Gulf of Maine Projects
Final Report: Evaluation of the Gulfwatch
Monitoring Program
Gulfwatch Review (cont.)
Future Program Design Modifications
Hypothesis Testing
Now that direct field experience has been acquired and the strengths
and limitations of this approach are specifically known, it is appropriate
to return to ask a specific testable question. This is a critical step,
from which the entire program should develop. There are differences between
monitoring to measure compliance with existing requirements (e.g., to
meet outfall permit requirements) and monitoring to determine spatial
and temporal trends in contaminant distributions. Both are important activities
but data collected for one purpose are usually not applicable for another
and the goals for data collection must be clearly stated at the outset.
The incentive to overuse monitoring data will be reduced if both data
collectors and resource managers clearly understand the inherent limits
on any data set. These limits become apparent if clear statements of purpose
are included with the data.
During this just-completed initial implementation phase of Gulfwatch,
a necessary focus on the mechanics of starting a new program has somewhat
obscured this very basic issue. The Gulfwatch program has responsibly
evolved in response to realities of the field sampling but this response
has focused on the logistical and programmatic details while losing sight
of the higher level mission questions. The program has reacted to logistical
issues and made methodological adjustments well. It is now an appropriate
point in time to use the 5-year experience base to reassess program mission
and as well as methods.
Sampling in Environmentally Relevant Scales
As Gulfwatch evolves, and if it expands in scope beyond tissue
sampling, samples need to be taken with an understanding of ecosystem
complexities and the processes governing variability in each compartment
of the ecosystem if meaningful data are to be generated. For example:
Water : The coastal water mass, driven by stormand tidal energy,
is changing vertically and horizontally on the time scales of minutes-to-hours;
extremely low concentrations of toxic organics in sea water cannot be
detected easily by standard analytical methods.
Biota: Tissue concentrations can vary with physiological state,
age and health of the organism and season. The organism's location in
the food web and habitat characteristics can have an effect on contaminant
distribution and concentration.
Sediment : Concentrations of organic contaminants in sediments
are usually correlated with the organic carbon content and grain size
distribution; therefore a sampling strategy for organics should first
assess such bulk parameter distribution. Muddy, high carbon sediments
are usually a long-term reservoir of toxic contaminants and should be
the primary sampling focus. For nutrient-related issues, sediments are
the site of major nutrient transformation within the system and more frequent
sampling may be required to resolve the rates and extent of nutrient biogeochemistry.
A sampling strategy involving sediments and may require sampling in different
spatial and temporal scales to resolve issues of toxics, nutrients and
pathogens.
One specific Gulfwatch issue related to sampling with respect
to environmental complexities relates to clearing of the gut prior to
analysis. Evidently, the program decision not to allow the guts to clear
was to provide parity with NS&T as well as to simplify the procedure.
This decision presents a problem in areas with high suspended sediment
loads. The observed order of magnitude difference in aluminum (Al) concentrations
clearly represents something other than body burden but this question
cannot be resolved with existing data. Perhaps this is an area where further
testing could be done; one possibility is to routinely allow the caged
mussels to clear their guts before analysis (if this aspect of Gulfwatch
continues; see discussion below) and then test deploy them in areas where
high suspended loads are suspected to be causing a problem. If this alternative
is attempted, possible loss of organics with gut clearing must be considered.
As collected, the caged mussel data are not directly comparable to native
mussels for some questions, in any case. This gut-clearing issue is also
related to data that was omitted from analysis because the iron (Fe) and
aluminum concentrations were unexpectedly high. The logical explanation
offered was that gut contents of sediment were probably high. Perhaps
the tissue data could be salvaged by analysis of the Metal/Al or Metal/Fe
ratios? It would also help the interpretation of metal concentrations
to know the concentrations in bottom and suspended sediments at the sampling
site. A reassessment of the program should include reconsideration of
sediment sampling in some cases.
