-------
TABLE 52. AGRICULTURAL SOIL NUTRIENT CHANGES APRIL - JULY 1975.
Crop
Fertilizer,
Sample Crop grown Planted Added-5/75 Winter - Feb 1975
Site // Summer '74
5/75
N
N
c c
Summer - July 1975 Incremental Change
p. p
i o
N
P. P
± o
N
38 Corn Alfalfa - - 210 350 18.1 1255 54 554 18.3 1075 -156 +204 +0.2 -179
39 Alfalfa Corn 16.3 19.9 226 294 23.5 1506 78 474 32.8 1213 -148 +180 +9.3 -293
41 Hay - 186 354 27.7 1483 140 308 18.3 1253 -46 -46 -9.4 -229
u>
Sample Sites 38, 39 plowed under 5/75, fertilized and planted; Site 41 not replanted or plowed
under (uncut hay).
Amounts in kg/ha. Sample Sites 38, 39 received approximately 610 kg/ha cow manure (43.6 kg/ha/week
for 14 weeks in between soil samples); Sample Site 41 not spread with manure.
Amounts in ppm.
-------
TABLE 53. NUTRIENT OUTPUT IN MILL CRF.EK FATERSHED
Item
1. Farm Crops Sold
a. Corn PQ
No
b. Hay P0
No
c. Alfalfa PO
No
d. Vegetable PO
No
2. Livestock Products
Cows Milk PQ
3. Forest Lumber Removal
White-Red Pine PQ
Spruce-Fir T
No
Oak-Pine PQ
Oak Hickory PQ
Unit Multipl ier
Loading (7 reroved
(kO from watersbe^l)
P92a* 32.6
590°
1072 49.4
8473
1763 10.0
17571
42° 10.3
750
851250b' O.«oc*
85125^ 1.2d*
1003f* n.76p<
9091 0.76
308 0.°7
25F.2 O.Q7
162 ".56
1587 0.56
332 0.23
Item
Output
O'.g/yr) Ref.
323 Table
1026
530 "
4186 "
335 "
3338 "
44
78
792 Table
10215 Table
7.62
6.Q1
3.0
25.0
0.91
8.80
0.76
47
31
31
(continued)
1-134
-------
TABLE 53 (continued)
Item
3. Forest Lumber
Elm-Ash Maple
Maple-Beech
Birch
Aspen-Birch
TOTAL
Removal
po
No
po
Ko
po
Ko
Unit
Loading
(Vp)
(cont'd.)
85°
7062
1291
12727
126
1174
Multiplier
(7 removed
from watershed
n.37p<
0.37P'
0.76
0.14
0.14
Item
Output
(kp/yr) Ref.
3.18
29.46
9.81
96.72
0.18
1.64
25.46 PQ
178.8 NQ
c.
d.
e.
f.
* Total watershed crop nutrients (Tahle 47)
'" Personal communication with watershed dairy fanner; milk/cow/yr
6810 kg with 125 producinp cows in watershed
Reference 59
Reference 59 gives 3.6% K-compounds, 7N estimated as 1.27
By forest type, maximum value
Tahle 46, for total watershed
Tahle 23
1-135
-------
REFERENCES
1. Pear, F. F., ed. Chemistry of the Soil. Reinhold Publishing Corp.,
New York. New York, 196*.
2. Duvigneaud, P. and P. Denaeyer-deSmet. Biological Cycling of Minerals
in Temperate Deciduous Forests. In: ^coloeical Studies, Vol. 1,
Analysis of Temperate Forest Fcosysteirs. D. E. Keichle, ed. Springer-
Verlag, Berlin, 1°70.
3. Ovington, J. D. Ouantitative Ecology and the Woodland Ecosysteir
Concept. In: Advances in Ecological Research. J. B. Cragg, ed.
Vol. 1, Academic Press, Inc., New York, New vorv, 1062.
4. Newbould, P. J. Methods for Estimating the Primarv Production-Forests.
IBP Handbook No. 2. Blackwell Scientific Publications, Oxford, 1Q67.
5. Spurr, S. P. Forest Ecology. The Ponald Press Co., New vorV, New YorV,
1964.
6. Cole, P. V., S. P. Gessel, and S. F. Dice. Distribution and Cycling of
Nitrogen, Phosphorus, Potassium, and Calcium in a Second-Growth Douglas-
Fir Ecosystem. In: Symposium on Primary Productivity and Mineral
Cycling in Natural Ecosystems. P. E. Young, ed. Univ. of Maine Press,
Orono, Maine, 1967.
7. Bakuzis, E. V. Forestry Viewed in an Ecosystem Perspective. In: The
Ecosystem Concept in Natural Resource Management. G. **. van T>yne, ed.
Academic Press, New York, New YorV, 1969.
8. Bormann, F. P. and G. F. Likens. The Vatershed Ecosystem Concept and
Studies of Nutrient Cycles. In: The Ecosystem Concept in Natural
Resource Management. G. M. Van Dyne, ed. Academic Press, New York,
New York, 196°.
9. Johnson, F. L. and P. G. Risser. Eionass, Annual Net Primarv
Production, and Dynamics of Six Mineral Elements in a Post Oak-Blackjack
Oak Forest. Ecology. 55:1246-125?, 1974.
10. Geyger, E. Green Area Indices of Grassland Communities and Agricultural
Crops Under Different Fertilizing Conditions. In: Ecological Studies,
Vol, 2, Integrated Experimental Ecology. P. Ellenherg, ed. Springer-
Verlap, Berlin, 1Q71.
1-136
-------
11. Gosz, J. R., G. F. Likens, and F.. F. Bormann. Nutrient Release from
Decomposing Leaf and Branch Litter in the Pubhard Brook Forest, New
Hampshire. Ecological Monographs. Vol. 43, No. 2, pp. 173-191, 1°73.
12. Edwards, C. A., P. F. Reichle, and T>. A. Crosslev, Jr. The Role of Soil
Invertebrates in Turnover of Organic vatter and Nutrients. In: Eco-
logical Studies. Vol. 1, Analysis of Temperate Forest Ecosystems, P. F.
Reichle, ed. Springer-Verlag, Berlin, 1970.
13. Satoo, T. A Svnthesis of Studies by the Farvest Method: Primary
Production Relations in the Temperate Deciduous Forests of Japan.
In: Ecological Studies, Vol. 1, Analysis of Temperate Forest Fcosys-
tems. P. F. Reichle, ed. Springer-Verlap, Berlin, 1970.
14. Likens, G. F. and F. H. Bormann. Chemical Analvsis of Plant Tissues
from the Puhhard *rook Ecosystem In New Hampshire. Bull. No. 79. Yale
University, School of Forestry, New Paver, Connecticut, 1970.
15. Siccama, T. G., F. H. Bormann, and G. E. Likens. The Pubbard Brook
Ecosystem Study: Productivity, Nutrients, and Phytosociologv of the
Ferbaceous Layer. Ecological Monographs, 40(4):38°-402, 1970.
16. Eber, W. The Primary Production of the Ground Vegetation of the Luzulo-
Fagetum. In: Ecological Studies, Vol. 2, Integrated Experimental
Fcology. F. Ellenburg, ed. Springer-Verlag, Berlin, 1971.
17. Likens, G. F. and F. F. Bormann. Nutrient Cvcling in Ecosystems. In:
Ecosystem Structure and Function. J. A. Wiens, ed. The Orepor State
University Press, U.S.A., 1972.
18. Fortescue, J. A. C. and G. G. Marten. Micronutrients: Forest Fcolopy
and Systems Analysis. In: Fcological Studies, Vol. 1, Analysis of
Temperate Forest Ecosystems. P. F. Reichle, ed. Springer-Verlag,
Berlin, 1°70.
19. Loehr, R. C. Characteristics and Comparative Magnitude of Non-Point
Sources. J. WPCP, 46(8):1840-1871, 1974.
20. Heller, H. Estimation of Btomass of Forests. In: Ecological Studies,
Vol. 2, Integrated Experimental Ecology. F. Fllenberp, ed. Springer-
Verlag,, Berlin, 1971.
21. Woodwell, G. M. and D. B. Blotkin
by Gas Exchange Techniques:
Studies, Vol. 1, Analysis of
Reichle, ed. Springer-Verlpg
22. Runge, M. Determination of
Vol. 2, Integrated
Verlag, Berlin, 1971.
Experimental
Metabolism of Terrestrial Ecosystems
The Brookhaven Approach. In: Ecological
Temperate Forest Fcosystenrs. D. F.
Berlin, 1970.
Energy Values. In: Ecological Studies,
Fcology. P. F.llenberg, ed. Springer-
1-137
-------
23. Plant Tissues and Organs: Mineral Composition, Sec. 53. In: Biolopy
Data Book, Vol. 1, 2nd ed. P. L. Altman and n. S. Mttmer, eds.
Federation of American Societies for Experimental Biology, Bethesda,
Maryland, 1972.
24. Odum, F. P. Fundamentals of Fcology. W. B. Saunders Co., Philadelphia,
Pennsylvania, 1971.
25. Ulrich, G., E. Ahrens, and M. TTlrich. Soil Chenical Differences Between
Beech and Spruce Sites - An Example of the Methods Used. In: Ecolop-
ical Studies, Vol. 2, Integrated Experimental Ecology. P. F.llenherp,
ed. Springer-Verlag, Berlin, 1971.
26. Likens, G. F. and F. H. Bormann. Linkages Between Terrestrial and
Aquatic Ecosystems. Bioscience, 24(8):447-456, 1°74.
27. Gihble, E. B. Phosphorus and Nitrogen Loading and Nutrient Budget on
Lake George, New York. Rensselaer Polytechnic Institute, Troy, New
York, 1974.
28. Spedding, C. R. W. Grassland Ecology. Oxford Hniversitv Press,
London, 1971.
29. Coupland, R. T., R. Y. Zacharuk, and E. A. Paul. Procedures for Study
of Grassland Ecosystems. In: The Fcosystem Concept in Natural Resource
Management. G. M. Van Dyre, ed. Academic Press, New York, New York,
1969.
30. Speidel, B. and A. Weiss. Primary Production of a Meadow (Trisetetum
flavescentis hercynlcum) with Different Fertilizer Treatments - Prelim-
inary Report. In: Fcological Studies, Vol. 2, Integrated Experimental
Ecology. H. Ellenberg, ed. Fpringer-Verlae, Berlin, 1971.
31. Lewis, J. K. Range Management Viewed In the Fcosystem Framework. In:
The Fcosystem Concept in Natural Resource Management. G. M. Van Dyke,
ed. Academic Press, New York, New York, 196°.
32. Stevenson, F. J. Origin and Distribution of Nitrogen in the Soil. In:
Soil Nitrogen. W. V. Bartholonew and F. F. Clark, ed. American Assoc.
of Agronomy, Inc., Madison, Wisconsin, 19f>5.
33. Davis, G. K. Mineral Problems in Forage Utilization. In: Grasslands.
H. B. Sprague, ed. American Assoc. for the Advancement of Science,
Washington, D.C. 1959.
34. Lunt, 0. R. Problems in Nutrient Availability and Toxicity. In: The
Biology and Utilization of Grasses. V. B. Younger and C. M.. McKall,
eds. Academic Press, New York, New York, 1972.
35. Billings, W. D. Plants and the Ecosystem. Wadsworth Publishing Co.,
Belmont, California, 1966.
1-138
-------
36. Humphrey, R. R. Range Ecology. The Ronald Press Co., New York, New
York, 1962.
37. Milner, C. and R. E. Hughes. Methods for the Measurement of the Primary
Production of Grassland. IBP Handbook No. 6. Blackwell Scientific
Publications, Oxford, 1967.
38. Washko, J. B. The Place and Contribution of Grasslands to the Agricul-
ture in the Northeast. In: Grasslands. P. B. Sprag,ue, ed. American
Assoc. for the Advancement of Science, Washington, D.O., 1^5°.
39. Weaver, J. Prairie Plants and Their environment. University of
Nebraska Press s 1968.
40. Tesar, M. B. The Place and Contribution of Grasslands to the Agricul-
ture of the Cornbelt. In: Grasslands. P. B. Sprague, ed. American
Assoc. for the Advancement of Science, Washington, D.C., 195°.
41. Properties of and Excretion Products in Urine: Mammals Other than
Sec. 186. In: Biology Data Book, Vol. 3, ?nd ed. P. L. Altman and
D. S. Dittmer, eds. Federation of American Societies for Experimental
Biology, Bethesda, Maryland, 1972.
42. Composition of Cereal Grains and Forage. Committee on Feed Composition
of the Agricultural Board, National Academy of Sciences - National
Research Council, Publication 585, Washington, D.C., 1958.
43. Gotaas, H. B. Compostinp, Sanitary Disposal, and Reclamation of Organic
Wastes. World Health Organization, Geneva, 1°56.
44. Heinke, G. W. and J. D. Norman. Condensed Phosphates in Lake Water and
Waste Water. Paper presented at the 5th International Water Pollution
Research Conference, San Francisco, California, July-August 1970.
45. Pointer, H. A. and Bywaters, A. Composition of Sewage and Sewage
Effluents. Paper presented at the Institute of Sewage Purification,
London, December 1960.
46. Friedlan, A., T. Shea, and P. Ludwip. Ouantitv srd Ouality Relation-
ships for Combined Sewer Overflows. Conference Preprint - Paper
presentee" at the 5th International Water Pollution Research Conference,
f.r.n Francisco, California, July-August 1970.
47. Popel, F. Sludge Digestion and Disposal, Third revised edition.
Techr.ische Hochschule, Stuttgart, Germany, 1967.
48. Barshied, R. and H. M. El-Paroudi. Physical Chemical Treatment of
Septic Tank Effluent. J. WPCF, 46(11), 1<>74.
49. Capital District Transportation and Regional Planning Study. Area Map
RC No. 6952, Sheet J-10, Troy South, Fast Greenbush ouad, 1971.
1-139
-------
50. Natural and Cultural Features Survey, 1968. Fensselaer County Planning
Board, Troy, New York, October 1969.
51. Scavia, D. A Study of Lakes in Rensselaer County with Proposals for
Environmental Management. Fresh Water Institute, Rensselaer Polytechnic
Institute, Troy, New York, December 1972.
52. Soil Conservation Service. Unpublished soil data from aerial photo-
graphs of Rensselaer Countv. V.S.D.A. Soil Conservation Service,
Fyantskill, New York, 1974.
53. Soil Survey Interpretations of Soils in New vorv State. Prepared by the
Dept. of Agronomy, Cornell Univ., Ithaca, New York, and U.S.D.A. Soil
Conservation Service, Syracuse, New York, 1972.
54. Climatological Data - New York Annual Summary 1972. U.S. Dept. of
Commerce, National Oceanic and Atmospheric Administration, 83(13).
55. Ferguson, R. H. and C. E. Mayer. The Timber Resources of New York
Ftate. Bull. NF.-20, U.S.D.A. Forest Service Pesource, 1970.
56. Fetling, L. J. and I. G. Carcich. Phosphorus in Waste Water. Vater
and Sewage Works, 12" (2): 59, 1973.
57. Nev York State Environmental Conservation Law, Section 35-0105.
58. CRC Handbook of Chemistry and Physics, 52nd ed. Chemical Rubber Co.,
Cleveland, Ohio, 1971.
59. Lange, N. A. Handbook of Chemistry, 7th ed. Handbook Publishers,
Inc., Sandusky, Ohio,
1-140
-------
APPENDICES
1-141
-------
APPENDIX A
ANALYTICAL PROCEDURES
GENERAL
The principle sources of error in soil sampling are: (1) variability
among different samples drawn from same volume; (2) variability introduced
among subsamples of same soil sample; (3) variability from one chemical
determination to another.
The main source of error is (1). Therefore this variability must be
delineated or reduced. This introduces the concept of composite soil
sampling. A composite soil sample gives a mean analytical value represen-
tative of the soil sampling volume from which the composite sample was
drawn. (Any analytical value is a mean for the individual soil particles,
hence the futility of an argument against obtaining mean-value analyses.)
Analyses for C, N, P and pH made on composite samples have been found to be
equivalent to the mean of analyses of individual cores. Individual cores
tend to be subject to large variations not significant to plant growth
individually (and likewise nutrient distribution).
FIELD SAMPLING
Soil samples were obtained using a 3.175 cm diameter auger. A Number
10 tin can with a 3,81 cm hole in the bottom was used as a dirt catcher.
To obtain a soil core, the can was placed on the ground surface, bottom
down, and the auger inserted through the hole. Any dirt clumps forced out
of the hole during the drilling process were retained inside the can and
composited with the core sample. Since the screw end of the .auger is only
20.3 cm long, 30.5 cm cores were taken in two increments. The major part
of the soil sample was retained in the screws of the auger and was scraped
off by hand and deposited into the composite sample bag.
Ten soil cores were obtained from within a 6.1 meter square staked off
area at each sample site. A zig-zag pattern was employed in locating cores
in order to obtain a random, representative sample, after the sampling pro-
cedures in A.O.A.C. Methods of Analysis, (llth ed.).^ soil cores taken
from agricultural fields were spread over the majority of the field and
taken in the same zig-zag pattern from both crop rows and furrows. All
soil cores from each sample site were composited and transported to the
laboratory in labelled plastic bags.
1-142
-------
Plant detritus samples from the interface layer were obtained from
within a folding wooden 0.9114 meter square frame, comprising 0.836 nr of
ground surface. Live grasses, plants and any snow cover were removed by
hand from within the square and discarded. All dead grasses and decomposed
materials, including twigs and sticks, were collected by hand and with the
aid of pruning shears and a garden hoe. Samples were transported to the
laboratory in labelled plastic bags.
