Network Analysis of the St. Marks Wildlife Refuge
Seagrass Ecosystem.
Network Analysis
This web page provides the raw data for a seagrass ecosystem network
analysis, which is a series of mathematical algorithms for estimating
steady-state material flows among ecosystem compartments in a food-web
network. These network analyses have been developed for the Halodule-wrightii
seagrass
ecosystem in the papers:
Baird, D., J. Luczkovich and
R. R. Christian. Assessment of spatial and temporal variability in ecosystem
attributes of the St Marks National Wildlife Refuge, Apalachee Bay, Florida.
Estuarine, Coastal, and Shelf Science 47: 329-349.
Christian,
R. R. and J. J. Luczkovich. Organizing and understanding a winter's
seagrass foodweb network through effective trophic levels. Ecological Modelling
117: 99-124.
Right: A simplified food-web diagram showing
some of the carbon-flow links in the St. Marks National Wildlife Refuge
seagrass ecosystem. The producer base of this sub-web shows shoal grass
(Halodule wrightii); primary consumers shown are various species
of herbivores and detritivores including meiofauna (represented here by
a harpacticoid copepod and a nematode), amphipods, and polychaetes; secondary
consumers shown are juvenile spot (Leiostomus xanthurus) and pinfish
(Lagodon rhomboides); tertiary consumer shown is the gulf flounder
(Paralichthys albigutta). The full model has many more compartments
than those shown in the figure. In fact, the network models that we have
constructed present quantitative estimates of the amount of carbon flowing
between 51 compartments during the winter in the St. Marks seagrass ecosystem.
Using KINEMAGE to visualize the St. Marks Food web:
The 3D network and molecule visualization software KINEMAGE was used
to display the food web from St. Marks. The method used is detailed
in the paper below:
Jeffrey
C. Johnson, Stephen P. Borgatti, Joseph J. Luczkovich, and Martin G. Everett.
2001. Network Role Analysis in the Study of Food Webs: An Application
of Regular Role Coloration. Journal of Social Structure, Vol. 2, No. 3,
May 8, 2001. (note this is an on-line peer-reviewed journal).
Click here to view the food web
in KINEMAGE
Rotate the kinemage by moving the cursor arrow into the kinemage figure
(the graph), left-clicking the mouse, and moving the cursor arrow in the
direction of the preferred rotation. If the graph is oriented with
top predators (blue) on top and primary producers (green and yellow)
at the bottom, then the vertical dimension of the graph basically reflects
the trophic position of a species or compartment. Keeping this trophic
orientation in mind, you can animate the graph to reveal the structure
of the individual isotrophic classes. Left click in any of the boxes to
the right, which can turn on the labels for each compartment. Start
with all the boxes unchecked. Next, click on "Estuary", "Biomass",
"In Flux", "Out flux"; this will show the biomass fo each compartment,
along with arrows representing the flow of carbon out of and into each.
Next click on each of the boxes labeled " Plankton" , etc. to see the biomass
and flows of each group. Finally, click on "label" to see the name
of each compartment. In addition, each of the isotrophic classes
can be shown without ties between them (for clarity's sake) by turning
off the "in flux" and "out flux" boxes, and successively turning on each
of the isotrophic classes desired (check box). View compartment labels
by pointing to the desired node and left-clicking.
Network analysts:
-
Dan Baird, Department of
Zoology, University of Port Elizabeth, P.O. Box 1600, Port Elizabeth,
South Africa aladdb@zoo.upe.ac.za
-
Stephen Borgatti, Carroll School of Management, Boston College, Boston,
MA
-
Robert
Christian, Department of Biology, East Carolina University, Greenville,
NC 27858 christianr@mail.ecu.edu
-
Martin Everett, Department of Mathematics, University of Greenwich, London,
UK
-
Jeffrey C. Johnson Department
of Sociology and Institute for Coastal and Marine Resources, East Carolina
University, Greenville, NC 27858 johnsonje@mail.ecu.edu
-
Joseph J. Luczkovich,
Department of Biology and Institute for Coastal and Marine Resources, East
Carolina University, Greenville, NC 27858 luczkovichj@mail.ecu.edu
-
David Richardson, Department of Biochemistry, Duke University, Durham,
NC
-
Jane Richardson, Department of Biochemistry, Duke University, Durham, NC
Others who helped collect and analyze the data:
-
Hilary Neckles - US Geological Survey, Water Resources Office, Augusta,
ME hilary_neckles@usgs.gov
-
Bill Rizzo - US Geological Survey, Environmental and Contaminants Research
Center, 4200 New Haven Road, Columbia, MO 65201 bill_rizzo@usgs.gov
Participating Students, Department of Biology, East Carolina University:
-
Karen Halliday - (B.S Biology ECU '97)
-
Garcy Ward - (M.S. Biology, ECU '98) garcy_ward@h2o.enr.state.nc.us
-
David Gloeckner - (M.S. Biology Candidate), David.Gloeckner@noaa.gov
Some Related Internet Web Links:
Download the Network Analysis Data and Software
-
Raw data sets for our network analysis of this ecosystem are available
here
via anonymous FTP.
