the proper design of a sediment trap (Gardner, 1980a, b).
This is important, because traps with too large a ratio of
opening to height will undertrap, whereas those that are
taller, with smaller openings, will overtrap. Only traps
with the proper design, beginning with a height: diameter
ratio of about 3 will produce results that may be compared
among sites and locations. In addition, sediment should be
removed from the traps every few days, poisoned, and
archived. Much of the information in the literature derived
from sediment traps, in short, is either undependable or of
use only in a limited geographic area. In addition, the
material caught in a sediment trap represents both
resuspended and net sediment, and these need to be differentiated if a complete picture is to be developed.
There has also been wide divergence in the methods
used to measure sediment in the water column, SPM.
Engineers and water quality workers have developed
instruments, “nephelometers” that basically measure the
attenuation of light through a water column as a measure
of turbidity. The material in the water, however, may be
a richly confusing mixture of alive and dead organic matter, dissolved substances, and sediments of varying grain
sizes, colors, and mineralogy. Consequently, a stable synthetic material called “Formazin” with uniform particle
size is often used as a standard for calibration and reproducibility. The results are Formazin Turbidity Units, or
FTU. In the USA, the EPA has specified a white-light
source and certain aspects of the instrument, to obtain
Nephelometer Turbidity Units, of NTU.
The main attraction of these measurement techniques is
that they are quick and easy, but, like many quick and easy
methods, they often come with inherent errors. Because
the type of sediment and the proportions of plankton and
other organic matter can vary so widely, best practice is
to calibrate the nephelometer against actual SPM values
at each site and then, of course, the values so attained
may not readily be transferred to other sites.
The other method used to determine suspended particulate matter is direct measurement: a known volume of
water is filtered through a pre-weighed fine filter (generally 0.45 µ), and the filter is dried and weighed (the method
is described in Cortes and Risk, 1985). This process takes
more time, but provides precise estimates. Perhaps equally
important, the filters may be saved for later analysis of
mineralogy, micropaleontology, and trace elements.
FTU and NTU should not be used in reef research. The
quality of the data so produced is inadequate for research
purposes, although the ease of acquiring such data may
be attractive. Although NTU have been used by, for example, USEAP to set water quality guidelines, the high SPM
values that characterize many reef sites, and the critical
importance of being able to identify sediment provenance,
mitigates against the instrumental methods in use.
Discussion of sediments in the literature has also suffered from some imprecision. Resuspension rates are not
the same as sedimentation rates. From the point of view
of the coral, an area with constant resuspension of bottom
sediments is just as stressful as net deposition of the same
amount of sediment, yet the two situations represent very
different settings. Corals react differently to fine versus
coarse sediments – in addition, the textural and chemical
properties of sediment are a function, at least in part, of
the mineralogy.
Ideally, then, any study of sediment stress on a reef
should therefore provide information on sediment deposition rates, resuspension rates, and SPM values, as well as
data as to grain-size distribution and mineralogy. The
number of published articles satisfying these requirements
is disappointingly small.
The reasons for this may lie in the fact that much of the
literature on the effects of sediment leaves much to be
desired. Because coral reefs are a geobiological construct,
their study – even more so than other ecosystems –
requires a fundamentally interdisciplinary approach. The
need for this has been a subject of comment many times
(e.g., Risk, 1999), but the tendency still is to a onedimensional approach.
Not only has some of the previous work on sediment
stress been less than ideal, there has been delay in adopting
some useful techniques. That corals incorporate sediment
into their skeletons has been known for decades (Barnard
et al., 1974). Cortes and Risk (1985) used this phenomenon to assess sediment input by decalcifying coral samples
and estimating the insoluble residue. True, it only is effective where the incoming sediment is siliciclastic; the high
variance encountered in Budd et al. (1993) arise from
mixing samples, from sites with predominantly siliciclastic
sediments as background input with samples from carbonate terrain. Davies (1992) showed that corals incorporate
sediments through micro-lesions in the tissue, which in
turn may be caused by sediment stress. Where others have
used this technique to assess sediment stress, it has proven
to be useful. Macintyre et al. (1994) traced the decline of
a Porites-dominated reef in Golfo Dulce, Costa Rica, by
analyzing the mineralogy of sediments trapped in coral
colonies. David (2003) estimated input of heavy metals
from mine tailings in the Philippines, and Edinger et al.
(2008) traced the impact of gold mining on corals of Sulawesi, Indonesia. Brown et al. (1990) reported the incorporation of iron into coral skeletons in Thailand. Recently,
Risk et al. (2009) have suggested that it may be possible
to tease out the mixed influences of sediment and nutrients
by plotting trapped sediment in coral samples (insoluble
residue) against d
15
N of the coral tissue; this is an area that
needs more work.
The corals themselves also may not cooperate with studies designed to show linear responses to sediment stress. In
nature, incoming sediment usually will contain additional
nutrients and plankton, so it represents a potential food
source as well as a potential stressor. Many coral species
are very plastic in their feeding behavior, and have been
shown, for example, to shift to a more heterotrophic feeding
mode when faced with increased sediment. Anthony
(2000) has shown this experimentally, and Risk et al.