Upon return to the laboratory at the end of a sampling day, soil and
plant detritus interface samples were immediately weighed to an accuracy of
jf 0.1 gm (field weight) and stored at 1.7°C prior to drying and grinding,
which was generally accomplished within a few days although some samples
remained in cold storage (4°C) up to 2 weeks.
SAMPLE PREPARATION
Moisture determination and sample drying were accomplished in the same
operation for soil and plant detritus Interface samples. The entire sample
was dried on tin foil covered trays at 75°C for 48 hours in a forced air
oven. After cooling, sample dry weight was recorded. Moisture was
calculated using field weight and dry weight figures.
Dried soil samples, including rocks, were ground in a Wiley Mill to
pass through a 1.0 m screen within the grinder. When grinding soil
samples, the interior cutting blades of the Wiley Mill were removed so that
clearance -between the revolving blade head and cutting chamber wall was
approximately 2.5 cm. Grinding in this manner reduced dry soil clumps to
0.5 mm size and retained rocks and gravel on the grinder screen. All rocks
and gravel were removed from the grinder after grinding a sample and then
weight was recorded. The 0.5 mm soil sample was then thoroughly parti-
tioned using a riffle and stored in 0.94 liter plastic containers with a
tightly sealed top. About 200 g of 0.5 mm soil was passed through a 0.15
mm screen and stored in 250 ml screw top glass jars prior to analysis. All
samples, once dried and in sealed containers, were stored at room tempera-
ture. Immediately prior to weighing for analysis, samples were re-dried at
75°C for 2 hours and cooled under desslcation.
Dried plant detritus interface samples, including twigs and small
branches, were ground in the Wiley Mill, with cutting blades installed, to
pass through a 1 mm screen. Ground dry samples were thoroughly partitioned
using a riffle and then stored.
Prior to analysis, the sample was re-dried in the same manner as the
soil samples. Plant samples were handled carefully prior to grinding.
PHOSPHORUS EXTRACTION
Phosphorus extraction for soil and detritus interface samples was
accomplished after methods in Jackson (pp. 171-174).!
1-143
-------
Apparatus: centrifuge, 60 ml centrifuge tubes, 250 ml volumetric flasks,
50 ml beakers, drying oven, fume hood, 100 ml volumetric
flasks, pipets
Reagents: 0.5N NaOH, cone, hydrochloric acid (HC1)
Soil Extraction Procedure: 0.5 gm of re-dried soil (passed through 1 mm
screen) is added to a 60 ml centrifuge tube. Add 10 ml cone.
HC1 and heat in a water bath at 70°C for 10 minutes under a
fume hood. Remove from water bath, add another 10 ml HC1 and
stand at room temperature 1 hour. Centrifuge and pour super-
natant liquid into a 250 ml volumetric flask containing 50 ml
distilled H20 (this is acid extract). Add 40 ml 0.5N NaOH to
soil sample in centrifuge tube. Mix and stand 1 hour at room
temperature. Centrifuge and pour supernatant into volumetric
flask containing acid extract. Add 50 ml 0.5N NaOH to centri-
fuge tube cover with inverted 50 ml beaker and heat in 90°C
oven for 8 hours. Centrifuge and add the supernatant liquid to
the 250 ml volumetric flask containing the acid extract. This
solution, now the mixed extract, is then diluted to 250 ml with
distilled water and thoroughly mixed.
Plant Tissue Extraction Procedure: 0.5 g of re-dried interface material
(ground to pass 1 mm screen) is extracted in the same manner as
soil samples. This extract is used for P£ analysis only.
PT analysis is accomplished by dry ashing procedure.
PHOSPHORUS DETERMINATION IN EXTRACTS
Analytical determinations of phosphorus for soil and detritus extracts
were accomplished by procedures detailed in Standard Methods.^
1. Inorganic P Soil Extract: After mixing, the suspended sediment in the
250 ml mixed extract is allowed to settle and 25 ml is care-
fully pipetted into a 100 ml volumetric flask. The sample is
diluted to 100 ml and the remainder of the determination is by.
the stannous chloride method as contained in Standard Methods.
Sample absorbance is read on a Bausch and Lomb Spectronic 20
spectrophotometer at 690 urn after 10 minutes at 24°C using a
path length of 1.0 cm. A standard curve for phosphorus is con-
tained in Figure A-l.
Sample Inorganic Phosphorus is calculated as follows:
Measured cone, (ppm) x (dilution vol (l)/sample vol (1)) x
(total vol (1) of extraction solution/wt soil in kg) - ppm P in
dry soil
Total mg P in sample « ppm x wt sample (kg)
Example: measured ppm x (.100/.025) x (250/.0005 kg) - ppm P
2. Inorganic Phosphorus: detritus interface extract is determined in the
same manner as soil extract inorganic phosphorus.
I-J44
-------
I
I—t
Ui
0.8
0.6
o
JO
JS 0.4
0.2
I I I I I I I I I I I I I I I T
I I I I I I I I I I 1 I I I I I
0 O.I
0.2 0.3 0.4 0.5
Concentration P(ppm)
0.6 0.7 0.8
FIGURE A-l. Phosphorus standard curve.
-------
3. Total Phosphorus Soil Extract: the flask containing mixed extracts is
shaken thoroughly to suspend all sediments and 25 ml is immedi-
ately pipetted into a 75 ml distillation flask. 10 ml cone.
nitric acid (IWH^) and 2 ml 72 percent perchloric acid
(HCIO^ are added. A microkjeldahl distillation apparatus is
used to heat samples to just boiling for 1 hour. Fumes col-
lected by this distillation apparatus are collected under glass
and drawn through a bubbler of ION NaOH and finally disposed of
in water suction. The entire distillation apparatus is placed
under a hood for safety, to prevent escape of volatile and
dangerous perchloric acid fumes. When cooled, the digested
colorless liquid is transferred to a 100 ml volumetric flask
and 15 ml of 6N NaOH is added to neutralize excess acidity.
The sample if then diluted to 100 ml and analyzed by the
stannous chloride method^ as described previously.
Sample Total Phosphorus is calculated as follows:
(measured cone, (ppm) x 0.100 1 x .250 I/.025 l)/wt sample (kg)
= ppm P in dry soil
Sample Organic Phosphorus is calculated by difference
mg PT - mg Pf - mg PQ
PHOSPHORUS DETERMINATION IN PLANT TISSUE
Total Phosphorus in plant detritus interface is accomplished after ,
methods for total phosphorus determinations in AOAC Methods of Analysis.''
Apparatus: Gooch crucible, pipets (5, 10 ml), muffle furnace, filtering
flask and vacuum, Whatman #42 filter paper, rubber spatula,
100 ml volumetric flasks.
Reagents:
Procedure:
0.5N Mg (OAc)
one liter.
liter.
) made by diluting 53.6 g Mg (C H_0_)_
2N H2S04, made by diluting 56 ml^conC.
4H 0 to
to one
0.5 g of interface sample (ground to pass 1 mm screen) is
placed in a gooch crucible, add 5 ml of Mg(OAc)2 and 10 ml
^0. Bring to boil on a hot plate and allow crucible to dry.
Place crucible in a 600°C muffle furnace for 30 minutes until
ash is uniformly gray in color. After cooling add 10 ml 2N
^30^ to crucible and rotate to bring acid in contact with
entire ash. Add 15 ml t^O and evaporate to 5 ml on a hot
plate. Remove, add 20 ml ^0 and cool. The contents are
then transferred to a 100 ml volumetric flask, rubbing down the
crucible with a rubber spatula and rinsing with distilled HO.
The 100 ml volumetric flask is diluted to the mark with dis-
tilled H20. 5 ml of this volume is then taken and diluted
to 100 ml for analysis by the stannous chloride method in
Standard Methods.^
1-146
-------
Calculations:
(measured ppm x 0.1 1 (dilution volume)) /sample wt mg *
Vg P/g dry soil
TOTAL KJEHLDAHL NITROGEN (TKN) DETERMINATION
Digestion of soil and interface samples follow the procedure' outlined
in Jackson.1 Distillation and tltration follows Standard Methods.-*
Apparatus: All equipment listed in Standard Methods^ for TKN determina-
tion. A 0.15 mm screen (100 mesh), analytical balance, and
11 cm filter paper are also required.
Reagents: Standard 0.02N t^SO^, cone. H-jSO^, granulated zinc.
Indicating boric acid solution is as described in Standard
Methods-* for TKN determination. Sodium hydroxide-sodium
thiosulfate reagent is made by dissolving 72 g Na2S.O~«5H 0
in 300 ml distilled H2° and adding this solution to 600 ml of
50 percent w/w NaOH. Digestion reagent is made in two parts;
solution and powder. Digestion solution: 80 g CuSO, «5H 0
diluted to 500 ml. Mix the two solutions. Powder is maae by
mixing 6 g HgO with 2 g Se powder.
Preparation: 4.5 g soil sample (0.5 plant tissue) is ground to pass a
0.1 mm screen and weighed into an 11 cm filter paper. 0.16 g
digestion powder is added. The sample-powder mixture is wrap-
ped and dropped as a packet into an 800 ml Kjeldahl flask.
40 ml distilled 1^0 and 10 ml digestion solution are added.
(Soil samples only, soak 30 minutes.) Add 35 ml cone. H SO,
down side of flask, avoiding packet. Add boiling chips.
Digestion: .Digest over very low heat 30 min, until frothing stops, rotat-
ing flask at intervals. Digest for 1 hour + 0.25 hour after
solution has cleared (light yellow-green or gray color). At
end digestion, heating is stopped, fume exhaustion continues
until fuming stops. Cool just until crystals start to form
then add 300 ml distilled H20.
Distillation: Use 25 ml indicating Boric acid in 500 ml Erlenmeyer. Start
cooling I^O, and have tube end below surface. Mix flask, add
several pieces zinc and boiling chips, then add 125 ml NaOH-
Na-S-O, solution down side of flask without mixing. Attach to
still, mix thoroughly, heat to boiling. Distill 150 ml. If
bumping is a problem transfer contents to another flask, leav-
ing residue, and continue with procedure. 100 ml NaOH-Na?S_0
solution may be used.
Determination: Titrate with 0.02N H2SO^. At end point, green color
disappears. One drop excess turns solution purple.
Calculations: % N in dry soil - (ml sample titration - ml blank titration)
x (normality of acid, this case - 0.02N) x (1.4 /sample wt g)
I-K7
-------
AMMONIUM DETERMINATION
The extraction procedures for soil and plant tissue samples are after
Jackson. 1 Determination of ammonia by distillation follows the procedure
for TKN in Standard Methods.3
Apparatus: Balance and 500 ml Erlenmeyer are needed for sample prepara-
tion. An 11 cm Buchner funnel, filter flask, Whatman #42
filter paper and vacuum pump are needed to separate the leach-
ate from the sample. 800 ml Kjeldahl digestion flasks and
distillation apparatus are required.
Reagents: Extraction solution is a 10 percent NaCl solution acidified to
pH 2.5 with cone. HC1. Ammonia is distilled into indicating
boric acid as described in Standard Methods3 for TKN determi-
nations. Distillate is titrated with standard 0.02N H SO^.
Extraction of Ammonium: Weigh out 100 g fresh soil sample into 500 ml
Erlenmeyer (5 g plant tissue). Make separate moisture determi-
nation (1 mm fraction). Add 200 ml acidified NaCl solution and
shake thoroughly at first, then intermittently for 0.5 hours.
Moisten filter paper and seat by suction on filter, add NaCl
sol. through filter. Add 250 ml more NaCl in increments, first
increment to rinse out flask.
Determination: Transfer NaCl leachate to 800 ml digestion flask. Add
80 ml 50 percent w/w NaOH down side of flask and distill about
125 ml into 25 ml boric acid. Titrate with 0.02N H2S04 until
green color turns purple.
Calculation: % N determined as in TKN determination. Organic N - TKN
minus
NITRATE AND NITRITE DETERMINATION
The extraction of nitrate and nitrite nitrogen is after Jackson.1
Determinations follow Standard Methods.3
Extraction Apparatus: 500 ml wide-mouth Erlenmeyers with stoppers, filter-
ing flask, Whatman #42 filter paper, Buchner funnel, and a
balance are required.
Extraction Reagents: Ca(OH)2» MgC03, activated charcoal, IN CuSO^,
0.6 percent Ag2S04 solution (6g/l). Extraction solution is
made by adding 200 ml CuS04 solution to 1 1 Ag2S04 solu-
tion and diluting the result to 10 liters.
Preparation: 50 g soil sample is ground to 0.5 mm. 10 g plant tissue
sample is ground to 1 mm.
1-148
-------
Extraction: 50 g soil (10 g plant tissue) is weighed and placed in 500 ml
bottle. 250 ml extraction solution is added. Shake 10 minutes
and add 0.4 g Ca(OH)2» shake 5 minutes more. Add 1 g
MgC03> mix thoroughly and filter on dry filter paper, discard
first 20 ml. Color development follows.
Colored Extracts: Add 1 g activated charcoal to filtered extract. Shake
15 to 30 minutes and filter again.
Nitrate Determination: Nitrate was determined using the brucine-sulfanilic
acid method as described in Standard Methods^ and compared to
the standard curve shown in Figure A-2.
Nitrite Determination: Nitrite analysis followed Standard Methods^ but
nitrite was rarely found in measurable concentrations.
Soil and Leachate Analyses
1. Total Carbon, Soil: was measured by ignition in a Leco carbon analyzer
following manufacturer's instructions. 25 mg of soil was added
to a ceramic crucible. One scoop each of iron accelerator and
tin copper catalyst was added. The crucible was ignited in a
stream of oxygen (1.5 liters/min) for 6-10 minutes. C02
produced was absorbed in an ascarite bulb and determined
gravimetrically by the following formula:
Weight absorption Bulb (mg) 07 ,Q .
w,_ . . * / \ X t.l»t.y = A W>
Weight sample (mg)
The error of 6 glucose and 6 steel standards analyzed ranged
from 1.7 to 4.9 percent.
2. Total Dissolved Carbon: measured on a Beckman Carbonaceous Analyzer
Model #137879 using 20 ul liquid sample. A calibration curve
is shown in Figure A-3.
3. Dissolved Organic Carbon: measured on a Beckman Carbonaceous Analyzer,
following manufacturer's instructions. 0.2 ml of 2N NCI was
added to 20 ml sample, and bubbled with N2 gas for 10 minutes
prior to analysis.
4. Chlorides: measured by silver nitrate metnod as described in Standard
Methods. 50 ml of sample were used.
5. pH: measured on a Fisher Accumet pH meter, Model 320.
6. Dissolved Oxygen: D.O. was measured with a Delta Scientific Oxygen
Meter and probe, Model 75.
1-149
-------
Ul
o
10
20 30 40 50 60 70 80
Concentration Total Dissolved C (ppm)
90 100
FIGURE A-2. Total dissolved carbon standard curve.
-------
0.8
0.6
o
M
<0.4
0.2
Concentration NO^-N
0.6 07
— (ppm)
FIGURE A-3. Nitrate nitrogen standard curve.
-------
7. Inorganic Phosphorus: 100 ml leachate sample was used without dilution
in the procedures previously described for P^ determination
on phosphorus extract.
8. Total Phosphorus: 25 ml sample was used without dilution in the pro-
cedure previously described for P^. determination on digested
phosphorus extract.
9. Nitrate Nitrogen: 10 ml sample was used without dilution in the pro-
cedure previously described for nitrate determination on
nitrate extract.
10. Ammonia Nitrogen: 350 ml sample or total volume remaining after other
analyses was used without dilution in the procedure previously
described for ammonia determination of ammonia extract.
1-152
-------
REFERENCES
1. Jackson, Marion L. Soil Chemical Analysis. Prentice Hall, Inc.,
Englewood Cliffs, New Jersey, 1958.
2. Assoc. of Official Agricultural Chemists. Official Methods of
Analysis, llth ed. Assoc. of Official Analytical Chemists, Washington,
D.C., 1970.
3. APHA, AWWA, WPCF. Standard Methods for the Examination of Water and
Waste-water, 13th ed. American Public Health Association, Washington,
D.C., 1971.
1-153
-------
APPENDIX B
1. Sample Resident's Survey Form, with cover letter and two completed
forms.
2. Layman's Summary of Project with Area Map and cover letter.
1-154
-------
RENSSELAER POLYTECHNIC INSTITUTE
TROY, NY 12181
March 25, 1975
STUDY OF MILL CREEK WATERSHED LAND NUTRIENTS
conducted by the Department of Chemical and Environmental Engineering for
the New York State Department of Environmental Conservation, United States
Environmental Protection Agency, and the International Joint Commission
Project Advisor: Dr. Hassan El-Baroudi, Associate Professor of
Environmental Engineering, Rensselaer Polytechnic
Institute
Dear Resident:
You may remember the summary of the above project which we delivered
to you a few weeks ago. Since that time we have been working hard on
sampling and analyzing soils and calculating the amount of nutrients in
forested and other land. Good progress has been made, but in order to cover
the residential and farmland areas of the watershed WE NEED YOUR HELP in
obtaining the necessary information.
Enclosed is a questionnaire which we think is suitable for our study.
We hope that you will find it reasonable to fill out and mail it in the
enclosed envelope. If you have any suggestions or questions, do not
hesitate to call or write.
Best regards and gratitude.