The data files below are for individual stations sampled in January
and February 1994 arranged in SCOR (Scientific Committee for Oceanographic
Research - a UN Committee) format saved as ASCII text files. SCOR format
is the standard input format for the Netwrk 4.2 software:
-
jans1.dat - Data from Site 1 (Live Oak Island Seine Beach) in January 1994
-
jans2.dat - Data from Site 2 (Wakulla Beach) in January 1994
-
jans3.dat - Data from Site 3 (Sprague Island) in January 1994
-
febs1.dat - Data from Site 1 (Live Oak Island Seine Beach) in February
1994
-
febs2.dat - Data from Site 2 (Wakulla Beach) in February 1994
-
febs4.dat - Data from Site 4 (St. Marks Lighthouse Point) in February 1994
Data sets below are averaged across compartments within months or over
both months:
-
jans12.dat - Data averaged from Sites 1 & 2 in January 1994
-
febs12.dat - Data averaged from Sites 1 & 2 in February 1994
-
janfeb.dat - Data averaged from sites 1&2 in January and February 1994
You may click on the highlighted link above, which will return an index
of filenames similar to those listed above. To transfer the files to your
PC via your web browser or FTP software, just click on the highlighted
file name, BUT HOLD DOWN THE SHIFT KEY (Netscape users) or RIGHT CLICK
THE MOUSE (Microsoft Internet Explorer users) or MAC people can click and
hold (then choose to download). Your web browser will ask you on what directory
or your hard drive you want to save it.
-
Download the Network Analysis Software
After you have downloaded a file, you may then input these files into
the Netwrk 4.2 software, which is available from:
-
Bob Ulanowicz's
Ecosystem Network Analysis page.
-
Our Download Network Software
home page. - I have Bob Ulanowicz' permission to distribute the
DOS software package NETWRK 4.2 for big networks (BIGNET.EXE for compartments
numbering up to 128). This will allow one to analyse our 51-compartment
models listed above. Other files on this index page are needed before
you can run NETWRK 4.2. You will need to download DOSXMSF.EXE to
run this software. Also on this page are the software packages for
calculating trophic impacts (IMPACT.EXE), a practice data set (CHESBAL.DAT)
and the documentation file for the trophic impact calculations (IMPACTS.DOC).
There are also software packages available for entering raw data on your
own network (BIGDAT.EXE) and balancing your model under steady-state assumptions
(DATBAL.EXE)., should you be so inclined. Please visit Bob Ulanowicz's
page above for more details.
The data files are in the following format (SCOR Format) that imports
directly into Netwrk 4.2:
| Line 1 |
A text title; the units of measurement |
| Line 2 |
The number of total compartments in the
model and the number of living compartments |
| Lines 3- 54 |
A list of the compartment labels in the
model |
| Lines 55 - 106 |
A list of compartment numbers and biomasses |
| Line 107 |
End of variable code |
| Lines 108 - 159 |
A list of compartment numbers and inputs
from outside the system; |
| Line 160 |
End of variable code |
| Lines 161 - 212 |
A list of material flows from compartments
in column 1 into compartments listed in column 2; flows are listed in column
3 |
For example:
Flow network of St Marks, JANUARY Site 1; mgC/m2/[day]. 6
51 48
phytoplankton
Halodule
micro-epiphytes
macro-epiphytes
benthic algae
bacterio-plankton
micro-protozoa
zooplankton
epiphyte graz amphipods
...etc.,
1 .8100000E+02
2 .4232000E+04
3 .2212000E+03
4 .6760000E+02
5 .1151000E+04
6 .8160000E+01
7 .4640000E+01
8 .5600000E+00
9 .1014800E+03
...etc.
