups are
displayed: a terrigenous group (shaded light gray) and a marine group (shaded light blue).
Table 1. Marine sediment major and trace element concentrations (top) and flux rates
(bottom) plotted by developed sites in Coral Bay and undeveloped sites in Lameshur Bay.
Geochemical Grouping. HCA of major and
trace elements generally grouped samples
of similar geographic locations and similar
sample types (Fig. 1). Marine shore
sediments collected below the developed
watershed were more similar geochemically
to the bedrock sediments than were those
collected below the undeveloped
watershed. Principle components, PC1 and
PC2, accounted for 67% and 61% of the
variance in the major and trace element
concentrations, respectively (Fig. 6). The
PCA of watershed rocks and sediments
collected in developed and undeveloped
watersheds overlapped, suggesting
generally similar geochemical composition
of watershed rocks. PCA generally grouped
marine and watershed samples separately
but the geochemistry of some of the marine
shore samples in the developed area
grouped with the watershed samples (Fig.
Developed/Undeveloped differences in
element concentrations and flux rates.
Major and trace element concentrations and
flux rates (in 10 marine sediment traps)
were greater at the shore than on reef
environments and always greater below the
developed watersheds (Table 1, Figure 7).
Modeled sediment delivery and yields.
Estimated sediment yields into Coral Bay
from its highly developed watershed were
517 Mg yr-1, or 12.6 times higher than the 41
Mg yr-1 yields reaching Lameshur Bay.
Figure 3. Rock and sediment thin section photomicrographs
shown in order from ridge (top) to reef (bottom). Stream bed
sediments contain rhyolite (A) and basalt grains (B) and Reef
trap sediments contain rhyolite (A) and carbonate grains (B).
Table2. Watershed size, sediment delivery and yield estimates from (Ramos-Scharrón
and MacDonald 2007, Ramos-Scharrón et al. 2012). Ratios between developed areas
and undeveloped areas are calculated for comparison.
Figure 7. Ratios of developed to
undeveloped major and trace element
concentrations and flux rates for the
shore (solid color) and reef (diagonal
lines) (Table 1) are displayed.
The major and trace-element concentrations as well as the mineralogy of marine terrigenous sediments demonstrate a
watershed source for some major and trace elements. Near-shore sediments in the developed areas were
characterized by mineralogy and major and trace element concentrations similar to watershed rocks and sediments,
suggesting a strong watershed influence in the developed areas. However, near shore sediments below the
undeveloped watershed contained less watershed derived metals and minerals. Marine major and trace element
concentrations were typically between 5 and 10 times higher at developed compared to the undeveloped shore sites
and between 2 to 52 times higher at developed compared to undeveloped reef sites. Marine metal flux rates were
typically between 2 to 7 times higher at developed shore sites and 33 to 430 times higher at developed reef sites than at
corresponding sites in the undeveloped areas. These differences were generally consistent with modeled sediment
yields, which were 12.6 times higher for the developed compared to the undeveloped watershed.
Figure 5. Box and whisker plots showing variation in
median metal concentrations from ridge to reef in the
undeveloped (left) compared to the developed (right)
watershed. The box represents the inter quartile range
(IQR) between the 25th and 75th percentile, and the
whiskers represent a value 1.5 times the IQR above and
below the 75thand the 25th percentile. The median is the
center line in the box. The dots show outliers. The ridge
to reef concentration patterns of Zn, Ba, and K were
similar to those depicted above.
This work has been supported by funding from the NOAA Coral Reef Conservation Program and NOAA/ARRA to the
VI-RC&D. Thanks to San Diego State University Geological Sciences Laboratory and the Washington State University
Geoanalytical Laboratory for their help and training with XRF and XRD. Also thank you to Aaron Pietruszka, Amalia
DeGrood, Whitney Sears, Zoe Hastings, Thomas DeCarlo, Amanda Greenstein, Nick Przyuski, Esther Araiza, Deserae
Rawling, Marverick Carey, Yi-Chen Hsieh, Matt Knoblock, Napoleon Guidino and USVI community field assistants
including Phil Strenger, Hewitt Schlereth, Roy Proctor and Bruce Swanson.
 Brooks, G.R., Devine, B., Larson, R.A., Rood, B.P. (2007). Sedimentary development of Coral Bay, St. John, USVI: A shift from natural to anthropogenic influences. Caribbean
Journal of Science 43(2): 226-243.
 Gray, S.C., Sears, W.T., Kolupski, M.L., DeGrood, A.M., and Fox, M.D. (2012). Factors affecting land-based sedimentation in coastal bays, US Virgin Islands. Proceedings of the
12th International Coral Reef Symposium, Cairns, Australia July 9th-13th, 2012.
 Ramos Scharrón, C. E. (2010). "Sediment production from unpaved roads in a sub-tropical dry setting -- Southwestern Puerto Rico." Catena 82(3): 146-158.
 Ramos-Scharrón, C. E. and L. H. MacDonald (2007). "Runoff and suspended sediment yields from an unpaved road segment, St John, US Virgin Islands." Hydrological
Processes 21(1): 35-50.
 Ramos-Scharrón, C.E., Atkinson, S., Day, K., Devine, B., Reale-Munroe, K., Swanson, B. (2012). “USVI Coastal Habitat Restoration Through Watershed Stabilization Project;”
NOAA-ARRA: 2009-2012 – Terrestrial Monitoring Component Final Report. Unpublished report submitted to NOAA-CRP.
 Williams, J. J., Dutton J., Chen C.Y., Fisher N.S., (2010). "Metal (As, Cd, Hg, and CH3Hg) Bioaccumulation from Water and Food by the Benthic Amphipod Leptocheirus
Plumuous." Environmental Toxicology and Chemistry 29(8): 1755-1761

similar documents