Yours very truly,
Kevin Walter
Graduate Student (270-6367)
Deborah James
Graduate Student
1-155
-------
March 25, 1975
TO THE RESIDENTS OF MILL CREEK WATERSHED
1. Name •
2. Address
3. Number of persons in your household
4. Type of land included in your property Lawn acres
(approximate): Forested acres
Grassland acres
Cropland acres
5. Do you have a septic tank, cesspool, or sewer connection?
(please circle)
6. How do you dispose of your household garbage? (please circle)
garbage disposal in sink compost pile
collected and hauled away other (please specify)
7. Do you keep animals other than house pets on your property?
horses - number pigs - number
cows - number chickens - number
other (please specify) number
How do you dispose of manure?
8. Do you have a vegetable garden or grow crops for your own consumption?
Yes No
If yes, please list crops acres
acres
acres
9. Do you grow crops beyond your own consumption? Yes No
10. Did you apply fertilizer to your lawn last year? Yes No
If yes, please describe amounts and types
11. Did you apply fertilizers to your garden last year? Yes.... No
If yes, please describe amounts and types
12. Do you plan to fertilize this spring? Lawn - Yes No,
Garden - Yes No,
If so, will you use more, less, or the same amount?
(Use back for further comment)
1-156
-------
March 25, 1975
TO THE RESIDENTS OF MILL CREEK WATERSHED
1. Name
2. Address
3. Number of persons in your household
4. Type of land included in your property Lawn acres
(approximate): Forested acres
Grassland acres
Cropland. acres
5. Do you have a septic tank, cesspool, or sewer connection?
(please circle)
6. How do you dispose of your household garbage? (please circle)
garbage disposal in sink compost pile
collected and hauled away other (please specify)
7. Do you keep animals other than house pets on your property?
horses - number pigs - number
cows - number chickens - number
other (please specify) number
How do you dispose of manure?
8. Do you have a vegetable garden or grow crops for your own consumption?
Yes No.....
If yes, please list crops acres
acres
acres
9. Do you grow crops beyond your own consumption? Yes..... No
10. Did you apply fertilizer to your lawn last year? Yes No
If yes, please describe amounts and types
11. Did you apply fertilizers to your garden last year? Yes.... No....
If yes, please describe amounts and types
12. Do you plan to fertilize this spring? Lawn - Yes...... No,
Garden - Yes No,
If so, will you use more, less, or the same amount?
(Use back for further comment)
1-157
-------
RENSSELAER POLYTECHNIC INSTITUTE
TROY, NY 12181
January 20, 1975
Dear Resident:
The attached summary describes a research project currently planned
for the Mill Creek area. This research will be conducted by two graduate
students, Kevin Walter and Deborah James, from the Environmental Engineer-
ing Department at R.P.I, in conjunction with the New York State Department
of Environmental Conservation.
In the near future we will be seeking from the residents of Mill Creek
Watershed, permision to obtain soil samples from forest, cropland and
grass- land areas. We would greatly appreciate your cooperation in helping
us obtain these soil samples and providing us with some information either
personally or by a forthcoming questionnaire.
We are looking forward to a productive research assignment in which
your help and interest are greatly needed and appreciated.
With best wishes (and a Happy New Year).
Sincerely yours,
(270-6369) Hassan M. El-Baroudi,
Faculty Advisor
(270-6367) Deborah James,
Graduate Student
(270-6367) Kevin Walter,
Graduate Student
1-158
-------
A Research Project Conducted by R.P.I. in Rensselaer County
Faculty Advisor: Hassan El-Baroudi - Environmental Engineering
Research Engineers: Kevin Walter and Deborah James - Graduate Students
INVENTORY OF NUTRIENTS IN MILL CREEK WATERSHED
(Oct. '74 - Sept. '75)
The soil of our planet is very important to us because it stores, con-
ditions and provides all living things with nutrients. During the animal
and plant cycles of life and death, the soil, together with water, air and
solar energy are responsible for dispersing, converting and transporting
various materials such that continuation of life becomes possible. It is
mainly through understanding these transformations and interactions that
the hope of our future welfare and survival lies. As a starting step,
engineers and scientists divide the total environment into separate units
or ecosystems, each of which has its specific activities within its bound-
aries and particular relations with adjacent units. Examples of these
ecosystems are the fresh water lakes, the swamps, forestlands and stream
watersheds. This project has the objective of studying the environmental
activities and cycles in Mill Creek Watershed which are responsible for
adding, losing or transforming three essential elements in plant and animal
nutrition: carbon, nitrogen and phosphorus. It is hoped that quantities
of these nutrients and their forms will be evaluated for the winter and the
summer seasons.
Mill Creek Watershed in Rensselaer County covers an area of about
10 square miles which is divided between the towns of North Greenbush and
East Greenbush. According to recent statistics, 55.4 percent of the water-
shed is forestland, 40.7 percent is used for agriculture and the remaining
area is occupied by water bodies, residential, recreational and urban
facilities. Small streams in the watershed join until they form Mill Creek
which crosses the City of Rensselaer and discharges into the Hudson River.
The inventory of nutrients in Mill Creek Watershed is only one part of
a larger study sponsored by the International Joint Commission, the U.S.
Environmental Protection Agency and the New York State Department of
Environmental Conservation. The Department will concurrently sample and
analyze Mill Creek on a daily basis. In addition, a research team from
Civil Engineering at R.P.I, will study the sediments in the streams. The
project is designed for the main purpose of providing knowledge and testing
techniques that are needed in the current studies of water quality in the
bordering lakes between the U.S. and Canada, and in the many similar set-
tings nationally and Internationally. Examples of the questions for which
the study is seeking answers are: How often should streams be sampled for
identifying water quality with a desired precision? How can the stream
sediments be sampled effectively and how do they interact with water
quality? What are the effects of the various activities and nutrients in
the drainage area and their seasonal fluctuations on the quality of stream
water?
1-159
-------
Needless to say, the task of inventory-taking of the nutrients on the
land can benefit greatly by the interest, input and communications with the
residents of the watershed. This contribution from residents will be
sought and welcomed by the Project team throughout the period of study.
1-160
-------
APPENDIX C
FORTRAN PROGRAM;
BEST STRAIGHT LINE FIT USING
LEAST SQUARES ANALYSIS
DIMENSION ICARD(80),X(4),Y(4)
DATA ISTAR /'*'/
IN-5
IOUT-6
120 READ (IN,100,END=300) ICARD
100 FORMAT (80A1)
IF (ICARD(l) .EQ. ISTAR) GO TO 200
BACKSPACE IN
N»4
SUMX=0.0
SUMY=0.0
SUMX2=0.0
SUMXY=0.0
READ (IN,110) (X(J),Y(J),J=1,4)
110 FORMAT (10X,4(F5.2,F10.4)
ICOUNT-ICOUNT+1
IF (X(4) .EQ. 0.0)N=3
DO 130 1=1,N
SUMX=SUMX+X(I)
SUMY=SUMY+Y(I)
SUMXY=SUMXY+(X(I)*Y(I))
SUMX2=SUMX@+X(I)**2
130 CONTINUE
CDET=(N*SUMX2) - (SUMX**2)
IF(ABS(CDET) .LT. l.OE-10) GO TO 190
YINT=((SUMY*SUMX2) - (SUMXY*SUMX))/CDET
SLOPE=((N*SUMXY) - (SUMX*SUMY))/CDET
WRITE (ICUT,140) ICARD
140 FORMAT (///,' DATA READ WAS: ',80A1,/)
GO TO 210
190 WRITE (IOUT,220)
220 FORMAT (' COEFFICIENT DETERMINANT .LT. l.OE-10, PROBABLE DATA ERRO
1R')
GO to 120
210 WRITE (IOUT,150) N,SLOPE,YINT
1-161
-------
150 FORMAT (' THE NUMBER OF POINTS IS f,Il,5X,'THE SLOPE IS ',
1F10.A, 5X,'THE Y-INTERCEPT IS \F10.4,
GO TO 120
200 WRITE (IOUT.160) ICARD
160 FORMAT ('I1,1X,80A1)
GO TO 120
300 WRITE (IOUT.170) ICOUNT
170 FORMAT(////,f PROCESSING COMPLETED - ',15,'SETS OF POINTS ANALYZE
ID1)
STOP
END
1-162
-------
REPORT II
EVALUATION OF THE BOGARDI T-3 BEDLOAD SAMPLER
by
Thomas F. Zimmie
Assistant Professor of Civil Engineering
Young S. Paik
Graduate Assistant
Carsten H.L. Floess
Graduate Assistant
Renssalaer Polytechnic Institute
Civil Engineering Department
Prepared for
New York State Department of Environmental Conservation
and
United States Environmental Protection Agency
-------
REPORT II
CONTENTS
Abstract *
Figures ii
Tables , iv
Acknowledgements • v
1. Introduction 1
Background 1
General 1
Bedload samplers 3
2. Conclusions 6
3. Laboratory flume experiments 7
Introduct ion 7
Equipment • 8
Procedure 10
Laboratory results and discussion \i
4. Field testing 39
General 39
Procedure 41
Analysis of field data 44
Comparison with bedload formulas 5g
References 65
Appendix: Field methods for fluvial bedload measurement using the
Bogardi T-3 bedload sampler 68
-------
ABSTRACT
This study consists of three parts:
I) Evaluation of the Bogardi T-3 bedload sampler in laboratory flume tests;
2) Field sampling using the Bogardi T-3 sampler;
3) Comparison of the field data .with theoretical values.
The Bogardi T-3 bedload sampler, designed by Professor Bogardi, was studied
in a non-circulating flume. Efficiency and selectivity of this pressure
difference-type sampler were evaluated at various water velocities and with
different bed sediment materials. Although the efficiency varied with
different velocities and bed materials, an overall efficiency of 40 percent
was found adequate for practical use. The selectivity of the sampler was
negligible, and the material sampled was representative of the bedload being
transported.
The sampler was used to measure the bedload transport in Mill Creek, which ~
is located-in eastern New York State. The watershed was approximately 26 km
in area above the sampling station.
Although the Bogardi T-3 sampler was originally designed for use in small
streams, it can also be used in large rivers.
The sediment-discharge relationships of bedload trensport versus water dis-
charge and bedload transport versus water velocity were obtained from field
sampling and USGS hydrological data for Mill Creek. The total amount of bed-
load transport was also computed.
Field measurements were compared with bedload formulas. The Schoklitsch for-
mula, the Meyer-Peter formula and the Einstein formula were examined. As for
most cobble-bedded streams, the mean size of the bed material in Mill Creek
is too large to give adequate amounts of bedload as computed by the formulas.
Reasonable agreement between computed results and actual field data could be
obtained by using the mean diameters determined from bedload samples instead
of the mean diameter of the bed material.
Il-i
-------
FIGURES
Number Page
1 Bogardi T-3 bedload sampler 5
2 Flume arrangement 9
3a Grain size distribution - Grit #2 22
3b Grain size distribution - Grit #1 22
3c Grain size distribution - Grit #1/2 23
3d Grain size distribution - Grit #40 23
3e Grain size distribution - Mixture 24
4a Total sediment movement vs. velocity - Grit #2 26
4b Total sediment movement vs. velocity - Grit #1 26
4c Total sediment movement vs. velocity - Grit #1/2 27
4d Total sediment movement vs. velocity - Grit #40 27
4e Total sediment movement vs. velocity - Mixture 28
5 Total sediment movement vs. D-0 size. 28
6a Velocity vs. efficiency - Grit #2 30
6b Velocity vs. efficiency - Grit #1 30
6c Velocity vs. efficiency - Grit #1/2 31
6d Velocity vs. efficiency - Grit #40 31
6e Velocity vs. efficiency - Mixture 32
7a Grain size distribution - Test #64 35
7b Grain size distribution - Test #68 35
Il-ii
-------
FIGURES (continued)
Number Page
7c Grain size distribution - Test #71 36
8 Mill Creek watershed, Renssalaer County, New York AO
9 Method of bedload sampling for low flow periods... 42
10 Method of bedload sampling for high flow periods 42
11 Bedload sediment-discharge relation for Mill Creek 45
12 Bedload sediment-discharge relation for Clearwater River 45
13 Bedload sediment-velocity relation for Mill Creek 52
14 Bedload sediment-discharge relations for Mill Creek: field
data and bedload formulas •. 64
-------
TABLES
Number Page
1 Laboratory data 12
2 Material characteristics 1?
3 Efficiency of the Bogardi T-3 bedload sampler -
statistical summary 33
4 Bedload data '. 46
5a Rating table for Mill Creek (11/7/74 - 4/2/75) 53
5b Rating table for Mill Creek (from 4/2/75) 54
6a Discharge occurrence (1/23 - 4/2/75) 55
6b Discharge occurrence (3/3/75 - 6/2/76; missing 8/22-
8/27/75) 56
Il-iv
-------
ACKNOWLEDGEMENTS
The authors wish to thank M. Comer, who was involved with the early labora-
tory and field testing, J. Murray and K. Nagel, who carried out the labora-
tory calibration tests, Drs. L. Hetling and G.A Carlson, New York State
Department of Environmental Conservation, and many members of the Water
Resources Division, Albany Office of the United States Geologic Survey, who
cooperated on this project. The authors also wish to thank Betty Alix who
typed this report.
This study was carried out as part of the Task C activities of the Pollution
from Land Use Activities Reference Group, International Joint Commission'and
was funded through the United States Environmental Protection Agency and the
State of New York.
II-v
-------
SECTION 1
INTRODUCTION
BACKGROUND
The International Joint Commission (IJC), consisting of Canadians and
Americans, has established various reference groups to study specific USA -
Canada water quality problems. These problems deal either directly or
indirectly with the Great Lakes. A number of watershed studies are being
carried out as part of the activities of the Pollution from Land Use
Activities Reference Group (PLUARG). This study was carried out as part of
the Genesee River Watershed Project.
Sedimentation investigations are an important part of the watershed
studies. In order to measure total sediment transport in a watercourse,
both suspended sediments and bedload sediments must be sampled. The -
specific aim of this research was to evaluate and calibrate a bedload
sampler for use in the sedimentation studies. Although this research was
performed as part of the Genesee River Watershed Study, the intent is to
utilize the bedload sampler in other USA - Canada watershed studies also.
GENERAL
Total sediment load in a stream can be considered to consist of two
types of sediment, bedload and suspended load. Bedload can be defined as
sediment that is transported in a stream by rolling, sliding, or skipping
along the bed or very close to it. Einstein (1948) suggested that the
thickness of the layer within which bedload moves is different for each
watercourse; it should be at least equal to the diameter of the largest
grain. The suspended load may be defined as that sediment which does not
spend any time on the bed of the channel. These sediments are carried
along at approximately the same velocity as the water phase. The
separation between bedload and suspended load is rather artificial, since
in reality there will be interchanges between the two. The behavior of a
particle of sediment in a stream is dependent both on the characteristics
of the sediment (weight, size, shape, etc.) and the fluid flow conditions
(velocity, direction, turbulence, etc.) (Graf, 1971). In a quiet water
zone suspended particles will settle and become part of the bedload,
whereas after storm events, high stream flow velocities will place bedload
sediments in suspension. Nevertheless, for a given sampling event at a
given sampling station, the concept of bedload and suspended load Is
practical.
II-l
-------
Bedload sediment investigations have not been as numerous in the
United States as those of suspended load because many of the important
rivers have beds composed of fine material and therefore carry much more
sediment in suspension than along the bed (Carlson and Miller, 1956). For
those watercourses, the bedload portion of the stream sediments usually
accounts for only a small percentage of the total sediment discharge
(Colby, 1964). In streams where the sediment load consists primarily of
silt and clay sized particles, the suspended sediment discharge is very
nearly equal to the total discharge (ASCE, 1969). However, the ratio of
bedload transport to total sediment transport is highly variable, and
depends on the character of the watercourse. For example, on the Middle
Loup River at Dunning, Nebraska, more than 50 percent of the total is
bedload whereas on the Milk River near Nashua, Montana, the bedload is
estimated at 5 to 10 percent of the total transport (Carlson and Miller,
1956).
In some streams, the sediment discharge can be measured in a
contracted section of a channel where the velocity and the turbulence are
sufficient to suspend all the particles in transport. Such a section is
sometimes called a natural turbulence flume. In small streams, it may be
practical to install an artificial turbulence flume (ASCE, 1969). For
other methods of measuring sediment discharges, see (ASCE, 1969).
This study is concerned only with the measurement of bedload sediment.
Movement of bedload is largely responsible for changes in the geometry of
alluvial channels. Thus it is virtually impossible to plan, design, and
maintain river basin projects without quanitifying bedload transport rates
(ASCE, 1969).
Assessment of pollution effects requires not only knowledge of the
efficiency of the sediment as a carrier, but also knowledge of quantities
and rates of sediment transport (Tywoniuk, 1972). In order to estimate the
amount of polychlorinated biphenyls trapped on bedload sediments in the
Hudson River, some samples were obtained using the Bogardi sampler.
The current trend in determining the total sediment-transport rate of
a watercourse is to combine an analytical method with a suspended sediment
sampling program. The transport rate of coarse particles (bedload) is
computed according to the existing theories, and the measured results
(suspended sediments) are used to determine the transport rate of fine
particles (Chien, 1954). Thus there have been many more measurements of
suspended load than of bedload. One of the reasons is the large number of
practical problems and difficulties in obtaining satisfactory bedload
samples.
Several computational methods for determining bedload or total load
have been developed, but none is universally acceptable to all sediment
sizes, bed configurations, and flow regimes (Graf, 1971). Also, there are
complications in the actual application of these formulas. These methods
include both analytical and empirical relations. Thus, direct measurements
are necessary to determine the amount of sediment load or to check these
computational methods (Graf, 1971).
