This overview is designed to give a brief description of potential aquatic plant
measurement methods that might be employed in an aquatic plant monitoring program.
The major methods are discussed in much greater detail in the companion volume,
published as a series of papers in the Journal of Lake and Reservoir Management.
However, additional qualitative methods are discussed to provide direction for potential
approaches at all stages of an aquatic plant monitoring program. The methods presented
are summarized in the Summary of quantification techniques section. When either
an agency or group is affiliated with a given type of program, the address and contact
person are both indicated in
appendix 1.
Plant Identification Program
A plant identification program would entail central or regional locations to which
citizens could send unknown aquatic plant samples for identification. Both the State
of Vermont and New York State Department of Environmental Conservation (NYS DEC)
have informal arrangements of this sort. The Rensselaer Fresh Water Institute (RFWI)
has conducted a formal program since 1987. The program is publicly advertised and
records were kept of samples received. The focus of the program was Lake George,
with a secondary area of concern being northeastern New York State, but all specimens
sent in were identified. Before 1987, plants were identified by RFWI on an informal
basis. All of the above programs are free to the public, with the purpose of locating
new sites with exotic plant species while also providing beneficial public relations
and allaying, or confirming, public fears concerning the locations and spread of
exotic aquatic plant species.
Although this sounds like a good system for locating new infestations of exotic
species, it has several drawbacks. First, large numbers of nontarget species are
sent in for every target species identified. In the first two years of the RFWI
program, less than half (42.9%) of the specimens identified were of a target species
(Madsen and Taggett, 1990). Second, search locations are completely nonsystematic,
with coverage completely dependent on the interest or concern of a few individuals.
Third, individual locations will be identified and reported repeatedly. Last, these
locations would need to be visited by a trained person once the sighting is made
for effective follow-up, so some minimal level of commitment must be made to a formal
program. However, the public relations benefits of such a program alone probably
justifies the effort made to maintain it, given that a minimal commitment of funding
to a centralized coordinating agency and "resident expert" is made. The organizing
agency and expert would be needed for other programs, as well. In addition, this
type of program will assist in identifying new infestations of exotic species, and
could be used for other target species of concern (e.g., zebra mussel).
Citizen "Plant Watcher's" Program
A citizen "Plant Watchers" program is one in which lay persons are trained to identify
and search for exotic species of concern. The locations and extent of these populations
are qualitatively recorded, and records sent to a central organizer. This type of
program could be modeled after the successful program used by the State of Vermont
Milfoil program (see Appendix 1). The "Milfoil Watchers Program" in Vermont trains
volunteer citizens to identify the target species, and follow appropriate procedures
to search for this species. Plants thought to be Eurasian watermilfoil are sent
to the program organizer for verification, along with field notes identifying locations.
Most commonly, lake groups or individual homeowners would volunteer to monitor their
own lake. This program, not unlike a citizens' lake assessment program, helps indicate
if a new species arrives at the lake and might provide "early warning" for a new
infestation, early enough for simple yet effective control measures to be taken.
This type of program would also ensure that a given area is effectively patrolled
for the presence of nuisance exotic species, and provide at least anecdotal information
on areal extent or spread. An organized plant watchers program is the best use of
citizen volunteers, and is strongly recommended as a preferred component of a regional
management program.
Citizens Semi-quantitative Monitoring
In this type of program, citizens are trained to semi-quantitatively sample areas
for the presence and abundance of vegetation, and send the samples in for analysis
or perform some evaluation of abundance themselves. This type of monitoring has
recently been incorporated into the Citizen Statewide Lake Assessment Program (CSLAP)
in New York State, jointly operated by NYS DEC and the New York Federation of Lake
Associations (see Appendix 1).
Although this type of program is preferable to no systematic searches or organized
effort, we think that it is almost impossible for citizens to gather reasonably
consistent quantitative data to be useful in monitoring and assessment. The effort
expended is therefore misdirected, and should rather be directed towards a qualitative
plant watchers program. When needed, trained technicians should be employed to gather
quantitative data for monitoring and assessment. These technicians may work for
such as private consulting groups, universities, and municipal or state agencies.
Semi-quantitative methods that have been employed by volunteer citizens include
rake-sample surveys (e.g., CSLAP) and quantifying plant height using echolocation
gear (i.e., fish locators). Such efforts might assist the local lake association
in planning for management needs, but does not replace the need for competent assessment
in larger management programs.
Plant Vouchers
With any program involving the identification of aquatic plants, the importance
of maintaining a collection of voucher specimens cannot be overstated. Vouchers
must be collected and maintained by experts as a record for the future, and to ensure
proper identification and verification of specimens. Hellquist (1992) provides an
excellent overview of this step in an aquatic plant monitoring program. The hazards
of neglecting this step could result in confusion over the identity of nuisance
versus native species, and cause questions over the timing of exotic species invasions
(Newroth, 1992).
