Warnke, D.A.

, J
.
Richter, and C. Oppenheimer. 1973. Charac-
teristics of the nearshore environment off the south coast of
Anvers Island, Antarctic Peninsula. Liinnology and
Oceanography, 18: 131-142.
-
-
Wood
Ecosystem studies and
mathematical model for Lake
Bonney, Taylor Valley, southern
Victoria Land
LEONARD W. HOWELL, J R. , RI CHARD G. KRUTCHKOFF,
and BRUCE C. PARKER
Virginia Polytechnic Institute and State Unwersity
Blacksburg, Virginia 24061
The fifth and final austral summer of our study of the
Lake Bonney ecosystempursuant to a mathematical model
was undertaken fromNovember 1976to January 1977.
Measurements throughout the season included dissolved 02,
pH, sulfates, NH,,-N, NO3 -N, NO2 -N, primary produc-
tivity, algal counts, temperature, alkalinity, and PO4 -P. In-
creasing emphasis was given to assessment of glacial
meltwater input to the lake, inorganic nitrogen and or-
thophosphorus composition of the meltwater and the ice-
cemented permafrost, and a newinvestigation of carbon-14
organic substrate uptake by the plankton microorganisms.
These data will be reported subsequently following detailed
analysis and interpretation.
In addition, we are investigating the growth of Lake Bon-
ney, the volume of which has more than doubled since its
discovery and description by Scott's field party in 1903 . The
lake's growth in recent years is emphasized by the necessity
to abandon the hut in January 1977when the water level
reached its foundation (figure 1). This hut was erected in
1962 and was several meters above Lake Bonney ice at that
time. The remainder of this report describes the
mathematical model, nownearly complete and to our
knowledge the first such model for an arheic, meromictic
lake.
Our modeling approach for the Lake Bonney ecosystem
has been to employ three categories of information to
develop a model as a tool for exploring interactions between
various lake organisms and their environments. These three
categories are (1) real data or measurements taken at Lake
Bonney (e.g., nutrient levels, productivity rates), (2 ) real
data or measurements fromother aquatic ecosystems (e.g.,
nutrient uptake rates, cell composition) and (3 ) intuitive
guesses (e.g., death rates, consumption rates). The Lake
Bonney model is deterministic and consists of 2 1
simultaneous differential equations, 2 1parametric equa-
tions, 4 2 submodels, and 180 coefficients. We have written a
FORTRAN programthat solves these equations numerically,
giving the solutions graphically and/or in tabular form.
Four programs permit a variety of input-output options.
Figure 1. Lake Bonney hut (built 1962) and analytical
chemistry lab (built 1971) in early January 1977, showing
lake water surrounding the hut and at a level nearly
touching the foundation.
Each programcalls each of the subprograms, and many of
the subprograms call each other as they run on the IBM
3 70/158 computer. Other important features of the model
are mass-balance, which checks the consistency of the equa-
tions, and a confidence interval submodel that allows for
confidence intervals in prediction.
Differential equations used in our model consist of
derivatives of biomass and nutrient (or other variable) con-
centrations with respect to time, as follows: (1) suspended
(plankton) algae, (2 ) suspended (plankton) bacteria, (3 )
suspended (plankton) yeasts, (4 ) water column phosphorus,
(5) water column inorganic nitrogen, (6) water column in-
organic carbon, (7) water column particulate organic mat-
ter (nonliving), (8) water column dissolved organic matter,
(9) water column dissolved oxygen, (10) mat algae (referring
to benthic attached felts), (11) mat yeasts, (12 ) mat bacteria,
(13 ) mat consumers (referring to ciliates, rotifers, tar-
digrades and nematodes in mats and which are not found in
the plankton), (14 ) mat inorganic nitrogen, (15) mat
phosphorus, (16) mat inorganic carbon, (17) mat dissolved
organic matter, (18) mat particulate organic matter (nonliv-
ing), (19) mat dissolved oxygen, (2 0) mat volume, and (2 1)
lake volume (living part only).
Parametric equations, which are those for environmental
variables influencing the biomasses and nutrients of interest,
are: (1) boron concentrations in the water column, (2 ) boron
concentration in the mat, (3 ) salt concentration in the water
column, (4 ) salt concentration in the mat, (5) glacier and
mountain blockage of sunlight, (6) daily/seasonal sunlight
intensity, (7) water column temperature, (8) mat tempera-
ture. (9) water column pH. (10) mat pH. (11) freezing/
melting of lake ice, (12 ) meltwater input of algae
(tychoplankton). (13 ) meltwater input of bacteria, (14 )
meltwater input of yeasts, (15) meltwater input (freshwater),
(16) meltwater input of inorganic carbon, (17) meltwater in-
put of dissolved organic matter, (18) meltwater input of par-
ticulate organic matter, (19) meltwater input of phos-
phorus, (2 0) meltwater input of inorganic nitrogen, and (2 1)
meltwater input of dissolved oxygen.
Sensitivity studies have been conducted to examine the
depth of various concepts in the submodels. Such submodels
October 1977
2 3
WATER ( . ULUMN ALGAE SCALE =3. 18E- 02
L6
15