11-2
-------
Bedload samplers for the direct measurement of bedload have been
developed and used in many countries in Europe, including Switzerland,
Germany, Hungary, Russia, and Poland. Calibration of these samplers have
indicated a mean efficiency of about 45 percent for basket or pan types of
samplers and about 70 percent for the best designed pressure difference
type. The efficiency of a sampler is the ratio of trapped sediment to that
actually moved as bedload per unit time. The efficiency varies with
sampler characteristics, method of supporting the sampler, hydraulic
conditions, particle size, bed stability, and bed configuration (ASCE,
1969).
Development of bedload samplers in the United States has been largely
limited to minor changes in design of some of the European models without
calibration tests (Hubbell, 1964). However, the USGS has recently devel-
oped a bedload sampler primarily intended for California coastal streams
(Helley and Smith, 1971). Thus the need for the development and calibra-
tion of bedload samplers is apparent.
BEDLOAD SAMPLERS
Bedload samplers come in a large array of types. Each type has been
developed to serve satisfactorily under slightly different conditions.
Usually the difference among the samplers lies in the type of sample it
will efficiently catch. Some samplers catch primarily large size
sediments, while others trap a broader range of sizes, Including the
smaller size particles.
The basic types of samplers are: the box- or basket-type, the
pressure difference-type, the pan- or tray-type, and slot- or pit-type.
The box or basket samplers operate by retaining sediment that is
deposited in the sampler because of a reduction in the flow velocity and
(or) that is screened from the flow (Hubbell, 1964). The basket-type only
catches material larger than the mesh of the basket, while the box-type is
able to catch a large number of fines in its baffled settling chamber. In
general, the basket-type has been adopted in preference to the box-type,
and the average efficiency of a basket sampler is about 45 percent
(Hubbell, 1964). The basket samplers were used not only for solving
specific river problems but also for checking the validity of bedload
formulas.
Due to the flow resistance of the box- or basket- type samplers, a
velocity decrease and pressure increase at the entrance are produced. As a
result, efficiencies are extremely variable and samples are unrepresenta-
tive. To remedy these disadvantages, pressure difference-type samplers are
designed so that the entrance velocity and the stream velocity are approxi-
mately the same. The bedload settles out of the water as it passes through
the low pressure area towards the rear of the device. The low pressure is
caused by an expansion in a flow chamber. The details and geometry vary
-------
with the type of sampler. The efficiencies of these samplers may approach
70 percent, depending on site conditions (Graf, 1971).
The Helley-Smith Sampler is an example of the modified basket type and
the Bogardi is a modified box-type. The Bogardi is similar to that devel-
oped by Karolyi (Graf, 1971) in design and concept, although the model used
in the present tests is much smaller than Karolyi's 90 kg device. A
refinement of this sampler, called the VUV by Novak (Graf, 1971) has a
higher effficiency, closer to 70 percent. This refinement was attained by
studying the exact flow characteristic inside the box to determine the best
shape for the plates. It has been found that these samplers will also
provide improved performance when provided with flexible bottoms and extra
weight for stability on the stream bottom.
The pan-type is designed for small quantities of bedload in streams of
low velocity. This type consists of a flat pan placed on the bed which may
or may not have vertical baffles to hold the bedload material which passes
over the top of the pan (Graf, 1971).
The pit-type is a very simple type of sampler and consists of an exca-
vation in the stream bed. This pit is allowed to fill up with bed
materials over a specified length of time. At predetermined intervals, the
pit is re-excavated and the sample recovered. This type of sampler is most
accurate in larger rivers with very high flows and sediment loads, where
the errors inherent in such a measurement would be insignificant (Hubbell,
1964). However, in spite of their high and relatively constant efficiency,
the use of pit-type samplers is limited to the shallow rivers because they
must be placed into the stream bed. In order to overcome this deficiency,
a diver-operated bedload sampler has been developed for deep rivers. A
scuba-equipped diver places a portable pit-type sampler into the stream
bed. According to Waslenchuk (1976), this type sampler is technically and
economically feasible. For a further discussion of bedload samplers see
Hubbell (1964) or Report 2, "Equipment Used for Sampling Bed Load and Bed
Material," of the Federal Inter-Agency River Basin Committee (1940).
The bedload sampler that was evaluated in this project was the Bogardi
T-3 Bedload Sampler (Figure 1), developed by Professor J. L. Bogardi of
Hungary. The sampler belongs to a family of similar pressure difference-
type bedload samplers developed by Karolyi (Hubbell, 1964). Novak (1959)
found in laboratory calibration tests that the sampling efficiency of these
samplers was about 45 percent. Although this efficiency may be considered
low, one of the advantages of this type of sampler is that the sampling
efficiency does not vary radically with velocity or particle size as does
the efficiency of most other trap-type samplers (Hubbell, 1964). Novak's
results were similar to those found in this project for the Bogardi
sampler, as will be discussed later.
The sampler was constructed of 1.6 mm sheet aluminum plates and 6.4 mm
plexiglass walls. The sampler is 51 cm long, 14 cm wide, and 23 cm high
(not including the stabilizing fin). The front opening is 10.2 cm high by
14 cm wide. The sampler weighs only 2.3 kg; however, weights can be added
II-4
-------
to increase the sampler stability in high water velocities. In this
project the maximum amount of weights used was 13.6 kg, which increased the
total sampler weight to 15.9 kg. More weight can be added if necessary.
The Bogardi T-3 Bedload Sampler has been used to a limited extent in
North America, primarily by Agriculture Canada to determine the amount of
pesticides on bedload sediments. In addition, a number of other organiza-
tions involved with IJC Great Lakes research projects have utilized the
sampler on occasion. However, the sampler had not been calibrated
previously, since these research projects did not deal with sedimentation
per se.
Ah*
Figure 1. Bogardi T-3 Bedload Sampler.
II-r5
-------
SECTION 2
CONCLUSIONS
From this report, the following conclusions may be drawn:
The Bogardi T-3 Bedload Sampler is primarily intended for use in small
streams and rivers. However, it is a versatile sampler and can be used in
large rivers also. It is very lightweight and highly portable.
The use of an overall sampling efficiency of 40 percent is justified
for field work with the Bogardi sampler.
No evidence of sample selectivity was indicated with the Bogardi
sampler, and the sample obtained can be considered to be a representative
portion of the bedload sediment being transported.
The sampler has been successfully used in water velocities of 2.4 m/s,
and can be used in higher velocities.
A bedload sediment-discharge relation was obtained for Mill Creek
using the Bogardi sampler.
Satisfactory bedload sediment-discharge relations can be obtained for
Mill Creek by the use of bedload formulas if the appropriate sediment size
parameter is obtained from bedload samples rather than from bedload
material.
Total bedload transport for Mill Creek was about 100,000 kg during the
period January 23, 1975, to June 2, 1976, an average daily value of about
240 kg/day. However, bedload transport is very event-dominated, and about
65 percent of the bedload movement occurred during the 10 days of highest
flows.
II-6
-------
SECTION 3
LABORATORY FLUME EXPERIMENTS
INTRODUCTION
The laboratory experiments were primarily concerned with the
controlled calibration of the Bogardi Bedload Sampler as used in the field
tests. Such measurements require the use of a flume to simulate the stream
conditions, since the calibration cannot be determined in natural streams
because the actual transport rate is unknown (Helley and Smith, 1971).
Therefore, in order to calibrate the sampler, a flume was required which
develops the characteristics critical for bedload movement in a manner
similar to that of the stream to be measured.
The important parameters that affect bedload movement can usually be
limited to sediment size characteristics, sediment transport rate, water
discharge, mean velocity, depth, bed slope and roughness. Determinations
of the parameters to be controlled must be made by the experimenter in each
type of experiment. True independence or dependence of variables is dif-
ficult to determine, as the interaction among them is not well understood
(Williams, 1967 and Yalin, 1972). In some reported experiments (Williams,
1967, 1970) the independent variables were the sediment characteristics,
water depth and sediment transport rates, while the dependent variables
were water discharge (mean velocity), slope and bed roughness. These types
of experiments' were done using a non-recirculating-type flume (see Williams
(1970) and Guy, Simons and Richardson (1966) for additional explanation and
diagrams). Other experimenters have used a recirculating-type flume
(Simons, Richardson and Haushild, 1963). The discharge rate, sediment size
and sediment load were chosen, and the slope, depth and bed roughness were
allowed to change in accordance with each other. In general, the flume
arrangement, the selection of variables and experimental procedure are
governed by the results to be obtained.
The flume in which the experiments were conducted allows a number of
hydraulic variables to be controlled. However, for the efficiency deter-
mination required for this project, the effects of water velocity (hence
discharge rate) and sediment size characteristics were considered. It was
found that the water depth had no effect on efficiency, as long as the
depth of the water was sufficient to completely cover the sampler. This is
an important point, since the efficiency is essentially zero if the sampler
is not covered completely. Thus, a minimum water depth of 25-30 cm is
required to sample.
II-7
-------
Five different kinds of sediment were used. These included four sands
of fairly uniform size, and one sand that was an equal mixture of the four
uniform sands. The V$Q sizes of the sand varied from 0.49 mm to
1.65 mm. A maximum water velocity of 0.82 m/s was used in the flume.
After each test, the amount of bedload sediment trapped in the sampler and
the total amount of bedload movement was determined. Efficiency, expressed
in percent, was calculated by dividing the weight of sampler sediment by
the total weight of bedload sediment movement. For example, if 4.5'kg of
bedload sediment passes by a 14 cm width of stream bed (the width of the
sampler) in a given time, and 2.3 kg of sediment is trapped in the sampler
during the same unit of time, the efficiency is 50 percent.
EQUIPMENT
The flume was a non-recirculating-type (sediment did not recirculate)
with a usable length of approximately 7.3 m and a width of 74 cm. The
flume arrangement is shown in Figure 2. The bottom was horizontal (no
slope) and non-adjustable. Two pumps were used to circulate water. One
pump had a rated capacity of 2.66 m^/min and fed a 15.2 cm diameter line.
The other pump had a rated capacity of 9.5 m-Vmin and fed a 26.7 cm line.
The pumps discharged into a stilling area and then went through a baffled
chamber, to reduce excess turbulence before entering the mainstream.
At a distance of 137 cm downstream of the last baffle, a plywood board
97 cm long and extending the entire width of the flume was secured to the
flume bottom. A small metal strip acting as a ramp was placed in front of
this board, and the board had a sand coating affixed to it to give it a
rough surface. The purpose of the board and the metal strip was to have
the flowing water approach and impact the sand bed in a relatively uniform
manner. The Bogardi T-3 Sampler was placed on a similar plywood board
51 cm long, with a rough finish and sand particles affixed to it. This
board was added to the flume to decrease the degree of observed scour.
Between these two boards was the sand bed. The sand bed had an overall
length of 2.57 m and a depth of 4.4 cm (this depth was equal to the
thickness of the plywood boards). The collection box was placed directly
behind the sampler. This box was constructed of aluminum, and consisted of
two pieces to facilitate its handling and removal from the flume. When in
place, the box had an overall length of 180 cm and a width of 30 cm. Since
it was desirable to negate effects created at the flume edges, the width of
the collection box was less than the overall flume width. The aluminum
utilized for the collection box was 0.8 mm thick.
To adjust the velocity and height of the water, three weirs were used.
The heights of these weirs were 38.1, 25.4 and 17.8 cm. Either of these
weirs could be placed in the slots at the extreme end of the flume.
A Price-type current meter was used to measure water velocity rates
and determine discharge rates. This is a cup-type current meter used
throughout the country by the USGS for stream gaging (Grover and
Harrington, 1966).
II-8
-------
PUMP
DISCHARGE
PIPES
BAFFLES
METAL STRIP
PLYWOOD
BOARD
SAND BED
REMOVABLE
METAL PLATE
PLYWOOD
BOARD
BOGARDI T-3
SAMPLER
COLLECTION
BOX
WEIR
W+4+-
T
104 cm
137 cm
97 cm
257 cm
51 cm
180cm
15. cm
FIGURE 2. Flume arrangement.
II-9
-------
PROCEDURE
The first step in a test was to prepare the sand bed. The sand ac-
cumulated behind the sampler from a previous test was removed and placed in
the bed area. This was done by using a trowel and shoveling the sand into
the bed area. When using a mixture, the sand was mixed thoroughly before
being smoothed out into a level and uniform layer. The collection box was
then secured in place by placing it into pre-cut notches in the plywood on
which the sampler rests. The collection box also had some lead weights
placed in it so that it would not move while the water was flowing.
Next the sampler was placed on the center of the plywood board and
secured in place. The sampler was placed so that its upstream opening
coincided with the edge of the plywood board. Thus, the sampler was
located at the edge of the sand bed. A piece of sheet metal measuring
25.4 cm by 30.5 cm was placed on the bed directly in front of the sampler
and a small weight was used to secure it to the sand bed. The purpose of
this piece of metal was to help reduce scour which occurred as the initial
wave of water reached the sampler.
It was necessary to fill the flume with water before a test could be
run. The filling operation was performed slowly to prevent disturbance of
the sand during this period of constantly changing flow conditions across
the sand surface. First, the area behind the sampler was filled using a
1.9 cm hose. When the water started to flood the sand bed, the 2.66
m^/min pump was started in a throttled down condition, to fill the flume
from the opposite direction. These opposing flows prevented disturbances
and wash-outs of the sand near the sampler.
Once the water level reached weir height, the 9.5 m/min pump was
started and the other pump was throttled to its desired position. The
timer for the experiment was started and the small metal plate in front of
the sampler was removed from the flume. The water was circulated for a
predetermined length of time. At the conclusion of the test, the pumps
were shut down and the sampler removed from the flume. Some of the water
in the flume was allowed to drain by raising the weir. Sand found on the
plywood board directly in front of the collection box was shoveled into the
box. This sand was considered to be part of the overall sample, since it
passed the cross section containing the upstream opening of the sampler.
Since the collection box was constructed in two pieces, it was necessary to
collect the sand into one box while the box was still in the flume. This
was done by tilting one box and washing the sand into the other half. This
half was then removed from the flume and the sand was trowelled, brushed,
and washed into pans and bowls. The sample obtained in the Bogardi T-3
Sampler was also placed into pans and bowls through a similar procedure.
These pans and bowls were then placed into an oven and the sand was allowed
to dry. With unifom sands it was only necessary to weigh the dried samples
from the Bogardi T-3 Sampler and the collection box, to compute the effi-
ciency. If a mixture was being tested, sieve analyses of both samples were
conducted, in addition to the calculation of overall efficiency. The sieve
analyses were performed in accordance with ASTM standards (ASTM Designation
11-10
-------
C136-46) using U.S. Standard Sizes #10, 20, 40, 60, 100 and 200 sieves.
LABORATORY RESULTS AND DISCUSSION
The laboratory data is presented in Table 1. The most common test
velocities were 0.427, 0.595 and 0.824 m/sec. These velocities corre-
sponded to maximum pumping rates using the 38.1, 25.4, and 17.8 cm weirs
respectively. It was preferable not to take velocity measurements during
every test, since the presence of the current meter could have a signif-
icant effect upon the efficiency measurement. However, the velocities were
verified routinely during the experimental phase and remained fairly con-
stant. The water velocity was measured near the upstream opening of the
sampler and the recorded velocity is an average of the velocity readings.
Table 1 also lists depth of water, weight of material found in the sampler,
total weight of bedload movement, rate of bedload movement, and a computed
sampler efficiency. The efficiency is computed from:
width of collection box weight in sampler
Efficiency = width Qf sampler « total bedload weight
where: width of collection box « 30.48 cm,
width of sampler - 13.97 cm.
Then, as shown in Table 1:
ws
E •
.458 x Wt
where: E = efficiency
Wg • weight in sampler
Wt = total bedload weight.
Material size characteristics of the sands used in the tests are shown
in Table 2. The data was obtained from mechanical sieve analyses. The
sieve size openings are shown in the table just below the sieve number.
Also shown are the results obtained utilizing the sand mixture, including
size analyses of material retained in the sampler, material obtained in the
collection box, and the material in the sand bed. For each sand analyzed,
the D , D , and D sizes are shown. The D size is defined as the soil
particle diameter at which 60 percent of the soil weight is finer, DSQ
is the diameter at which 50 percent of the soil weight is finer, and
D,Q is the corresponding value at 10 percent finer (Lambe and Whitman,
1969). The uniformity coefficient (C ) is the ratio of the D60 size
to the DIQ size (Lambe and Whitman, 1969), and is also shown in
Table 2. All of the sand used in the tests will pass a #4 sieve. Complete
grain size curves for the mono-sized sands and a typical bed sample of the
mixture can be found in Figures 3a-3e.
11-11
-------
TABLE 1. LABORATORY DATA.
Test
*
1-7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
Date
1976
1/8
1/9
1/11
1/11
1/11
1/11
1/11
1/14
1/14
1/14
1/16
1/16
1/16
1/19
1/19
Grit
i
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
Time
(min)
5
9
30
6
6
6
26
32
31
10
10
10
5
8
15
Vel.
(mps)
0.59
0.59
0.41
0.64
0.64
0.64
0.52
0.45
0.46
0.82
0.82
0.82
0.82
0.59
0.58
Depth
(m)
PILOT
0.34
0.34
0.34
0.24
0.24
0.24
0.32
0.31
0.31
0.27
0.27
0.27
0.27
0.34
0.23
Sample wt.