Semi-quantitative Diver "Swimover"
Although it may be helpful to get an "expert" to look at your lake or plant problem
from a boat, an adequate assessment of the aquatic plant community, particularly
submersed vegetation, simply cannot be done from observations at the water surface.
Although a boat survey might be necessary to ensure coverage of a large proportion
of a lake, aquatic plant assessments must include diver swimovers; snorkeling at
a minimum; SCUBA diving preferable. Aquatic vegetation assessments should include
a comprehensive, systematic coverage of the lake, a comprehensive survey of all
depths in which rooted vegetation can grow, and a comprehensive list of species
and assessment of distribution and abundance. To adequately fulfill this need, some
diving will be necessary. Since diving is slow and cannot cover the entirety of
most lakes, systematic site selection is necessary to cover both habitat types and
geographic areas of the lake. The diver simply cannot jump in at one or two places
and get adequate coverage of the lake.
The absolute minimum effort by a trained technical crew is a diver "swimover" survey.
This type of survey will provide both a comprehensive species list and data on the
distribution and abundance of species. A swimover survey must be done by individuals
trained in plant identification and diving, and familiar with this sort of technique.
Divers select a number of transects that are representative of habitats (slope,
exposure and bottom type) and geographic areas of the lake (Photo 1).
|
|
Photo 1
|
These areas or transects are then examined by divers. One methodological approach
is to swim back and forth across the selected area, and note the occurrence and
abundance of all species by depth interval. Divers note both the minimum and maximum
depth of the occurrence of any target (e.g., nuisance) species and of all rooted
species. One method for estimating the abundance of observed species is the commonly
used abundance scale indicated in Table 2. However, there are several variations
of this scale. In tandem with a diver swimover survey, trained personnel might circumnavigate
the lake looking for the occurrence and distribution of nuisance target species
approaching the surface. Sketch maps of this information are also useful for management
decision-making as a first step.
|
|
Table 2
|
If enough transects are done, this technique approaches the usefulness of quantitative
data in that the occurrence of species is systematically noted around the lake.
However, this does not replace quantitative data collection. A diver swimover is
often a good first step in developing an information base for lake management decisions,
and provides some data on species presence and relative abundance for the purposes
of permit applications. It is especially appropriate for small lakes that do not
have large lake management budgets. Examples of diver swimover surveys that were
conducted as part of the RFWI Lake Assessment Program are available from the Rensselaer
Fresh Water Institute (e.g., Eichler, 1990). These assessments are primarily done
for small lakes on small budgets.
Transect Methods
Transect methods are primarily utilized to quantitatively determine the distribution
and relative abundance of aquatic plants as part of a monitoring program. Transect
methods should only be attempted by technicians trained in both aquatic plant identification
and SCUBA. Although transects can be done by snorkeling (Photo 1), SCUBA is the
best approach. Transect methods are useful for determining the distribution of native
and nuisance plants in a lake, long-term monitoring of populations, and mapping
the expansion of nuisance plants. Transects are also used to develop vegetation
maps in lakes. Although they have some applications to assessment of control techniques,
they are most useful for large-scale assessments of communities.
Titus' (1992) paper on transects gives excellent illustrations on the many ways
that transects can be utilized. Titus prefers transects of contiguous quadrats (0.2
m
2) in which species presence is recorded (Photo 2), with transects perpendicular
to shore, or parallel to shore at discrete depth intervals. We prefer perpendicular
transects, with either a line-intercept approach or quadrats (0.1 m
2) with species
presence or absence. Either method gives good results.
|
|
Photo 2
|
Transect methods are more costly, but give more quantitative and objective results
than diver swimovers. A diver swimover survey of a small lake might take two man-days,
while a transect study will take at least 8 to 10 man-days.
Transect methods have been very useful in monitoring populations over extended periods
of time. Given adequate documentation, the transects can be revisited in subsequent
years, and the data compared for changes. Permanent transects are especially useful
for this purpose. For instance, permanent transects and a grid system of contiguous
quadrats were used in Lake George to document the spread and increased density of
Eurasian watermilfoil, with the concurrent decline of native species (Madsen et
al., 1991).
Biomass
Biomass sampling also requires trained technicians to produce adequate results.
Biomass sampling is much more intensive than the other methods, but provides precise
information on plant abundance within a given area. Because of the intensity of
sampling, biomass plots are very localized. Biomass is not suited for large-scale
surveys of lakes, but rather for vegetation patterns within individual bays or areas.
