HEAVY METAL POLLUTION IN ROAD DEPOSITED SEDIMENTS,
PALOLO VALLEY, HONOLULU, HI
Joseph O. Bussen, Ross A.
Sutherland (Geomorph. Lab., Dept. of Geography, Univ. of Hawaii, Filip M.G.
Tack (Lab. of Anal. Chem. and Applied Ecochem., Univ. of Ghent); email: bussen@hawaii.edu.
Abstract
Road
deposited sediments were collected from Palolo Valley, Honolulu (n=20), a
source for the very polluted Ala Wai Canal estuary. Chemical properties and
particle size distributions were measured.
Spectrometrically determined total digestion concentrations for 18
metals yielded a median moderate pollution level for Cu, a median significant
pollution level for Zn, and a median extreme pollution level for Pb.
Concentrations from a 0.5 M HCl leach of the labile and potentially bioavailable
phase for 24 metals gave a median very high pollution level for Cu and a median
extreme pollution level for Pb and Zn.
Four-phase sequential analysis showed that 65% of the Cu, 92% of the Pb,
and 79% of the Zn are in the non-residual phases. Palolo is smaller in area and
in population than the better-studied adjacent Manoa Valley, but it appears to
be significantly more polluted with heavy metals.
Introduction
Sediments
deposited on urban streets are repositories of natural and anthropogenic materials
transported from basin slopes to the streets by water and wind and derived from
road wear and vehicle emissions and wear. Such road deposited sediments (RDSs)
that are not removed by street sweeping operations are easily mobilized and
transported through storm drains into streams. Heavy metals and other
pollutants affect water quality and are potentially toxic to stream biota. Humans may ingest these pollutants directly
(e.g., by inhalation of particles or ingestion by children), or indirectly
(e.g., by eating contaminated fish).
The
Ala Wai Drainage Canal through the former marshes of Waikiki, perhaps the most
polluted waterway in Hawaii, is fed primarily by the Manoa-Palolo Drainage
Canal, carrying the runoff from highly populated Manoa and Palolo Valleys
(Figure 1). There continues to be much concern about pollution in these
waterways and their contributing areas. Most attention has been given to the
Ala Wai Canal and Manoa Stream. This
study focuses on the Palolo Valley component of this system.
In
three surveys of heavy metals in whole fish in 109 streams across the U.S., the
highest levels of Pb were found in Manoa Stream; copper levels were also very
high (Schmitt and Brumbaugh, 1990). Sutherland (1999, 2000) found moderate
contamination signals for Cu and Zn and a very strong signal for Pb (up to 1080
mg/kg) in Manoa Stream bed sediments. In unpublished work on Palolo Stream
sediments, very high concentrations of copper (2670 mg/kg), lead (1385 mg/kg),
and zinc (1230 mg/kg) were found at one site.
Two
recent studies of background, roadside soil, and RDS metal concentrations in
Manoa Valley are Sutherland et al. (2000) and Sutherland and Tolosa (2000).
These studies found anthropogenically enriched amounts of Cu, Zn, and
especially, Pb, in roadside soils and RDS, which are attributed to emissions
and wear from automobiles. In this study
the RDS from 20 sites in the adjacent Palolo Valley are analyzed.
RDS
contamination has been more widely studied in recent years. Stone and Marsalek
(1996) studied street sediments in Sault Ste. Marie, Ontario. They found total
metal content (mg/kg) of Cu, Pb, and Zn to be 87, 91, and 227, respectively.
Kim et al. (1998) analyzed road dusts from the industrial city of Taejon, one
of the six largest cities in Korea. They found total digestion concentrations
(mg/kg) of Cu, Pb, and Zn of 47 (11-219), 60 (22-160), and 172 (75-336), for 9
“industrialized” samples; 57 (17-226), 52 (13-161), and 214 (67-495), for 31
“traffic” samples, and 24 (10-347), 28 (12-115), and 107 (59-323) for 41
“control” samples. Bris et al. (1999) used a wet vacuuming technique to collect
street samples from the center of Paris. They report “high contamination,
especially for copper and zinc”. Their “gutter” concentrations (mg/kg) of Cu,
Pb, and Zn were 200 (100-1700), 3300 (900-6100), and 2000 (1500-4600),
respectively. Sutherland and Tolosa (2000) list mean Pb concentrations (mg/kg)
from seven RDS studies in suburban, residential and low traffic areas: Libya,
417±187; London, 399±67; Ecuador, 298±288; Egypt , 293 (53-513); Nigeria, 172 (51-297);
Canada, 91±55; and Manoa, Hawaii,
151±116. A recent street
sweeping study in another Honolulu neighborhood yielded moderate concentrations
(mg/kg) of Cu (120±68), Pb (46±24), and Zn (184±168).