V (gin)
8
TEFTS
198.5
681.0
17.2
754.0
1218.0
2198.0
725.0
17.0
0.0
1276.0
3749.0
1796.0
903.0
58.0
—
Tot. wt.
1351.5
3004.0
1042.5
4421.0
5015.0
830.9'
3840.0
464.0
237.0
7433.0
10820.0
7236.0
6073.0
1642.0
—
Rate
(gm/min)
269.6
333.7
34.8
736.8
985.8
1384.8
133.8
14.5
7.6
743.3
1082.0
723.6
1214.6
205.3
—
Efficiency
W /.458W
s t
32.1
49.5
3.7
37.2
45.0
57.8
45.5
8.0
0.0
37.5
75.7
54.2
32.5
7.7
—
(continued)
-------
TABLE 1 (continued).
Test
1
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
Date
1976
1/19
1/19
1/20
1/20
1/21
1/21
1/21
1/28
1/28
1/28
1/28
1/20
1/30
2/2
2/2
2/2
Grit
1
1
1/2
1/2
1/2
1/2
1/2
1/2
1/2
1/2
1/2
1/2
1/2
40
40
40
Time
(min)
15
15
15
10
7
20
5
7
20
30
40
-
35
10
10
10
Vel.
(mps)
0.58
0.58
0.59
0.59
0.82
0.59
0.82
0.82
0.43
0.43
0.43
0.82
0.46
0.50
0.59
0.82
Depth
0.23
0.23
0.34
0.34
0.27
0.34
0.27
0.27
0.43
0.43
0.43
0.24
0.29
0.34
0.34
0.27
Sample wt.
W (gm)
S
—
—
615.0
305.0
2110.0
1808.0
2680.0
2358.0
110.0
307.0
270.0
—
0.0
621.0
491.0
990.0
Tot. wt.
—
—
4670.0
3490.0
8172.0
8142.0
8969.0
10334.0
1886.0
1692.0
2426.0
—
0.0
4696.0
3031.0
6881.0
Rate
(gtn/min)
—
—
311.3
349.0
1167.4
407.1
1793.8
1476.3
94.3
56.4
60.7
—
0.0
469.6
303.1
688.1
Efficiency
W /.458W
S t
—
—
28.8
19.1
56.4
48.5
65.2
49.8
12,7
39.6
24.3
—
—
28.9
35.4
31.4
(continued)
-------
TABLE 1 (continued).
Test
#
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
Date
1976
2/4
2/4
2/4
2/6
2/6
2/6
2/9
2/9
2/9
2/9
2/10
2/10
2/11
2/11
2/11
2/16
Grit
f
40
40
40
40
40
40
40
2
2
2
2
2
2
2
2
2
Time
(min)
15
10
17
15
15
25
20
15
15
15
20
10
10
6
5
30
Vel.
(rops)
0.82
0.82
0.82
0.59
0.43
0.43
0.43
0.82
0.82
0.82
0.82
0.82
0.82
0.82
0.82
0.59
Depth
(tn)
0.27
0.27
0.27
0.34
0.43
0.43
0.43
0.27
0.27
0.27
0.27
0.27
0.27
0.27
0.27
0.34
Sample wt.
W (BB)
9
2063.0
1130.0
3252.0
912.0
112.0
449.0
81.0
1387.0
455.0
2814.0
2358.0
1225.0
655.0
305.0
496.0
1716.0
Tot. wt.
Wt(*m)
12182.0
6774.0
11147.0
4242.0
780.0
2748.0
1108.0
6044.0
4978.0
12494.0
8080.0
5621.0
3528.0
2200.0
2893.0
6983.0
Rate
(ptn/min)
812.1
677.4
655.7
282.8
52.0
110.0
55.4
456.3
331.9
832.9
404.0
561.2
352.8
366.7
578.6
232.7
Efficiency
W /.458W
s t
37.0
36.4
63.7
46.9
31.4
35.7
16.0
44.2
20.0
49.2
63.7
47.7
40.5
30.3
37.4
53.7
(continued)
-------
TABLE 1 (continued).
Test
f
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
Date
1976
2/16
2/16
2/18
2/18
2/18
2/18
2/20
2/20
2/20
2/25
2/27
2/27
2/27
3/1
3/1
3/7
Grit
1
2
2
2
2
2
2
Mix
Mix
Mix
Mix
Mix
Mix
Mix
Mix
Mix
Mix
Time
(min)
15
30
32
30
22
135
10
10
15
10
15
20
30
45
60
30
Vel.
(mps)
0.59
0.59
0.59
0.43
0.82
0.43
0.82
0.82
0.59
0.82
0.59
0.59
0.43
0.43
0.43
0.59
Depth
(in)
0.34
0.34
0.34
0.43
0.27
0.43
0.27
0.27
0.34
0.27
0.34
0.34
0.43
0.43
0.43
0.34
Sample wt.
W (gm)
s
—
401.0
2175.0
0.0
4225.0
0.0
3156.0
5586.0
278.0
3515.0
154.0
1277.0
51.0
745.0
874.0
2946.0
Tot. wt.
Wt(pn)
—
3408.0
6327.0
0.0
16690.0
0.0
11755.0
17195.0
3536.0
11940.0
4106.0
7510.0
1396.0
3363.0
3453.0
10939.0
Rate
(gm/min)
—
113.6
197.7
0.0
758.6
0.0
1175.5
1719.5
235.7
1194.0
273.7
375.5
46.5
74.7
57.6
364.6
Efficiency
W /.458W
S t
—
25.7
75.1
—
55.3
—
58.6
70.9
17.2
64.3
8.5
37.1
8.0
48.4
55.3
58.8
(continued)
-------
TABLE 1 (continued).
M
M
1
1— i
Test
t
71
72
73
74
75
76
77
78
Date
1976
3/7
3/12
3/12
3/12
3/12
3/19
3/19
3/19
Grit
1
Mix
Mix
Mix
Mix
Mix
Mix
Mix
Mix
Time
(min)
20
15
12
7
10
40
50
30
Vel.
(mps)
0.59
0.59
0.82
0.82
0.82
0.52
0.52
0.52
Depth
(m)
0.34
0.34
0.27
0.27
0.27
0.30
0.30
0.30
Sample wt.
Wg(gm)
1987.0
798.0
2069.0
403.0
463.0
170.0
309.0
112.0
Tot. wt.
Wt(gm)
9316.0
4252.0
11233.0
3833.0
6186.0
2762.0
2939.0
1602.0
Rate
(gm/tnin)
465.8
283.5
936.1
547.6
618.6
69.1
58.8
53.4
Efficiency
W /.458W
8 t
46.6
41.0
40.2
23.0
16.3
13.5
23.0
15.3
-------
TABLE 2. MATERIAL CHARACTERISTICS.
Grain Size
Analysis
Grit #2
Grit #1
Grit #1/2
Grit #40
Test #61
Sample
Test #61
Box
Test #62
Sample
Test #62
Box
Test #63
Sample
Test #63
Box
Test #64
Sample
Test #64
Box
# 10
(2.00)*
63.02
99.60
99.87
99.91
92.71
91.21
93.19
90.04
92.68
89.64
93.13
89.15
Percent
1 20
(0.84)
2.16
22.23
65.39
72.03
49.48
45.78
48.35
44.65
46.68
42.12
46.57
40.24
Passing
# 40
(0.42)
1.04
1.10
0.66
38.28
12.04
10.62
11.11
11.15
16.89
10.58
10.66
7.79
Sieve #
# 60
(0.25)
0.47
0.29
0.10
9.81
3.22
2.77
2.77
2.71
8.50
7.98
2.49
1.37
Size in mm.
#100
(0.15)
0.13
0.11
0.05
0.07
0.09
0.05
0.02
0.00
0.07
0.04
0.03
0.05
#200
(.074)
0.03
0.09
0.04
0.05
0.06
0.03
0.01
0.00
0.03
0.03
0.00
0.04
Pan
0.01
0.01
0.00
0.01
0.01
0.01
0.01
0.00
0.00
0.00
0.00
0.02
D
max
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
D
60
1.90
1.28
0.75
0.57
1.03
1.10
1.05
1.15
1.03
1.13
1.08
1.17
D
50
1.65
1.15
0.65
0.49
0.86
0.91
0.87
0.95
0.89
0.97
0.90
0.95
D
10
1.02
0.57
0.45
0.26
0.37
0.41
0.40
0.40
0.28
0.39
0.40
0.47
C
u
1.86
2.24
1.68
2.23
2.78
2.68
2.65
2.91
3.68
2.90
2.70
2.49
(continued)
-------
TABLE 2 (continued).
Percent Passing Sieve #
Grain Size
Analysis
Test #65
Sample
Test #65
Box
Test #66
Sample
Test #66
Box
V Test #67
55 Sample
Test #67
Box
Test #68
Sample
Test #68
Box
Test #69
Sample
Test #69
Box
# 10
(2.00)*
96.38
95.76
95.66
91.40
90.97
90.34
91.74
88.66
91.28
90.26
# 20
(0.84)
45.41
37.92
52.97
43.47
54.85
41.82
43.53
41.95
48.48
46.78
I 40
(0.42)
19.21
8.65
12.68
10.33
31.80
11.67
12.20
12.16
15.03
16.56
# 60
(0.25)
9.44
2.05
3.20
2.42
15.86
3.59
2.90
3.16
4.30
5.10
#100
(0.15)
1.35
0.22
0.38
0.22
1.94
0.45
0.27
0.26
0.03
0.30
#200
(.074)
0.00
0.00
0.10
0.02
0.02
0.09
0.03
0.03
0.41
0.03
Pan
0.00
0.00
0.08
0.00
0.00
0.00
0.00
0.01
0.00
0.01
D
max
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
Size in mm.
D
60
1.00
1.12
0.96
1.15
0.95
1.18
1.15
1.19
1.10
1.05
D
50
0.85
0.97
0.79
0.96
0.72
1.00
0.96
0.98
0.89
0.86
D
10
0.26
0.44
0.37
0.40
0.21
0.38
0.38
0.38
0.31
0.34
C
u
3.85
2.55
2.63
2.88
4.52
3.10
3.07
3.13
3.54
3.08
(continued)
-------
TABLE 2 (continued).
Grain Size
Analysis
Test#61-69
Bed
Test #70
Sample
Test #70
Box
Test #71
Sample
V Test #71
S Box
Test #70-71
Bed
Test #72
Sample
Test #72
Box
Test #73
Sample
Test #73
Box
# 10
(2.00)*
90.52
97.21
95.13
96.88
96.88
96.81
96.14
94.18
96.43
95.25
Percent
# 20
(0.84)
40.77
43.47
37.98
45.02
46.49
43.32
42.22
39.24
45.34
41.78
Passing
# 40
(0.42)
10.70
9. 15
7.38
9.64
10.30
7.94
7.94
7.80
8.54
6.64
Sieve #
# 60
(0.25)
2.68
1.94
1.27
1.89
1.53
1.03
1.64
1.30
1.23
0.86
Size in mn.
#100
(0.15)
0.22
0.19
0.12
0.15
0.10
0.07
0.08
0.10
0.08
0.07
#200
(.074)
0.02
0.01
0.01
0.02
0.03
0.02
0.00
0.02
0.00
0.00
Pan
0.00
0.00
0.00
0.00
0.00
0.01
0.00
0.00
0.00
0.00
D
max
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
2.00
D
60
1.10
1.12
1.18
1.10
1.07
1.10
1.14
1.19
1.10
1.15
D
50
0.95
0.95
1.02
0.93
0.89
0.92
0.97
1.00
0.93
0.96
D
10
0.39
0.46
0.45
0.44
0.42
0.44
0.43
0,46
0.42
0.46
C
u
2.82
2.43
2.62
2.50
2.55
2.50
2.65
2.59
2.62
2.50
> _ . . . , j i - M
(continued)
-------
TABLE 2 (continued).
Percent Passing Sieve #
Grain Size # 10 1 20 # 40 I 60 #100 #200
Analysis (2.00)* (0.84) (0.42) (0.25) (0.15) (.074) Pan
Test #74 94>28 4Q>60 9>18 1>92 oao 0>QO Q>00
Sample
Test #74 95>91 40>77 6>6? 0>?7 Oi04 0>f)0 0>00
Box
Test #75 96>14 52>26 13<20 1>% 0>20 0>OQ 0>OQ
Sample
Test #75 9fi 5g 42>81 6>Q2 0>63 Q>05 0>OQ 0>00
Box
M
.i, Test#72-75 97_lg 49>Q9 g>86 1>y6 0>53 0>03 0>00
o Bed
Test #76 96<87 4g>71 u<22 1<83 0>12 0>OQ 0<00
Sample
Test #76 %>59 39<29 6>2? Q>60 0>Q3 Q<01 0>00
Box
Test #77 97.06 43.65 7.14 0.42 0.03 0.00 0.00
Sample
Test #77 96>52 39>16 5>59 Q>65 0>39 0>09 0>no
Box '
Test#76-77 g5 ?9 43>Q4 g>38 1>43 0>Qg n>01 0>00
Bed
Size in mm.
D D D D
max 60 50 10
2.00 1.15 0.98 0.44
2.00 1.14 0.97 0.45
2.00 0.97 0.79 0.36
2.00 1.11 0.97 0.47
2.00 1.03 0.85 0.42
2.00 1.03 0.86 0.40
2.00 1.18 1.00 0.46
2.00 1.11 0.95 0.45
2.00 1.17 1.00 0.46
2.00 1.12 0.96 0.56
C
u
2.61
2.53
2.69
2.36
2.45
2.58
2.56
2.47
2.54
2.00
(continued)
-------
TABLE 2 (continued).
Percent Passing Sieve v
Grain Size # 10 # 40 # 60 # 60 #100 #200
Analysis (2.00)* (0.84) (0.42) (0.25) (0.15) (.074) Pan
Size in mm.
D D D D C
max 60 50 10 u
Test #78
Sample
Test #78
Box
Test #78
Bed
95.80 41.81 6.87 0.69 0.10 0.00 0.00
98.27 38.57 6.02 0.54 0.03 0.00 0.00
96.00 43.53 10.41 2.41 0.23 0.00 0.00
2.00 1.12 0.96 0.45 2.49
2.00 1.15 1.00 0.48 2.40
2.00 1.11 0.95 0.40 2.78
* Sieve sizes are in mm.
M
I
NJ
-------
H
NJ
100
a:
ui
80
1.0 .8 .6 .4 .3 .2
PARTICLE SIZE (mm)
100
or so
Id
60
oc
ui
40
UI
£20
UI
a.
1.0 .8 .6 .4 .3 2
PARTICLE SIZE (mm)
FIGURE 3a. Grain size distribution - Grit #2,
FIGURE 3b. Grain size distribution - Grit #1,
-------
100
N)
U)
o:
UJ
80
u.
_i 60
tr
UJ
40
UJ
O 20
£E
UJ
O.
1.0 fl .6 .4 .3 .2
PARTICLE SIZE (mm)
2.0 1.0 B .6 .4 .3 .2
PARTICLE SIZE (mm)
FIGURE 3c. Grain size distribution - Grit #1/2.
FIGURE 3d. Grain size distribution - Grit #40.
-------
0 [•
2.0 1.0 .8 .6 .4 .3 .2
PARTICLE SIZE (mm)
FIGURE 3e. Grain size distribution - Mixture.
11-24
-------
In Figures 4a-4e plots of bedload sediment movement rate versus
velocity are shown for the different sediment materials. The plotted rate
of movement is the average obtained in the tests. For all grain sizes, the
rate of bedload movement increased with velocity. All the curves have the
same general shape and rates of bedload movement are increasing at an in-
creasing rate.
There is a critical velocity below which bedload movement will not
occur and above which sediment movement will commence. This critical
velocity will be different for different grain sizes (assuming cohesionless
material) and is higher for larger sized sediments (ASCE, 1975). In tests
using the Grit #2, It was found that a velocity of 0.42 m/sec was below the
critical velocity. This result is consistent with Figure 2.46 of the ASCE
Manual (1975) which shows that for a sediment size of 1.65 mm, the critical
velocity should be between 0.21 and 0.46 m/sec.
Figure 5 is a plot of bedload movement rate versus the grain size for
different velocities. The characteristic grain size used for the V$Q
size and bed sediment transport rates were obtained from the results
plotted in Figures 4a-4e. With the exception of Grit #40, which has the
smallest grain size, the bedload movement rate increased with decreasing
grain size, in the range tested.
The amount of total bedload movement obtained using Grit #40 at a
velocity of 0.82 m/sec is less than that obtained from using the next
larger sand. At a velocity of 0.82 m/sec some of the particles are enter-
ing the suspended load. At lower velocities the results are consistent
with the linear relations, since at these lower velocities less particles
are entering the suspended load. In order for a particle to enter the
suspended load, the maximum upward velocity component must be greater than
the settling velocity of the particle (Graf, 1971). From Figure 4.4 of
Graf (1971), the settling velocity of a Grit #40 is found to be
approximately 50 mm/sec. The maximum upward velocity component is computed
from formula 8.7 of Graf (1971):
(Vmax = °'17 (" D)°'46
where (Uy)max * maximum upward velocity component
U * mean water velocity
D = depth of flow
The upward velocity component is computed to be 49.3, 46.6 and 44.9 mm/sec
for the 0.82, 0.60 and 0.43 m/sec velocities respectively. Thus, for
Grit #40 and a velocity of 0.82 m/sec, some of the particles are entering
the suspended load. For larger sized particles, the settling velocity is
much greater than the upward velocity component and the coarser sediments
will remain as part of the bedload.