Because of the intensity of effort required, we would not recommend biomass sampling
for routine monitoring. However, it is the method best suited to finely quantify
the effectiveness of most control techniques, so it is especially suited to assess
management tactics.
Biomass should be performed by individuals trained in SCUBA and plant identification
(Photo 3). We have used trained SCUBA divers who were not trained in plant sampling
procedures before, but found they were more expensive than necessary and needed
continual supervision anyway.
|
|
Photo 3
|
Biomass methodologies and approaches are described more extensively by Madsen (1992).
In brief, we recommend that management tactic sites be evaluated using at least
one control (untreated) and one treated site; although it is far preferable to have
more than one site of each (see Spencer, 1992). The number of biomass samples taken
at each site should be statistically evaluated for adequacy; a good starting point
for preliminary sampling is ten samples per site. Samples should be taken following
a stratified-random pattern using a 0.1 m
2 quadrat or sampler. Larger samplers are
probably not necessary, and engender more labor in sorting. Samples should be oven-dried
and weighed, with appropriate statistical tests and transformations applied.
Manipulative Experiments
Manipulative experiments are less a method in and of themselves than an approach
to take when examining the effectiveness of a given management tactic. The experimental
approach is reviewed by Spencer (1992). Manipulative experiments are difficult to
set up and require statistical sophistication, but are much more powerful in terms
of the statistical inference in evaluating treatments. A more critical and analytical
approach can be taken with a management tactic using manipulative experiments; thus
they are worth the trouble if experts are available. Although all quantitative techniques
can be used to gather data in a manipulative experiment, biomass sampling is the
most common one employed given its greater power of resolution.
Remote Sensing
Remote sensing was attempted for quantifying vegetation using various satellite
and 35 mm film formats and sensors, but results were often very disappointing. However,
with the introduction of new technologies and new sensors that employ videotape
equipment, remote sensing may finally yield the types of results originally expected.
However, the application of these new technologies and techniques is still in the
research and development phase. Cibuzar (1992) gives an overview of the development
and application of one example of the new technology in the symposium volume.
Remote sensing is one of the quantitative tools that would be useful for both studies
of nuisance plant distributions and for comparisons of seasonal and inter-annual
comparisons of growth. Although it is most useful for monitoring purposes, it may
also be used to evaluate large-scale management tactics. However, there are still
drawbacks to remote sensing that will not allow it to totally replace field sampling.
"Ground truth" tests are still required to identify plants. As yet, remote sensing
cannot differentiate completely between species, cannot assess the abundance of
plants nor the depth to which they occur, and cannot identify plants when they are
below the surface any significant extent.
Simulation Modeling
Simulation models of aquatic macrophyte growth and the effects of various management
tactics have been attempted since the early 1970's (Titus et al. 1975). The recent
modeling efforts are probably centered in the attempts within the U.S. Army Engineers
Aquatic Plant Control Research Program at Waterways Experiment Station, Vicksburg,
MS. These efforts are coordinated by R. Michael Stewart (Appendix 1). Simulation
modeling efforts will continue in the future on Eurasian watermilfoil, Hydrilla,
and Waterhyacinth (Stewart, 1990). First generation models are completed for Hydrilla
and Eurasian watermilfoil (Wooten, 1990) as well as for several particular management
applications (Boyd and Stewart, 1990; Clifford et al., 1990). Future work will focus
on generalizing first generation models to new environments.
Simulation modeling has many attractive features. These models may become a powerful
tool in the "what if?" approach, of predicting the results of manipulations, giving
insight to management alternatives, and in asking appropriate management questions.
However, models will not replace sampling to evaluate the results of management
activities or other routine monitoring. On the contrary, modeling will require more
intensive monitoring and assessment in both the development and operational phases
of their use.
Expert Systems
Expert systems are beginning to receive a great deal of attention in all areas in
which novices or citizens need to benefit from the expertise of a few individuals.
An expert system is a computerized menu-driven system in which decision makers are
led through a hierarchical chain of decisions. On the basis of their answers to
simple queries, the program assists them in making the best decision based on the
experience of the systems designers. The concept shows great promise for making
lake management more consistent with the best management practices of the time,
as well as indicating areas in which more expertise is required. However, these
systems are very early in research and development for aquatic plants. One such
system is currently under investigation at the U.S. Army Engineers Aquatic Plant
Control Research Program at the Waterways Experiment Station (Lawrence and Lemmon,
1990).
However, there will always be shortcomings with any expert system. They are limited
by the experience of the experts involved in their development, and in the programming
available. Also, they rely on over generalizations to allow for classification of
situations. No two lakes or situations are completely identical, so there will always
be a role for site visits by experts. However, the expert system promises to be
another tool of promise.