Study Area
Palolo
Valley has an area of 10.9 km2, of which about 68% is undeveloped
and in forest reserve. There is some remnant agriculture (e.g., flower farms)
in the back the valley. There are very few service stations or automobile
repair facilities in the valley, although a few “back-yard” type operations can
be seen. In 1990, Palolo had 4097 households and a resident population of
13034. Of 2640 housing units in Palolo
in 1980, 80.1% were single family, 10.6% apartment buildings, 8.6% duplexes,
and 0.7% townhouses.
Methods
Samples
were collected from storm sewer inlets at 20 sites (Figure 1) with an
acid-washed Nalgeneä scoop, oven dried, and sieved through a 2000 mm nylon sieve.
Small subsamples were ground and sieved through a 125 mm nylon sieve. One aliquot from each site was
subjected to a total multi-acid digestion and then analyzed for 18 metals by
ICP-AES and for Pb by FAAS. Precision (coefficient of variation) and accuracy
(means compared to a standard) of this series were tested with four replications
of certified reference material (GSD-7)
stream sediment. Precision was excellent (2.35% or better) for all elements
except Cd; accuracy was good (3-5%) for the metals of most concern to us (Al,
Cu, Pb, Ni, Zn). A second aliquot was treated with a 0.5 M HCl leach, often
used as an indicator of potential bioavailability (Stone and Marsalek, 1996),
and then analyzed for 24 metals by ICP-AES. A third subsample was partitioned
into operationally defined acid-extractable (acetic acid), reducible (NH2OH×HCl), and oxidizable (H2O2)
fractions using the optimized BCR 3-step procedure (Rauret et al., 1999), and
the residual fraction was digested with aqua regia. Several RDS properties were measured: pH and conductivity (5.0 g of < 125 mm sediment was mixed with 25 ml distilled water and
shaken 30 minutes with a mechanical shaker), cation exchange capacity (CEC) by
the ammonium/sodium acetate method, and percent organic carbon (%OC) by chromic
acid digestion.
Particle
size fractions were measured by the hydrometer method (Gee and Bauder, 1986)
with the coarser fraction put through a set of sieves (1000, 500, 250, 125, and
63 mm) for 20 minutes in a
mechanical shaker. Percentages of sand, silt, and clay were determined.
A pollution index was calculated for both HCl
and total digestion, using an Enrichment Ratio (ER) relative to Al, and a
ranking system, after Sutherland (1999):
where CnRDS is the concentration of
element n in a given fraction, CAlRDS is the concentration of Al in
the fraction, CnBackground is the element’s concentration in Koolau
basalts (100, 190, 6, and 110 mg/kg for Cu, Ni, Pb, and Zn, respectively), and
CAlBackground is the Al concentration in Koolau basalts, 7.7%.
Categories of enrichment/pollution level (e/p) are assigned as follows:
ER<2=depletion to minimal e/p, ER 2-5=moderate e/p, ER 5-20=significant e/p,
ER 20-40=very high e/p, and ER>40=extreme e/p.
Results
Sediment
chemical variables pH, conductivity, CEC, and %OC were determined for all 20
samples. Medians and median absolute deviations are shown in Table 1. Total
digestion concentrations and HCl extractable concentrations for selected metals
are also shown in Table 1. The highest four concentrations found for the three
metals are listed in Figure 1, linked to the sites where they were found. Mean
ERs for the three metals under both digestions are also shown in Table 1. The
labile HCl extracted phase shows a very high e/p for Cu and an extremely high
e/p for Pb and Zn. Results from Manoa RDS are included in Table 1 for
comparison. Sequential extraction data for Ni (lithogenic
control) and Cu, Pb and Zn (anthropogenic control) are shown in Table 2.
Road sediments were predominantly sand (63.8 ± 9.3 %), followed by silt (29.9 ± 6.8 %) and clay (6.3 ± 3.7 %). Scattergrams were generated to examine
relationships between sediment properties, sand/silt/clay fractions, metal
concentrations and enrichment ratios. No correlations with particle size or
between chemical properties and metal concentrations were apparent. Zinc and Cu concentrations show some
relationship (HCl leach, r2=0.56), but neither were significantly
correlated with Pb at a=0.05.