11-25
-------
1200
o>
1200
UJ
UJ
800
I-
UJ
UJ
o
800
M
I
N5
400
o
UJ
(O
UJ
o
UJ
CO
400
<
o
0.4 O.5 0.6 0.7 0.8 0.9
FLUME VELOCITY (mps)
o
0.4 0.5 0.6 0.7 0.8 0.9
FLUME VELOCITY (mps)
FIGURE 4a. Total sediment movement vs. velocity
- Grit #2.
FIGURE 4b. Total sediment movement vs. velocity
- Grit #1.
-------
1600
E
v.
E
1200
KJ
vj
til
UJ
2 800
h-
UJ
5 400
UJ
CO
O
1200
0.4 O.5 0.6 0.7 0.8 0.9
FLUME VELOCITY (mps)
UJ
UJ
O
UJ
5
UJ
CO
£
O
600
400
0.4 0.5 0.6 0.7 0.8 0.9
FLUME VELOCITY (mps)
FIGURE 4c. Total sediment movement vs. velocity
- Grit #1/2.
FIGURE 4d. Total sediment movement vs. velocity
- Grit #40.
-------
82-11
C5
I H
H- 0>
X M
rr
C CO
i-( (D
m a.
I
n>
3
01
O
n
TOTAL SEDIMENT MOVEMENT (gm/min)
o
o
o
o
8
O
m
P
H ^
^ P
3 oo
•o
(0
"^ p
I
w
H
O
(0
a.
(D
ca
Ul
o
N
m
TOTAL SEDIMENT MOVEMENT (gm/min)
o o
o o
.- o
01
s
O K
CM Ol
o
N
rn o
II
o
O
O
s
o
-------
Similar results were obtained for the tests utilizing the mixture. At
higher velocities the smaller grains became suspended, while at lower
velocities all the material remained as part of the bedload. At lower
velocities the results from the mixture are consistent with the observed
linear relations.
Plots of sampler efficiency versus water velocity are presented in
Figures 6a-6e. In general, sampler efficiency increases with velocity.
The increase in efficiency is large for lower velocities but decreases at
higher velocities. This behavior was observed in most of the tests. For
the Grit #1/2 size, the plot of the data appears to be of a slightly dif-
ferent shape, but follows the same general trend. Considering the highly
variable nature of bedload movement, the results shown in Figures 6a-6e are
quite consistent.
A statistical summary of efficiencies based on test results is
presented in Table 3. The two most important statistical summaries of
experimental data are the mean and the variance (Myers and Walpole, 1972).
The mean is the summation of all measurements divided by the number of
measurements. The variance quantifies the variability of the observed data
about the mean. The positive square root of the variance is defined as the
standard deviation, and the non-dimensional ratio of the standard deviation
to the mean is called the coefficient of variation (Myers and Walpole,
1972). The mean, standard deviation, and coefficient of variation for
sampler efficiency are presented in Table 3. Tests 22, 23, and 24 were not
included in the computations for Table 3 since there was insufficient depth
of water to cover the sampler in these tests. Tests 58 and 60 were also
not included since the velocity was too low to move the Grit #2 particles.
The variability of efficiency as measured by the standard deviation is
about one-half the mean value for most tests. Differences in coefficients
of variation are significant when comparing results for different grain
sizes. Coefficients of variation for the different grain sizes varied from
.341 for Grit #40 to .714 for Grit #1/2. With increasing velocity, more
consistent results were obtained and the variability decreased. For tests
at the lowest velocity, the coefficient of variation is extremely high
(.860), a value close to unity. However, the variability decreased at
higher velocities.
The mean efficiencies obtained for all tests using the sand mixture
were within 1 percent of each other. Also, the difference between the
coefficients of variation was less than 0.05.
To stimulate a typical bedload measurement in which the stream would
be moving at a constant velocity and in which the bed material would be a
graded sand, a series of tests were run using a velocity of 0.82 m/sec with
the sand mixture. The results show a mean efficiency of about 45.5 percent
with a coefficient of variation close to 0.5, This illustrates the kind of
variability to be expected when taking a bedload sample in the field.
11-29
-------
i
OJ
o
50
40
y 30
O
u_
(O
£5 10
0.4 0.5 0.6 0.7 0.8 0.9
FLUME VELOCITY (mps)
70
2 60
>-
O 50
UJ
Ul
30
oc
UJ
-120
<0
10
0.4 0.5 0.6 0.7 0.8 0.9
FLUME VELOCITY (mps)
FIGURE 6a. Velocity vs. efficiency - Grit #2.
FIGURE 6b. Velocity vs. efficiency - Grit #1.
-------
CO
70 r
60
50
UJ
S240
u.
u.
UJ
flC
UJ
30
20
I0
0.4 0.5 0.6 0.7 0.8 0.9
FLUME VELOCITY (mps)
70
60
UJ
O
C 40
u.
UJ
a:30
UJ
a! 20
10
0.4 0.5 0.6 0.7 0.8 0.9
FLUME VELOCITY (mps)
FIGURE 6c. Velocity vs. efficiency - Grit //1/2.
FIGURE 6d. Velocity vs. efficiency - Grit #40.
-------
70
geo
O 50
UJ
040
U_
U.
W 30
cc
3 20
0.
< 10
CO
7
KEY
— — MIXTURE
TOTAL
0.4 0.5 0.6 0.7 OB 0.9
FLUME VELOCITY (mps)
FIGURE 6e. Velocity vs. efficiency - Mixture and total.
11-32
-------
TABLE 3. EFFICIENCY OF THE BOGARDI T-3 SAMPLER - STATISTICAL SUMMARY*
DESCRIPTION
Tests Using
Grit 12
Tests Using
Grit #1
Tests Using
Grit #1/2
Tests Using
Grit #40
Tests Using
Mixture
All Tests
Tests At
Vel. - 0.43mps
Tests At
Vel. - 0.60mps
Tests At
Vel. = 0.82mps
Tests Using
Grit #2 at
Vel. - 0.82mps
Tests Using
Mixture at
Vel. " 0.82mps
NUMBER
OF TESTS
12
14
11
10
18
65
13**
22***
27
9
6
MEAN
EFFICIENCY^
45.24
34.74
31.31
36.28
35.89
36.65
21.78
38.46
44.50
43.15
45.55
STANDARD
DEVIATION
15.82
22.65
22.35
12.38
20.70
19.50
18.74
16.75
18.00
13.09
22.61
COEFFICIENT
OF VARIATION
.350
.652
.714
.341
.577
.532
.860
.436
.404
.303
.496
* Tests 22, 23, 24, 58, 60 not entered into statistics.
** Includes Tests 10, 15, 16.
***Includes Tests 8, 9, 11, 12, 13.
N-l Weighing Used.
11-33
-------
The high variability of the results obtained was expected. Helley and
Smith (1971) state that bedload transport is extremely variable in time and
space. They found a great deal of scatter in their data, with standard
deviations in some cases equal to the mean value. The standard deviations
obtained in this project were somewhat less. Considering the nature of
bedload movement, the variabilities obtained in this testing program appear
very reasonable.
One of the reasons for the high variability in the test results can be
attributed, in part, to the presence of scour and sand waves. Some degree
of scour and sand wave activity was observed in all of the tests. If a
test was concluded at a time when scour was prevailing near the sampler
opening, the corresponding efficiency was usually low. On the other hand,
if a test was concluded at a time when a sand wave had just reached the
sampler, the corresponding efficiency was high. The problems of sand waves
and scour are likely to occur in the field as well as the laboratory,
making the sampler efficiency at least as variable in the field as was
found in the laboratory.
To investigate the possibility of sampler selectivity, several grain
size curves were plotted. Selectivity is defined as the condition when the
sampler collects a limited range of particle sizes while allowing other
particle sizes of the total bedload to pass through the sampler. For
example, if the sampler allows fines to pass through while continuing to
catch coarser sized particles, the sampler is being selective. Complete
results of the sieve analyses for all tests can be found in Table 2. Using
representative grain size curves from the test results (see Figures 7a-7c),
a comparison was made between those of the total bedload transported and
those of the sampler material.
As shown in Figures 7a to 7c, which represent typical results, slight
differences in the grain size distribution curves were obtained for the
sediment collected in the box and the sediment collected in the sampler.
However, total bedload movement is the addition of the box material and
sampler material. Thus, differences in grain size curves between the bed-
load sediment movement and the sediment trapped in the bedload sampler are
even less than shown in Figures 7a to 7c.
Therefore, it Is concluded that the Bogardi Bedload Sampler collects a
representative portion of the total bedload transported. For the range of
sediment sizes and velocities utilized in these experiments, sampler
selectivity is not a problem.
As a result of the flume tests, some general conclusions can be made
relative to the efficiency of the Bogardi T-3 Bedload Sampler. For
velocities at or above 0.92 m/sec an efficiency of approximately 45 percent
was measured. Between 0.61 and 0.92 m/sec an efficiency of approximately
40 percent was obtained. These results are consistent with other tests
utilizing the Karolyi sampler (which is similar in design, but much larger
than the Bogardi T-3), in which Novak (1959) concludes that the average
efficiency is 45 percent. At velocities much below 0.61 m/sec the results
11-34
-------
100
KEY
— SAMPLE
— BOX
— BED
2.0 1.0 .8 .6 .4 .3 .2
PARTICLE SIZE (mm)
FIGURE 7a. Grain size distribution - Test #64.
100
— — SAMPLE
2.0 1.0 .8 .6 .4 .3 .2
PARTICLE SIZE (mm)
FIGURE 7b. Grain size distribution - Test // 68,
-------
100
— — SAMPLE
2.0 1.0 .8 .6 .4 .3 .2
PARTICLE SIZE (mm)
FIGURE 7c. Grain size distribution - Test #71.
11-36
-------
obtained were highly variable and an accurate general conclusion can not be
made. From a practical standpoint this may not be serious, since the
associated amount of bedload movement is small at low velocities.
Variability of sampler efficiency was an expected problem. For tests
conducted at velocities above 0.61 m/sec, coefficients of variation of the
resulting efficiencies were on the order of 0.4 to 0.5. The observed trend
indicates that the variability is reduced at higher velocities, however
this should be verified by further tests.
At significantly higher velocities than those tested, the sampler
efficiency may decrease markedly, as sediment may be able to wash through
the sampler without having a chance to be retained as part of the sample.
Additional tests to verify this possibility should be performed in a flume
where higher velocities can be attained (approximately 1.83 m/sec).
It is recommended that the sampler not be allowed to fill more than
one-third of capacity. When the sampler was filled beyond this amount, a
significant amount of scour was observed near the upstream opening of the
sampler causing a subsequent decrease in efficiency. Novak (1959) also
found that efficiency changes when bedload samplers are filled to more than
25 to 30 percent of their volume.
At the higher test velocities the sampler was not stable and moved
downstream if not restrained. In the laboratory, the sampler was provided
with a handle and was secured by wedging the handle against an overhead
discharge pipe. This solution to the stability problem is not feasible
under field conditions. Field sampling procedures are discussed in
Section 4 and the Appendix.
Some field sampling has been accomplished in Mill Creek and in the
Hudson River with water velocities approaching 2.4 m/sec. Under these
conditions the Bogardi T-3 Sampler had to be provided with 11 to 14 kg of
lead weights which were placed in the collection chamber.
Within the range of sediment sizes tested, the sampler was not ob-
served to be selective in obtaining a bedload sample. The sample obtained
was an accurate representation of the actual bedload movement.
Some general conclusions can also be made regarding the rate of bed-
load movement. An increase in water velocity or decrease in grain size
results in a large increase in the rate of sediment movement* Bedload
movement increases at an increasing rate with velocity while increasing
linearly with decreasing grain size (provided all the sediment remains part
of the bedload).
As noted previously, the results of the laboratory experiments using
the Bogardi Sampler are similar to those of Novak (1959), who used similar
types of bedload samplers. The sampling efficiency does not vary radically
with velocity or particle size. In addition, sampler selectivity does not
appear to be a problem with the Bogardi Sampler. Overall sampler
11-37
-------
efficiency values obtained were also very similar, i.e., about 45 percent
from Novak's experiments and about 40 percent for the Bogardi Sampler.
IIr38
-------
SECTION 4
FIELD TESTING
GENERAL
Although this project was part of the Genesee River Watershed Study,
the field testing program was conducted in a small pilot watershed in
eastern New York State. A number of other IJC-PLUARG studies were also
being conducted in the same watershed. The selection of the watershed
location was based on practical and economical considerations, since many
of the studies required intensive field efforts.
The Mill Creek watershed is located in the southwestern part of
Rensselaer County in eastern New York State (Figure 8), and lies in the
towns of North Greenbush and East Greenbush. It is a few miles south of
Troy, New York, and a few miles east of Albany, New York. It is a rela-
tively small watershed, approximately 26 knr in area above the sampling
station, semi-rural in character, and contains no significant point sources
of pollution. The water quality in Mill Creek is excellent. Additional
information about the watershed can be found in the detailed report by
El-Baroudi (1975).
The sampling station was located at the intersection of a roadway
bridge and Mill Creek. All samples were collected on the downstream side
of the bridge. A sampling platform was attached to the roadway bridge.
This allowed sampling during peak flow periods when personnel could not
enter the stream.
It was imperative that continuous discharge records be available, both
for this project and others being conducted at the sampling station.
United States Geological Survey, Albany Office, Installed and main-
tained a continuous recording gage station at the sampling station. The
stream was gaged at frequent intervals, and an excellent stage-discharge
relationship was available.
Field sampling was conducted over an 18-month period, January 1975
through June 1976. Samples were collected on a routine basis (typically
weekly) and on an event basis. This was necessary since bedload sediment
transport is event dominated, a characteristic common to many hydrologic
measurements. Most of the annual quantity of bedload transport occurs on a
few days of the year (major events). Conversely, routine sampling at
Mill Creek during periods of low flow often yields no measurable bedload
-------
Wothinqton Co
Sorotogo
Co.
MOHAWK
RIVER
MILL.
CREEK
WATERSHED
HUDSON
RIVER
Albany
Co.
FIGURE 8. Mill Creek watershed, Rensselaer County, New York.
11-40
-------
sediment movement.
PROCEDURE
A procedure that can be used for bedload sampling in the field in-
volves a water discharge measurement and the subsequent calculation of the
centroids of areas of equal water discharge. For example, the centroids of
areas representing 20 percent of the total discharge can be computed. As a
limit, the width of any area should be kept under 7.6 m (Helley and Smith,
1971). Bedload measurements should then be taken on the bed at verticals
through these centroids* Repeated measurements should be taken as long as
the flow is steady, because of the variability of bedload movement with
time (Hubbell, 1964). Average values should then be calculated after dis-
carding obviously erroneous measurements. For example, a very low value
might indicate forward tipping of the sampler with subsequent loss of
sample through the front orifice. A very high value might indicate drag-
ging of the sampler forward during placement or removal from the stream, or
it might indicate placement of the sampler mouth in front of a dune where
excessive quantities of sand might be scooped up (Helley and Smith, 1971).
Measurements with a bedload sampler would be made by lowering the
sampler to the stream bed and leaving it in position for a measured period
of time. Sampling time should probably be limited to that which would ap-
proximately fill the sampler to no more than 33 percent of capacity (Graf,
1971).
The rate of bedload movement per unit time per foot of cross-section
can be calculated from the average dry weight of the samples collected, the
time of sampling as measured by a stopwatch, and the sampler dimensions.
Total bedload transport for each area of equal water discharge is calcu-
lated by multiplying the corresponding rate of transport by the width of
the area. A factor considering efficiency should also be included. Total
stream bedload transport is the summation of the bedload discharges for
each area.
For sampling in Mill Creek, the above procedure was modified. Because
sampling took place immediately on the downstream side of a bridge with
vertical concrete abutments, the flow was nearly uniform under the bridge.
Therefore, only one sample was taken at approximately the middle of the
bridge. The rate of bedload transport was assumed to be constant for the
width of the stream flowing under the bridge (approximately 4.4 m). This
assumption was verified by sampling at various points across the stream
bed. Thus, the total amount of bedload transport for Mill Creek could be
calculated from the results of one bedload sample* In most instances only
one sample was obtained on each sampling date due to time and economic
reasons.
The easiest and most reliable method of sampling is to have personnel
enter the stream directly (Fig. 9). When this was impossible during periods
of high flow, longer rods were attached to the sampler sides and the sampler
was lowered to the stream bed from the sampling platform (Fig. 10). During
11-41
-------
FIGURE 9. Method of bedload sampling for low flow periods.
FIGURE 10. Method of bedload sampling for high flow periods,
11-42
-------
very high flows, a system of ropes was also utilized to provide stability
during the lowering and raising processes. This requires at least two
sampling personnel.
If a platform or low bridge is not available or the water depth is too
great, then the sampler must be lowered into the watercourse using ropes.
As mentioned later, this method was utilized on the bedload sampling pro-
gram in the Hudson River.
It was mentioned in the section on Laboratory Experiments that the
sampler must be completely immersed in water when sampling.
Once on the stream bed, the sampler was left in position for a period
of time measured by a stop watch.
Upon removing the sampler from the stream bed, the front of the
sampler was lifted up higher than the rear to prevent the loss of any
sample. The rear trap door of the sampler was also caulked prior to
sampling to prevent the loss of any sediment through the joints between the
door and the body of the sampler. The sample plus the water retained was
removed through the rear door and placed In a watertight plastic container
for shipment back to the laboratory. A velocity measurement was also made
at the point of sampling by means of a Price-type current meter. This
measurement was made at the six-tenths depth, according to the standard
USGS stream gaging procedure (Davis, Foote and Kelly, 1966). The tempera-
ture of the water sediment mixture was also determined, plus the approx-
imate gage height from the staff gage at the sampling station. Any other
pertinent information regarding the sampling was also recorded. Additional
sampling details are given in the Appendix.