Gulfwatch and NOAA
National Status and Trends (NS&T) sampling periods are offset
by 6 months. Previously published studies have shown that seasonal patterns
exist in mussel metal body burdens and concentrations (Phillips, 1976,
1980 ). Mucklow (1996) reports a similar result in the Gulf. For most
metals, Gulf concentrations were higher in the spring than in fall. In
addition, differences in the physiological state of the mussels in spring
vs. fall may affect contaminant burden and there are also likely to be
seasonal differences in inputs (although the NS site studied by Mucklow
was located near a sewage outfall which may have been a relatively constant
local source of metals). Mussels removed some distance from local sources
of pollutant may reflect contaminants from distant sources in the spring
when river discharges are at their highest and therefore show an even
greater seasonal change in concentration. These environmental and physiological
factors make the direct comparison of data between NS&T and Gulfwatch
difficult. In addition, there appear to be some differences in the analytical
methods which are discussed below. One suggestion might be for Gulfwatch
to sample the NS&T sites more often in one year in order to provide
information on seasonal variation at these sites and allow for a direct
comparison of data with the labs running the NS&T samples. Gulfwatch
certainly should regularly participate in the inter laboratory comparison
exercises conducted by NOAA.
Benchmark stations were chosen in an effort to quantify interannual
variability. It would seem best to choose these stations to represent
fairly clean sites which respond to regional scale pollution inputs. There
may also be strong reasons to choose benchmark stations in heavily contaminated
areas to address specific management questions; for example, Deer Island
would have value as a benchmark for many questions relevant to the Gulf.
By monitoring these benchmark stations annually there should be a reasonable
chance of picking up regional trends in 3-5 year time scales. However,
as presently designed, the number of benchmark stations is fairly small
and it's not obvious that they are sufficiently removed from local contamination
sources that might skew their value as representative of the Gulf-region.
It would be good to consider moving some of these stations and increasing
the number if this variability detection objective is to be met.
Reviewers note that the reported limit of detection for most analytes
is higher than that of NS&T. Trace concentrations in samples require
more elaborate, often non-standard analytical methods that account for
analytical artifacts and recovery levels (e.g., high resolution, low detection
limit, negligible "blank" concentrations, etc.). This issue
must be resolved by Gulfwatch through improved QA/QC practices and inter
laboratory comparison before data can be produced in relatively clean
areas to resolve the interannual question. A regional program such as
Gulfwatch has an opportunity to focus on regional issues such as long-distance
atmospheric transport that local programs, with their inherently narrower
focus, normally cannot. Some contaminants have a distant source (Pb from
gasoline, PCBs, etc.) as well as local sources and by selecting an adequate
number of stations, these multiple sources might begin to be sorted out.
Such questions are prime candidates for partnerships between a monitoring
program and academic research efforts.
Overall, the issue of benchmark stations should be re-examined
by Gulfwatch. By only monitoring the other sites every 3 years it will
take a decade or more to determine if an individual site is getting cleaner
or more polluted. This seems to be an acceptable time-frame, given the
regional nature of the program and the current funding level, but it also
implies a continued long-term source of program support.
Links to On-going Research
The bivalve monitoring concept has been successfully tested at
local, regional, national and international scales for more than three
decades. The best of these programs have allocated adequate resources
for communication of results and for QA/QC activities. These successful
programs have also kept up-to-date with these activities by maintaining
close links with academic researchers who are also working in the area
of coastal environmental quality. Successful monitoring programs are designed
to make as simple a measurement as possible, but monitoring measurements
must be interpreted in the context of ecosystem complexities. Data interpretation
and the decision of when to move up to more expensive measurements would
benefit close dialog with active research projects. The search for simple
answers to complex questions is a recipe for failure. Newer methodologies
and techniques may be necessary to address specific questions and monitoring
program links to academic research is essential if the program is to make
use of (or even be aware of) these new methods. Interesting peripheral
questions and some of the "high-tech" analyses can be addressed
through partnerships between the monitoring program and academia.
Gulfwatch is ideally situated to build these connections but does
not seem to have done so to date because of the initial phase focus on
start-up details. These links need to receive much higher priority as
the program matures and will require program resources.
Back to General Review | On
to Sampling and Analysis
To obtain a printed version
of this report, please download this document. You will need Adobe
Acrobat 3.0. It's easy to use and available
free
|