Discussion
This study confirms the prediction of high concentrations
and pollution indices for Cu, Zn, and especially Pb, in RDS in Palolo Valley.
Compared to the results of the pilot studies of Palolo Stream sediments,
extreme Pb and Zn concentrations (but not the medians) are even higher in the
RDS (Figure 1). The higher metal concentrations at certain sites do not seem to
be connected to traffic density. The highest suite of values at site 14 is
puzzling, since this site may have the smallest automobile impact. Perhaps this location simply has a longer
residence time, allowing for additional storage of RDS, despite lesser inputs.
Compared to Manoa RDS, Palolo RDS concentrations and ERs (total and HCl) are
higher in every instance (Table 1). The contrast is greatest for the more
labile HCl fractions, i.e., in the presumably more bioavailable fractions. The extremely high Pb concentrations and ERs
are especially notable. Compared to the one street sweeping study available for
urban Honolulu, Palolo concentrations are much higher, especially for Pb.
Palolo RDS total digestion concentrations are considerably higher than those of
Stone and Marsalek (1996) for Sault Ste. Marie and Kim et al. (1998) for
Taejon, but do not approach the extremely high concentrations of Bris et al.
(1999) for Paris. Compared to the Pb concentrations from suburban, residential,
and low traffic areas around the world listed by Sutherland and Tolosa (2000),
Palolo mean concentrations are smaller than those from Libya and London, about
the same as those from Ecuador and Egypt, and considerably greater than those
from Nigeria and Canada.
The 0.5 M HCl digestions yielded Pb, Cu, and Zn
concentrations that were 88%, 61%, and 72%, respectively, of their total
digestion concentrations, confirming the predominance of these
anthropogenically controlled metals in the labile phase. The corresponding concentrations of Al and
Ni, in contrast, were just 6% and 9% of their total digestion concentrations,
as expected for lithogenically controlled elements. Such elements are primarily bound within the crystal structure of
primary and secondary minerals, and are not likely to be released in solution
over short time intervals (Tessier et al., 1979). Sequential analysis showed
64.7% of the Cu, 91.8% of the Pb, and
78.8% of the Zn in the non-residual phases. Further study of these labile fractions may help us understand
the different biological pathways for uptake of heavy metals into organisms.
The correlation between Cu and Zn (but not Pb) is intriguing. Further work may show that these elements
are related in some way to present or past automobile, light industry, or
agricultural usage.
This
study was limited to RDS of Palolo Valley. It is planned to further examine
both the inputs from the valley slopes and the outputs in the sediments of
Palolo Stream and its tributaries.
References
Bris F-J, Garnaud S, Appery N, Gonzalez A, Mouchel J-M,
Ghassan C, and Thevenot D (1999), Sci. Tot. Environ. 235:211-220.
Gee GW, Bauder JW (1986), In: Methods of Soil Anal.,
Pt 1. Phys. and Minerolog.
Methods—Agron. Mon. no. 9, Madison, Am. Soc. of Agron.—Soil Sci. Soc. of Amer.,
chap. 15.
Kim KW, Myung J-H, Ahn JS, Chon H-T (1998), J.
Geochem. Explor. 64:409-419.
Rauret G, López-Sánchez JF, Sahuquillo A, Rubio R,
Davidson C., Ure A, Quevauviller Ph (1999), J. Envir. Monit. 1:57-61.
Schmitt C J, Brumbaugh WG (1990), Arch. Env. Contam.
Tox. 19:731-747.
Stone M, Marsalek J (1996), Water, Air, Soil Poll.
87:149-169.
Sutherland, RA (1999), In: Proc. of IUGG 99 Sym. HS5,
(JB Ellis, Ed.) Wallingford, UK, IAHS Press, pp. 269-278.
Sutherland, RA (2000), Env. Geol. 39(3-4):330-341.
Sutherland RA, Tolosa CA (2000), Environ. Poll. (in
press).
Sutherland RA, Tolosa CA, Tack FMG, Verloo MG (2000),
Arch. Env. Contam. Tox. 38:428-438.
Tessier A, Campbell PGC,
Bisson M (1979), Anal. Chem. 51(7):844-851.