In addition to the bedload samples obtained by using the Bogardi
Sampler, at frequent intervals bed material samples were collected by
scooping the top layer of the stream bed. These samples were also oven
dried and sieved.
The Bogardi T-3 Bedload Sampler is a small, lightweight sampler ori-
ginally developed for use in small streams. In spring 1976, bedload
sampling was conducted in the upper Hudson River between Fort Edward,
New York, and Troy, New York. Various highway bridges were utilized as
sampling stations since personnel could not enter the river due to water
depth. Because of the height of the bridges above the water, the sampler
was lowered into the water by using ropes. In most cases, weights were
added to the sampler to increase its stability. On one sampling day, a
major event occurred on the upper Hudson River. A 100-year flood occurred
and water velocities above 2.1 m/s were encountered. The sampler performed
quite well and satisfactory bedload samples were obtained in all cases.
A major consideration with a bedload sampler is the maximum water
velocity in which the sampler can be used. The stability of the Bogardi
Sampler can be greatly Increased by adding weights. Since the weights are
inside the sampler, there is no increased resistance to the flow of water.
11-43
-------
Because the weights are placed in the lower collecting box of the sampler
where the velocities are low, there is no measurable effect on sampling
efficiency when weights are used (maximum of 13.6 kg lead plates were
used). At Mill Creek, the highest velocity encountered was about 2.4 m/s.
It is felt that samples can be obtained in water velocities as high as 3.0
m/s and probably even at higher velocities.
ANALYSIS OF FIELD DATA
The general methods applied in computing sediment discharges depend on
the kind, accuracy and frequency of the available samples, and on on the
expected use of the computed discharges. The general procedure differs for
computations that are based on frequent sediment discharge measurements and
for those based on occasional sediment discharge measurements (Colby,
1963).
Sediment samples are generally obtained infrequently unless a reason-
ably accurate time distribution of sediment discharge and concentration is
desired. With occasional measurements of sediment discharge over the usual
range of flow in the stream, the sediment discharges or concentrations can
be plotted against the streamflow or gage height. A curve drawn through
these points defines an average sediment discharge or concentration re-
lation for the cross section. Usually, individual points scatter widely
from such a curve (Colby, 1963). For example, see Figures 11 and 12.
For a particular sediment measuring station, curves based on sediment
discharge measurements and concurrent streamflows can be used to compute an
approximate sediment discharge at any time when the streamflow is known.
Although sediment discharges computed from such curves may be reasonably
correct on the average, at individual times they may be greatly in error.
Sediment discharges for a day, storm period, month, year or longer
period can be computed directly by using a flow duration curve for the
period, and by using a sediment discharge curve that is assumed to be rep-
resentative for the period. The range of flow is divided into several sub-
ranges, and the amount of time that the flow was within each subrange is
determined. The sediment discharge per unit time is obtained for each sub-
range, and this is multiplied by the amount of time that the flow was in
that particular subrange. The total sediment discharge is the sum of these
products (Colby, 1963).
The above general procedure is usually modified to fit a specific
situation.
The field sampling results are shown in Figure 11 which is the bedload
sediment-discharge relation for Mill Creek. Total bedload transport is
plotted versus the stream discharge, assuming a constant efficiency of 40
percent. The best fit solid line was obtained by using a least squares
criteria. The bedload data is listed in Table 4, and values of total
bedload have been computed assuming sampler efficiencies of 40 percent.
The spread of data points is considerable in Figure 11, i.e. for a given
11-44
-------
*-
Ui
I05
S I02
CD
10'
EFFICIENCY* 40%
10
-I
10'
DISCHARGE (m3/sec)
^ if)3
o
<
o
o
id
CD
-J I0«
CLEARWATER RIVER
After Emmeft
IOZ I03 10
DISCHARGE (m3/sec)
FIGURE 11. Bedload sediment-discharge relation
for Mill Creek.
FIGURE 12. Bedload sediment-discharge relation
for Clearwater River.
-------
TABLE 4. BEDLOAD DATA.
•e-
Date
1975
Jan.
23
Feb.
24
25
26
27
Mar.
6
20
21
Apr.
3
10
25
Jun.
12
17
26
Station
11
11
13
13
12
12
12
13
13
11
12
10.5
10.5
10.5
Water
Temp.
(°C)
0
0
1
1
2
1
4
4
-
8
11
19
20
24
Gape
Ileipht
(cm)
-
128
113
101
95
87
109
101
122
88
82
84
81
74
Discharge
r-3/s
—
7.67
3.25
1.46
0.896
0.420
2.63
1.46
5.35
0.613
0.241
0.365
0.218
-
Velocity
m/s
0.205
1.95
1.19
0.528
0.534
0.248
0.976
0.573
2.59
0.323
0.206
0.261
0.20
0.062
Time
(min)
110
-
15
15
30
135
15
15
—
60
45
20
40
40
Sample
Weight
(R)
5.0
-
2470.3
115.6
122.0
4.1
3141.1
68.4
—
4.0
-
0.5
1.1
-
Kg/m/day
*
1.17
-
4235
198.2
104.6
0.787
5384
117.3
—
1.71
-
0.64
0.71
-
Total
Bedload
(kp/dayj*
5.19
-
18732
877
463
3.48
23819
519
—
7.56
-
2.83
3.15
-
(continued)
-------
TABLE 4 (continued).
Date
1975
Jul.
2
9
15
Aug.
7
8
15
21
Sep.
3
9
12
13
19
23
24
Station
10.5
10.5
10.5
10.5
10.5
10.5
10.5
10.5
10.5
10.5
10.5
10.5
10.5
10.5
Water
Temp.
(°0
23
23
23
17
17
18
-
16
15
16
15
16
14
14
Gage
Height
(cm)
73
88
80
87
117
81
76
76
76
76
75
76
84
81
Discharge
m3/s
0.613
0.162
0.524
4.59
0.195
-
0.076
-
0.076
-
-
0.348
0.229
Velocity
m/s
0.050
0.393
0.183
0.381
1.373
0.200
-
0.084
0.082
0.084
0.078
0.084
0.323
0.239
Time
(min)
60
30
40
60
2
40
-
45
90
45
-
-
90
60
Sample Kg/in/day
Weight *
(g)
0.2 0.09
5.1 4.37
1.3 0.83
16.0 6.86
1078.3 13864
-
-
0.1 .056
-
0.1 .056
-
_ _
4.0 1.14
- -
Total
Bedload
(kg/day)*
0.38
19.33
3.70
30.33
61324
-
-
0.25
-
0.25
-
-
5.06
-
(continued)
-------
TABLE 4 (continued)
i
*-
o>
Date
1975
Sep.
25
26
Oct.
3
10
11
15
18
19
20
21
Nov.
7
13
14
20
26
Station
10.5
10.5
10.5
10.5
10.5
10.5
10.0
10.0
10.0
10.0
10.0
10.0
10.0
10
10
Water
Temp.
<°C)
13
14
11
12
12
14
11
11
11
12
13
9
4
5
5
Gape
Height
(cm)
95
95
80
77
92
86
125
128
114
101
82
95
99
81
88
Discharge
m3/s
1.31
1.20
0.173
0.106
0.969
0.479
5.85
6.41
4.09
1.93
_
-
-
-
0.658
Velocity
m/s
0.659
0.580
0.188
0.121
0.448
0.375
2.00
2.21
1.13
0.671
0.217
0.519
0.647
0.241
0.421
Time
(min)
30
60
^
-
90
30
2
1
2
5
5
_
-
-
30
30
Fample
Weight
(R)
54.60
26.65
.
-
14.66
2.40
123.20
69.56
55.56
1067.90
14.60
_
-
-
0.50
10.05
Kg/in/day
*
46.81
11.42
-
4.19
2.06
1584
1789
714.36
5492
75.08
^
-
-
0.43
8.61
Total
Bedload
(kg/day)*
207.01
50.52
—
18.52
9.10
7006
7912
3160
24293
332.13
L
-
-
1.89
38.10
(continued)
-------
TABLE 4 (continued).
Date
1975
Dec.
1
10
17
1976
Jan.
14
27
M 27
*- 28
VO
28
29
Feb.
17
18
19
27
Station
10
10
10
10
10
10
10
10
10
10
10
10
10
Water
Temp.
(°C)
7
3
2
0
0
0
0
0
0
0
0
0
0
Gage
Keight
(cm)
96
95
92
107
146
146
148
148
123
113
102
113
94
Discharge
m3/s
1.37
1.27
0.91
2.88
0.49
9.49
9.77
9.77
5.45
3.84
2.16
3.84
1.12
Velocity
m/s
0.448
0.464
0.415
_
1.66
1.66
1.66
1.66
0.671
1.10
0.778
1.10
0.659
Time
(min)
30
30
30
30
2
2
2
5
5
5
5
5
5
Sample
Weight
(R)
19.70
98.35
-
7.30
40.60
62.80
22.90
11.00
59.30
35.2
-
-
2.6
Kg/m/day
*
16.88
84.30
-
6.25
522.01
807.44
294.43
56.56
304.97
181.04
-
-
13.37
Total
Bed load
(kg/day)*
74.69
372.89
-
27.67
2309
3572
1302
250.23
1349
800.75
-
-
59.15
(continued)
-------
TABLE 4 (continued) .
Date
1976
Mar.
12
26
30
Apr.
1
9
22
26
May
7
14
18
19
20
Jun.
1
11
21
Station
10
10
10
10
10
10
10
10
10
10
10
10
10
10
10
Water Gage
Temp. Height
(°C) (cm)
86
86
88
116
86
81
100
88
88
89
109
116
86
79
79
Discharge
m3/s
0.501
0.479
0.613
4.34
0.479
0.195
1.81
0.591
0.636
0.734
3.23
4.34
0.482
0.151
0.151
Velocity
m/s
-
.308
.305
1.16
0.296
0.149
0.814
0.314
0.323
0.363
1.11
1.42
0.218
0.198
0.198
Time
(min)
-
-
30
2
30
30
5
30
30
15
4
2
30
15
—
Sample Kg/m/day
Weight *
(e)
-
-
2.8 2.41
135.42 1741
0.75 0.64
-
-
1.4 1.20
_
2.5 4.26
1.1 7.07
212.0 2726
1.9 1.63
_
- -
Total
Bedload
(kg/day)*
-
-
10.61
7702
2.85
-
-
5.31
-
18.85
31.27
12057
7.20
-
-
* Sampler Efficiency Assumed «= 40%
-------
discharge the total bedload can vary considerably. This normal variability
of bedload transport is of much greater significance than that due to + 5
percent differences in efficiency. As previously noted, the efficiency
does not vary radically with velocity and discharge. However, the effi-
ciency does increase somewhat with discharge, approximately 35 percent at
low flows to 45 percent at high flows. Values of total bedload were
computed using efficiencies appropriate to the discharges, and the line of
best fit was obtained. Similarly, values were computed assuming constant
efficiencies of 35 percent, 40 percent (Fig. 11) and 45 percent. The four
lines virtually coincide when plotted on Figure 11. Therefore, from a
practical standpoint, the assumption of a constant efficiency value of 40
percent for the Bogardi T-3 Bedload Sampler is justified for field
sampling.
It appears necessary to establish the range of sampler efficiency when
calibrating a bedload sampler, perhaps to the nearest 5 or 10 percent.
However, extensive efforts to improve the accuracy of the laboratory ef-
ficiency values appear unwarranted.
Total bedload transport can also be plotted against velocity
(Figure 13). The plot looks very similar to Figure 11, a fact which is not
surprising considering the fairly uniform flow conditions at the Mill Creek
sampling station, i.e., discharge and velocity correlate very well. In
general this will not be true, especially for water courses with complex
cross-sections. Correlations of bedload transport and discharge can differ
radically from correlations of bedload transport and velocity.
As noted previously, the variability of the field sampling results
shown in Figure 11 is typical of bedload sampling programs and is not unique
to this project. Bedload transport in natural systems is highly variable
(Graf, 1966). Values of total bedload transport versus discharge derived
from a field bedload sampling program by Emmett (1976) are shown in
Figure 12. The similarity between Figure 11 and Figure 12 is evident.
Similar relationships are found for suspended load transport (for example,
Mansue and Anderson, 1974).
It was imperative that continuous discharge records be available, both
for this project and others being conducted at the sampling station. The
United States Geological Survey, Albany Office, installed and maintained a
continuous recording gage station at the sampling station. The stream was
gaged at frequent intervals, and an excellent stage-discharge relationship
(rating curve or table) was available (Table 5-a, 5-b).
In order to estimate the amount of bedload transport in Mill Creek,
the number of occurrences of mean gage height were tallied from the USGS
gage station records and corresponding water discharges were found using
rating tables (Table 5-a, 5-b and Table 6-a and 6-b). For each discharge,
the corresponding bedload transport was obtained from the bedload sediment-
discharge relation curve (Fig. 11). The total bedload movement in Mill
Creek was then calculated.
11-51
-------
3K
•o
V.
E
LU
X
o
CO
Q 10
O
<
o
_J
o
bJ
OQ
10°
EFFICIENCY»40%
10
-I
10
VELOCITY (m/sec)
FIGURE 13. Bedload sediment-velocity relation for Mill Creek.
11-52
-------
TABLE 5a. RATING TABLE FOR MILL CREEK (11/7/74 - 4/2/75).
Gage Height
(cm)
79.30
82.35
85.40
88.45
91.50
94.55
97.60
100.65
103.70
106.75
109.80
112.85
115.90
118.95
122.00
125.05
128.10
131.15
134.20
Discharge
(m3/s)
0.11
0.20
0.34
0.50
0.67
0.90
1.18
1.46
1.82
2.27
2.72
3.25
3.86
4.56
5.35
6.13
6.92
7.70
8.57
Difference
(o3/s)
0.09
0.14
0.16
0.17
0.23
0.28
0.28
0.36
0.45
0.45
0.53
0.61
0.70
0.79
0.78
0.79
0.78
0.87
11-53
-------
TABLE 5b. RATING TABLE FOR MILL CREEK (from 4/2/75)
Gage Height
(cm)
76.25
79.30
82.35
85.40
88.45
91.50
94.55
97.60
100.65
103.70
106.75
109.80
112.85
115.90
118.95
122.00
125.05
128.10
131.15
134.20
137.25
140.30
143.35
146.40
149.45
152.50
155.55
Discharge
(ir3/s)
C.08
0.15
0.26
0.43
0.66
0.91
1.20
1.54
1.93
2.38
2.83
3.33
3.84
4.34
4.84
5.35
5.85
6.36
6.86
7.36
7.87
8.37
8.93
9.49
10.05
10.60
11.20
Difference
(ro3/s)
0.07
0.11
0.17
0.23
0.25
0.29
0.34
0.39
0.45
0.45
0.50
0.51
0.50
0.50
0.51
0.50
0.51
0.50
0.50
0.51
0.50
0.56
0.56
0.56
0.55
0.60
11-54
-------
TABLE 6a. DISCHARGE OCCURRENCE (1/23 - 4/2/75).
Mean Gage
Height (cm)
82.66- 83.88
84.18- 85.40
85.71- 86.93
87.23- 88.45
88.76- 89.98
90.28- 91.50
91.81- 93.03
93.33- 94.55
94.86- 96.08
96.38- 97.60
97.91- 99.13
99.43-100.65
100.96-102.18
102.48-103.70
104.01-105.23
105.53-106.75
107.06-108.28
108.58-109.80
110.10-111.33
111.63-112.85
113.16-114.38
114.68-115.90
116.21-117.43
117.73-118.95
Average
(CO)
83.27
84.79
86.32
87.84
89.37
90.89
92.42
93.94
95.47
96.99
98.52
100.04
101.57
103.09
104.62
106.14
107.67
109.19
110.72
112.24
113.77
115.29
116.82
118.34
Discharge
(o3/s)
0.24
0.31
0.39
0.47
0.55
0.64
0.74
0.85
0.98
1.12
1.26
1.40
1.57
1.75
1.95
2.18
2.40
, 2.63
2.88
3.14
3.43
3.74
4.07
4.42
Bedload
(kg/day)
3.4
5.4
10.0
14.0
20.0
27.0
36.0
52.0
70.0
91.0
127.0
150.0
193.0
240.0
317.0
385.0
498.0
634.0
680.0
861.0
997.0
1404.0
1540.0
1812.0
Number
of
Occurrences
4
8
13
11
10
8
3
3
3
1
0
1
1
1
0
1
0
0
0
0
0
0
1
1
Total
Bedload
Total Eedload
(kg)
14.0
44.0
129.0
155.0
204.0
217.0
108.0
156.0
211.0
91.0
0
149.0
193.0
240.0
0
385.0
0
0
0
0
0
0
1540.0
1812.0
« 5648.0 kg
11-55
-------
TABLE 6b. DISCHARGE OCCURRENCE (3/3/75 - 6/2/76; missing 8/22-8/27/75).
