Transect Preparation
Prepare a transect line with markings every 1 m. I use a 1/4" or 3/8" nylon braided
rope, and I prefer a bright color such as yellow. The markings can either be permanent
marker or flagging. I prefer fluorescent flagging, and often use an alternate color
for the 5 and 10m intervals to help in keeping track of where you are on the transect
(Photo 1). The transect can either be a set length (e.g., 100 m) or the length between
permanent markers in the case of permanent transects.
Data Recording
Data is recorded on data sheets (see data sheet). For each 1 m interval, all species
or target species are marked if they are present (1) or absent (0) between the two
marks, below the line; in other words, if the species crosses or intercepts the
vertical plane made by the line segment. Additional data, such as water depth or
sediment type, can also be recorded for analysis of correspondence between species
occurrence and environmental conditions. For a given study, more than one transect
should be utilized. If the study involves a treatment effect versus reference, I
would recommend 400 intervals (preferably as 4 - 100 m transects) for each plot.
The more transects or intervals used, the greater the statistical power in subsequent
data analyses. For quantification of plants in a lake, I would recommend stratified-random
deployment of transects around the perimeter of the lake, and generally deployed
perpendicular to shore. However, transect methods can be adapted for many study
purposes and objectives.
Data Anyalysis
Data analysis can be made as easily as counting the number of intervals species
are present, and divide by the total number of intervals examined. This gives a
frequency of occurrence. For large numbers of intervals, this provides a rough approximation
of percent cover. Before and after analyses or comparisons between transects for
a given species can be made using a two-by-two or chi-square analysis using the
actual numbers of intervals with and without the species. For statistical analyses,
the actual numbers rather than frequencies should be used. For instance, a given
study of the effectiveness of herbicide X was setup in one plot in which four 100m
transects were used as a treatment plot, and a similar setup for a reference plot.
Transect analysis indicated that pretreatment transect counts found 320 intervals
with the target species, and 80 without. Six weeks after treatment, the analysis
noted 200 intervals with the target species, and 200 without. The two-by-two analysis
for the treatment plot would look like this:
The chi-square analysis, using the actual interval counts, indicated a significant
reduction in the presence (or frequency or cover) of the target species, using the
Pearson's Chi-square test (p<0.001).
Data Entry
Data entry for computer can be done by using a "1" for intervals with species present,
and a "0" for intervals in which species are absent; and then utilizing the appropriate
statistical package. Data entry will also facilitate calculation of indices of species
richness, presence/frequency of classes of plants (dicots, monocots, native, exotic),
and can also be used to construct maps of vegetation if transect locations are known.
Bibliography
Madsen, J.D., G.O. Dick, D. Honnell, J. Shearer, and R. M. Smart. 1994. Ecological
Assessment of Kirk Pond. Miscellaneous Report A-94-1, U.S. Army Engineer Waterways
Experiment Station, Vicksburg, MS.
Madsen, J.D., L.W. Eichler, J.W. Sutherland, J.A. Bloomfield, R.M. Smart, and C.W.
Boylen. 1993. Submersed Littoral Vegetatgion Distribution: Field Quantification
and Experimental Analysis of Sediment Types from Onondaga Lake, New York. Technical
Report A-93-14, U.S. Army Engineer Waterways Experiment Station, Vicksburg, MS.
Madsen, J.D., J.W. Sutherland, J.A. Bloomfield, L.W. Eichler and C.W. Boylen. 1991.
The decline of native vegetation under dense Eurasian watermilfoil canopies. Journal
of Aquatic Plant Management 29:94-99.