(1994) have shown that past metabolic shifts are recorded
in the isotopic signatures in the skeletons. Similarly, Risk
580
IMPACTS OF SEDIMENT ON CORAL REEFS
This is important, because traps with too large a ratio of
opening to height will undertrap, whereas those that are
taller, with smaller openings, will overtrap. Only traps
with the proper design, beginning with a height: diameter
ratio of about 3 will produce results that may be compared
among sites and locations. In addition, sediment should be
removed from the traps every few days, poisoned, and
archived. Much of the information in the literature derived
from sediment traps, in short, is either undependable or of
use only in a limited geographic area. In addition, the
material caught in a sediment trap represents both
resuspended and net sediment, and these need to be differentiated if a complete picture is to be developed.
There has also been wide divergence in the methods
used to measure sediment in the water column, SPM.
Engineers and water quality workers have developed
instruments, “nephelometers” that basically measure the
attenuation of light through a water column as a measure
of turbidity. The material in the water, however, may be
a richly confusing mixture of alive and dead organic matter, dissolved substances, and sediments of varying grain
sizes, colors, and mineralogy. Consequently, a stable synthetic material called “Formazin” with uniform particle
size is often used as a standard for calibration and reproducibility. The results are Formazin Turbidity Units, or
FTU. In the USA, the EPA has specified a white-light
source and certain aspects of the instrument, to obtain
Nephelometer Turbidity Units, of NTU.
The main attraction of these measurement techniques is
that they are quick and easy, but, like many quick and easy
methods, they often come with inherent errors. Because
the type of sediment and the proportions of plankton and
other organic matter can vary so widely, best practice is
to calibrate the nephelometer against actual SPM values
at each site and then, of course, the values so attained
may not readily be transferred to other sites.
The other method used to determine suspended particulate matter is direct measurement: a known volume of
water is filtered through a pre-weighed fine filter (generally 0.45 µ), and the filter is dried and weighed (the method
is described in Cortes and Risk, 1985). This process takes
more time, but provides precise estimates. Perhaps equally
important, the filters may be saved for later analysis of
mineralogy, micropaleontology, and trace elements.
FTU and NTU should not be used in reef research. The
quality of the data so produced is inadequate for research
purposes, although the ease of acquiring such data may
be attractive. Although NTU have been used by, for example, USEAP to set water quality guidelines, the high SPM
values that characterize many reef sites, and the critical
importance of being able to identify sediment provenance,
mitigates against the instrumental methods in use.
Discussion of sediments in the literature has also suffered from some imprecision. Resuspension rates are not
the same as sedimentation rates. From the point of view
of the coral, an area with constant resuspension of bottom
sediments is just as stressful as net deposition of the same
amount of sediment, yet the two situations represent very
different settings. Corals react differently to fine versus
coarse sediments – in addition, the textural and chemical
properties of sediment are a function, at least in part, of
the mineralogy.
Ideally, then, any study of sediment stress on a reef
should therefore provide information on sediment deposition rates, resuspension rates, and SPM values, as well as
data as to grain-size distribution and mineralogy. The
number of published articles satisfying these requirements
is disappointingly small.
The reasons for this may lie in the fact that much of the
literature on the effects of sediment leaves much to be
desired. Because coral reefs are a geobiological construct,
their study – even more so than other ecosystems –
requires a fundamentally interdisciplinary approach. The
need for this has been a subject of comment many times
(e.g., Risk, 1999), but the tendency still is to a onedimensional approach.
Not only has some of the previous work on sediment
stress been less than ideal, there has been delay in adopting
some useful techniques. That corals incorporate sediment
into their skeletons has been known for decades (Barnard
et al., 1974). Cortes and Risk (1985) used this phenomenon to assess sediment input by decalcifying coral samples
and estimating the insoluble residue. True, it only is effective where the incoming sediment is siliciclastic; the high
variance encountered in Budd et al. (1993) arise from
mixing samples, from sites with predominantly siliciclastic
sediments as background input with samples from carbonate terrain. Davies (1992) showed that corals incorporate
sediments through micro-lesions in the tissue, which in
turn may be caused by sediment stress. Where others have
used this technique to assess sediment stress, it has proven
to be useful. Macintyre et al. (1994) traced the decline of
a Porites-dominated reef in Golfo Dulce, Costa Rica, by
analyzing the mineralogy of sediments trapped in coral
colonies. David (2003) estimated input of heavy metals
from mine tailings in the Philippines, and Edinger et al.
(2008) traced the impact of gold mining on corals of Sulawesi, Indonesia. Brown et al. (1990) reported the incorporation of iron into coral skeletons in Thailand. Recently,
Risk et al. (2009) have suggested that it may be possible
to tease out the mixed influences of sediment and nutrients
by plotting trapped sediment in coral samples (insoluble
residue) against d
15
N of the coral tissue; this is an area that
needs more work.
The corals themselves also may not cooperate with studies designed to show linear responses to sediment stress. In
nature, incoming sediment usually will contain additional
nutrients and plankton, so it represents a potential food
source as well as a potential stressor. Many coral species
are very plastic in their feeding behavior, and have been
shown, for example, to shift to a more heterotrophic feeding
mode when faced with increased sediment. Anthony
(2000) has shown this experimentally, and Risk et al.
(1994) have shown that past metabolic shifts are recorded
in the isotopic signatures in the skeletons. Similarly, Risk
580
IMPACTS OF SEDIMENT ON CORAL REEFS