**** *******A*******
14
13
1
11
10
9
B
7
6
5,
4,
3,
2
I I I GAYS I I I
0 40 80 120 160 200 240
SUNLI GHT I NTNS. L PER DY SCALE =0. 35E 02
16
15
14

* *
13
12
11
10
9,
8
7

* *
61
51

* *
41
31
2
11
I I DAYS I I
0

40 80 120 160 200 e40

WATER COLUMN ALGAE SCALE 0. 14E- 02

SUNLI GHT I NTNS. I PER DY SCALE =0. 35E 02
161 161 *
15 I ****
15 1 ** *
14 I ****t ***** 14 1 * *
13 1 v**s*.
13 1 *
121 12! *
11 1 11 I *
10 1 10 I *
91 91 *
81 81 *
7! 71
61 61 *
51 51 * *
41 41
31 , 31 * *
2 I 2 I ****
1 1
1 I *******a*****ss ** **** -
I I DAYS I I I I I I 1 DAYS I 1
240 280 320 360 400 440 480 240 280 323 360 400 440 480
Figure 2. Examples of Lake Bonney mathematical model computer output. 2A. 480 day prediction of phytoplankton algal
biomass for East Lake lobe (g/m 3
=
y axis value x 0.18x10 2 and
=
y axis value x 0.14x10 2 , respectively). 2B. Sunlight inten-
sity over Lake Bonney (langleys per day = y axis value x 0.35x102).
include (1) the effect of boron toxicity on algal growth, (2)
the sink effect of the mat in nutrient uptake from the water
column and (3 ) primary production as a function of light in-
tensity, nitrogen, phosphorus, inorganic carbon, dissolved
organic matter, boron concentration, and salinity.
Although our field studies at Lake Bonney have covered
only the austral summer, our model has been expanded to
include the winter and day-by-day predictions for a period
of more than 2 years. Other programs cover periods of 120
and 480 days. An example of the computer output is il-
lustrated in figure 2A, which shows phytoplankton biomass.
Figure 2B illustrates graphically the parametric equation for
sunlight.
Having a total lake ecosystem model allows one to order
biological concepts in terms of their individual contribution
to the lake as a whole. It is also a useful tool in exploring the
environmental impact of perturbations on the lake. While
this model would require major alteration for typical dimic-
tic lakes or impoundments in exorheic regions, such as most
of North America, South America, and New Zealand, the
model could be adapted to other similar meromictic lakes
with only minor changes in coefficients and parametric
equations.
We gratefully acknowledge support of National Science
Foundation grant GV-3 51 71 and members of the 1976-1977
Lake Bonney field team: T. Allnutt, B. Bishop, D. Brown,
J.
Crate, L. Heiskell, Jr., W. Thompson, D. Wendt, and S.
Wendt.
Baseline microbiological data for
soils of the Dufek Massif
B.C. PARKER, A.B.
FORD,*
T. ALLNUTT, B. BISHOP,
and S. WENDT
Virginia Polytechnic Institute & State University
Blacksburg, Virginia 24061
Cameron and Ford (1974) report the first baseline
microbiological data for soils of the Pensacola Mountains.
Especially noteworthy are their findings that few organisms
and low species diversities are associated with these relatively
remote and pristine residual soils, thereby resembling the
earlier results from undisturbed sites in the dry valleys of
Victoria Land. We report here a preliminary assessment of
this earlier study using 23 aseptically collected soil samples
from the Dufek Massif, primarily between 82139'S.
5340'W. and 8230'S. 5148'W.
Collections were made 1-20 December 1976 and brought
to McMurdo Station within 2 weeks for immediate pro-
8 Alaskan Geology Branch, U.S. Geological Survey, Menlo Park,
California 94 02 5.
24
ANTARCTIC JOURNAL