Mean Gage
Height (cm)
71.98- 73.20
73.51- 74.73
75.03- 76.25
76.56- 77.78
78.08- 79.30
79.61- 80.83
81.13- 82.35
82.66- 83.88
84.18- 85.40
85.71- 86.93
87.23- 88.45
88.76- 89/98
90.28- 91.50
91.81- 93.03
93.33- 94.55
94.86- 96.08
96.38- 97.60
97.91- 99.13
99.43-100.65
100.96-102.18
102.48-103.70
104.01-105.23
105.53-106.75
107.06-108.28
108.58-109.80
110.11-107.97
111.63-112.85
113.16-114.38
Average
(cm)
72.59
74.12
75.64
77.17
78.69
80.22
81.74
83.27
84.79
86.32
87.84
89.37
90.89
92.42
93.94
95.47
96.99
98.52
100.04
101.57
103.09
104.62
106.14
107.67
109.19
110.72
112.24
113.77
Discharge Bedload
(m3/s) (kg/day)
_
-
-
0.10
0.14
0.19
0.24
0.31
0.40
0.50
0.61
0.73
0.86
1.00
1.15
1.31
1.47
1.66
1.85
2.07
2.29
2.51
2.74
2.98
3.23
3.48
3.74
3.99
_
-
-
0.50
1.02
1.90
3.44
5.44
10.4
15.9
25.4
36.5
52.1
72.5
92.9
125
163
217
272
353
417
544
657
702
770
1020
1270
1520
(continued)
11-56
Number
of
Occurrences
4
25
35
22
23
25
35
34
48
55
32
34
14
19
7
12
4
7
3
7
3
1
0
2
1
2
3
1
Total Bedload
(kg)
-
-
-
11.0
23.4
47.6
120
185
500
872
812
1240
729
1380
650
1490
652
1520
815
2470
1250
543
0
1400
770
2040
3800
1520
-------
TABLE 6b (continued).
Mean Gage
Height (cm)
114. 68-115. 90
116. 21-117. A3
117.73-118.95
119.26-120.48
120.78-122.00
122.31-123.53
123.83-125.05
125.36-126.58
126.88-128.10
128.41-129.63
129.93-131.15
131.46-132.68
132.98-134.20
134.51-135.73
136.03-137.25
137.56-138.78
139.08-140.30
140.61-141.83
142.13-143.35
143.66-144.88
145.18-146.40
146.71-147.93
148.23-149.45
149.76-150.98
151.28-152.50
152.81-154.03
154.33-155.55
Average
(cm)
115.29
116.82
118.34
119.87
121.39
122.92
124.44
125.97
127.49
129.02
130.54
132.07
133.59
135.12
136.64
138.17
139.69
141.22
142.74
144.27
145.79
147.32
148.84
150.37
151.89
153.42
154.94
Discharge
(m3/s)
4.24
4.49
4.74
5.00
5.25
5.50
5.75
6.00
6.26
6.51
6.76
7.01
7.26
7.52
7.77
8.02
8.27
8.54
8.82
9.10
9.38
9.66
9.94
10.22
10.50
10.78
11.06
Bedload
(kg/day)
1630
1830
2130
2310
2630
2900
319D
358 D
3990
4100
4510
4980
5440
5780
6120
6340
6800
7700
8150
8610
9290
9970
10900
11300
12200
12700
13400
Number
of
Occurrences
0
1
1
2
1
1
1
0
1
0
0
0
0
0
0
0
0
0
0
0
1
0
0
0
1
1
1
Total
Bedload
Total Bedload
(kg)
0
1830
2130
4620
2630
2900
3190
0
3990
0
0
0
0
0
0
0
0
0
0
0
9290
0
0
0
12200
12700
13400
- 93720 kf
Therefore, total bedload in Mill
5648.0 + 93720 «= 99368 kg; i.e.,
Creek from Jan. 23, 1975 to Jun.2, 1976 is
approximately 100,000 kg.
11-57
-------
Total bedload transport in Mill Creek during the period of January 23,
1975 to June 2, 1976 (missing data 8/22/75 to 8/27/75) (491 days) was esti-
mated to be about 100,000 kg, an average daily amount of about 240 kg/day.
However, as noted previously, bedload is event dominated, and approximately
65 percent of the total bedload was transported during the ten days when
flows were highest* Almost 50 percent of the total bedload transport
occurred during the four days of highest flows.
COMPARISON WITH BEDLOAD FORMULAS
If bedload samples are not available, bedload formulas are commonly
used to calculate the quantity of bedload sediment transport. Many bedload
formulas have been developed and an extensive amount of literature is
available. The reader is referred elsewhere (Graf 1966, ASCE 1975, Yalin
1972) for additional information.
Historically, bedload formulas have been developed from closely con-
trolled laboratory conditions, and are empirical relations. Strictly
speaking, they should only be applied within the experimental boundaries.
The developers of bedload formulas should point this out quite clearly.
The range of variables encountered under natural conditions is almost uni-
versally greater than that of the laboratory. In addition, the numbers of
variables encountered in the field is much greater than in the laboratory.
Therefore, it is not surprising that no universally applicable bedload
formula has been developed (Graf, 1966), and that no formula has been
developed which clearly predicts with a sure degree of accuracy, the bed-
load transport (Falcone, 1974). Yalin (1972) states: "It is regrettable
that the authors of (sediment) transport formulae do not always indicate
the validity regions of their expressions, and therefore the transport
formulae are often used for cases where they should not be used. As a
consequence, the transport formulae are often criticized on the grounds
that they give the transport rates that are 'five or ten times different'
from the measured rates. In this context it should be remembered that if
the limits are not indicated, it does not mean that they do not exist. All
it means is that it is up to the user of the formula to discover the
limitations not mentioned by the author, and to use the formula
accordingly."
Notwithstanding the above, the engineer faced with the task of comput-
ing bedload yield must select a formula, and in practice the formulas are
applied well outside of the experimental boundaries.
There are essentially three different approaches to the bedload com-
putation problem. They are as follows (Graf, 1971):
1. The duBoys-type equations, considering a shear stress relation-
ship.
2. The Schoklitsch-type equations, considering a discharge relation-
ship.
11-58
-------
3. The Einstein-type equations, based on statistical considerations
of the lift forces.
The duBoys-type equations are based on the oversimplified model of
bedload moving in sliding layers. The Schoklitsch-type equations consider
discharge and the critical discharge at which sediment begins to move.
Einstein's equation represents a departure from the other types of equa-
tions. This is largely because Einstein's physical model makes ample use
of the advancements in fluid mechanics (Graf, 1971).
For bedload computations, Shulits and Hill recommended the following
formulas from the 14 formulas examined (Falcone 1974, Tywoniuk 1972):
1. Skoklitsch, 1934: A discharge formula
2. Meyer-Peter & Muller: A tractive stress formula
3. Straub: A tractive stress formula.
In order to compute bedload transport using bedload transport for-
mulas, the required basic data are:
(1) stream width, average depth
(2) mean velocity or water discharge
(3) energy gradient
(4) particle-size analysis of bed material
(5) water temperature.
We can define some of this data fairly accurately while some is not easy to
obtain.
It is of interest to compare the Mill Creek field results with those
predicted by bedload formulas. Following Shulits and Hill's recommenda-
tion, the Schoklitsch, 1934 Formula and Meyer-Peter Equation are used,
while the Straub Equation is eliminated because it yields bedload values
too large in comparison with the other formulas and field measurements in
Mill Creek. The Einstein bedload formula is also used because of its
classical importance. Bedload formulas are often modified after original
development, but may still carry the original name. Therefore, the reader
is referred to the sources of the formulas used herein: 1. Einstein for-
mula as presented in Graf (1971), 2. Meyer-Peter formula (Gray, 1973), and
3. Schoklitsch formula (Falcone, 1974).
The Schoklitsch (1934) formula is engineering units is given the fol-
lowing form by Shulits and Hill and presented in Falcone (1974):
g - 25 (s'-'/D'-
ir-59
-------
where:
gg; bedload discharge transported (Ib/s/ft)
S ; water surface slope
Dg; effective grain diameter (ft)
q ; water discharge (cfs/ft)
; critical discharge (cfs/ft).
Schoklitsch gives the critical discharge in engineering unit as:
4/3
qoj = 0.0638 Dfi/S
where:
q0l; critical discharge (cfs/ft)
D£ ; mean grain diameter for given fraction (ft)
S ; water surface slope.
For a non-uniform mixture, the grain size distribution curve is divided
into a number of fractions, and for each fraction the mean diameter and
percentage of the total mixture is obtained using the above equation. Bed-
load transport is then computed for each fraction and summed to compute
total bedload using the relationship:
g - ag + bg ,+ eg + ...
s sa sb sc
where:
g , g ,,...; bedload discharge for any given set of size
fractions
a,b, ... ; percentage of total mixture which each mean
grain diameter represents.
Since the grain size curve can be arbitrarily divided Into parts, this
procedure can yield a wide range of results.
For coarse material and flat slopes this equation should be used with
care (Graf, 1971).
The limitations of this formula are (Falcone, 1974):
water surface slope 0.003 - 0.024
specific gravity of grains 2.53 - 2.69
mean grain diameter (mm) 0.305 - 7.01.
11-60
-------
Gray (1973) presents the Meyer-Peter bedload formula in a simplified
form. In engineering units it is:
2/3 2/3
q Z/J = 39.25 q ' S - 9.95 D
where:
qs = sediment discharge in Ib/sec/ft width of channel,
q * water discharge in cfs/ft width,
S * slope of the channel in ft/ft,
D = mean size of sediment in feet as defined below
D =
where pj is percent by weight of bed sediment with a mean size d^,
where d^ is in feet.
D is determined from a size analysis of a representative sample of bed
material. The size distribution is divided into a convenient number of
size fractions, and the mean size and weight percentage of each is deter-
mined.
The Meyer-Peter equation was formulated after a long series of experi-
mental measurements and was carried out using a wide range of character-
istics (Falcone, 1974):
water surface slope 0.004 - 0.2
specific gravity of grain 1.25 - 4.00
mean grain diameter (mm) 0.400 - 30.0.
The Einstein bedload formula can be written in the following form
(Graf, 1971):
* = fct
., Ps -p
SRh
Ys Pg -P gd
1_
11-61
-------
where:
g; bedload discharge
pg; density of sediment
p ; density of water
d ; average diameter
S ; slope of channel
g ; gravitational constant
Yg; specific weight of sediment
R.*; hydraulic radius, due to particles
* ; intensity of bedload transport
t ; intensity of shear on particles.
The Einstein bedload formula is dimensionally homogeneous and has been
checked out for a very wide range of characteristics. Graphs are available
for practical calculations (Graf, 1971). For further discussion of this
formula see Einstein (1950), Shen (1972), Graf (1971) or Yalin (1972).
The bedload formulas have a D dimension term, representing some size
characteristics of the sediment. If no bedload samples are available, bed
material samples are obtained by scooping, dredging, or using piston
sampling devices (ASCE, 1975). Generally, the mean grain size will be much
greater for the bed material than for the sampled bedload, since the mean
sediment size of the bedload in transport will depend on the discharge and
velocity of the stream. In other words, bedload is the material that is in
motion along the stream bed while bed material is the material making up
the stream bed that is available for transport. Therefore, particle-size
analyses of bed material and bedload at a given point and time are not
usually the same.
\
In order to compute bedload transport, it is necessary to obtain
representative samples of bed material. There are four major classes of
bed material samplers available in use to date:
(1) drag bucket
(2) grab bucket
(3) vertical type or piston sampler
(A) rotating bucket scoop type.
11-62
-------
In shallow streams, satisfactory samples can often be obtained using
shovels, scoops, or even by handpicking. However, there are certain phases
of the procedure that are not yet adequately defined, for example (Carlson
and Miller, 1956):
(1) How many samples are necessary to define the mean?
(2) Is there a change in the mean bed size with discharge?
(3) How is a representative sample obtained in a cobble-bedded stream
(for example, Mill Creek)?
(4) How are representative samples obtained under deep flowing water?
(5) How deep should a sample be taken to give particle size informa-
tion applicable to the various bedload formulas?
Obviously, the normal variability of bed material samples is as high
or higher than that for suspended and bedload sediment samples.
Total bedload transport was computed for Mill Creek using the
appropriate D size obtained from bed material samples and by using the D
sizes obtained from bedload samples. The bed material derived values
appear unsatisfactory, largely due to the large D sizes. Use of the
Einstein and Meyer-Peter formulas predict essentially no bedload transport.
The Schoklitsch formula can yield a range of sediment-discharge relations
depending on the method of analyzing the grain size distribution curves
(Falcone, 1974). The field results fall within this range; however, in
general, one would have no idea where the actual answer lies.
Sediment-discharge relations predicted by using the appropriate D
sizes obtained from bedload samples are presented with the field data (data
points omitted for clarity) in Figure 14. The values computed by the use
of the bedload formulas agree quite well with each other and yield satis-
factory agreement with the field data.
The Schoklitsch formula was used to compute bedload yields using the
bedload sampling data of Emmett (1976), shown in Figure 13. As for Mill
Creek, the agreement is satisfactory and the computed sediment-discharge
relation is essentially parallel to and shifted to the left of the field
relation.
It appears that bedload formulas can yield satisfactory results if the
correct sediment size characteristic is used. Of course, the 0 size of the
bedload in transport will not be known without bedload sampling, in which
case the use of bedload formulas becomes academic.
The previous conclusions appear valid for Mill Creek and the Clear-
water River, where a wide range of sediment sizes exist. If a watercourse
has a narrow range of sediment sizes, the D sizes of the bedload sediment
transport and the bed material sediment may be similar and, thus, quite
different results may be obtained.
11-63
-------
10'
5 *o4
o»
Q I03
O
_J
O
A'SCHOKLITSCH
B'MEYER-PETER
C'EINSTEIN
D'FIELD DATA
IO'1 10° 10
DISCHARGE (m3/sec)
FIGURE 14. Bedload sediment-discharge relations for Mill Creek:
field data and bedload formulas.
11-64
-------
REFERENCES
American Society of Civil Engineers. 1969. Sediment Measurement Tech-
niques: A. Fluvial Sediment. Journal of the Hydraulic Division,
ASCE, Vol. 95, No. HY5, Proc. Paper 6756. September, pp. 1477-1514.
American Society of Civil Engineers. 1975. Sedimentation Engineering.
Manual and Report on Engineering Practice, No. 54. New York.
American Society for Testing Materials. 1956. Standard Method of Test for
Sieve Analysis of Fine and Coarse Aggregates. Selected ASTM Engineer-
ing Materials Standards. ASTM, Philadelphia, Pennsylvania.
Carlson, E.J. and Miller, C.R. 1956. Research Needs in Sediment Hydrau-
lics. Journal of Hydraulics Division, Proceedings of the American
Society of Civil Engineers, Vol. 82, No. HY2. Proc. Paper 953.
April, pp. 1-33.
Chien, N. 1954. The Present Status of Research on Sediment Transport.
Transactions, ASCE, Vol. 121. Paper No. 2824. p. 833.
Colby. 1963. Fluvial Sediments - A Summary of Source, Transportation,
Deposition, and Measurement of Sediment Discharge. United States Geo-
logical Survey Water Supply Paper 1181-A, Washington, D.C.
Colby. 1964. Discharge of Sands and Mean-Velocity Relationships in Sand
Bed Streams. United States Geological Survey Water Supply Paper 462-A,
Washington, D.C.
Davis, Foote and Kelly. 1966. Surveying, Theory and Practice, Fifth
Edition, McGraw-Hill Book Company, New York.
Emmett, W.W. 1976. Bedload Transport in Two Large Gravel-Bed Rivers, Idaho
and Washington. Proc. of the Third Federal Inter-Agency Sedimentation
Conference, Denver, pp. 4-101 - 4-114.
El-Baroudi, H. 1975. Inventory of Forms of Nutrients Stored in a Water-
shed. Report prepared for New York State Department of Environmental
Conservation and IJC. Rensselaer Polytechnic Institute, Troy, New
York, August.
Einstein, H.A. 1948. Determination of Rates of Bedload Measurement. Pro-
ceedings Federal Interagency Sedimentation Conference, United States
Department of Interior.
11^65
-------
Einstein, H.A. 1950. The Bed-Load Function for Sediment Transportation in
Open Channel Flows. United States Department of Agriculture, Soil Con-
servation Service Tech. Bull. 1026.
Falcone, F.E. 1974. Bedload Discharge for Irregular Cross-Sections. Pre-
print 2385, ASCE Annual and National Environmental Engineering Conven-
tion. October, pp. 1- 30.
Graf, W.H. 1971. Hydraulics of Sediment Transport. McGraw-Hill Book
Company, New York.
Gray, Donald. 1973. Handbook on the Principles of Hydrology. Water Infor-
mation Center, Canada.
Grover, N.C. and Harrington, A.W. 1966. Stream Flow-Measurements, Records
and Their Uses. Second Edition, General Publishing Company, Ltd.,
Canada.
Guy, H.P., Simons, D.B., and Richardson, E.V. 1961. Summary of Alluvial
Channel Data from Flume Experiments. 1956-61, United States Geological
Survey Prof. Paper 462-1. Washington, D.C.
Guy, H.P. and Norman, V.W. 1970. Field Methods for Measurement of Fluvial
Sediment. Techniques of Water-Resources Investigations of the United
States Geological Survey, Book 3, Chapter C2. Geological Survey.
Washington, D.C.
Helley, E.J. and Smith, W. 1971. Development and Calibration of a Pres-
sure-Difference Bedload Sampler. United States Geological Survey
Open-File Report. Menlo Park, California.
Hubbell. 1964. Apparatus and Techniques for Measuring Bedload. United
States Geological Survey Water Supply Paper 1748. Washington, D.C.
Lambe, T.W. and Whitman, R.V. 1969. Soil Mechanics. John Wiley and Sons,
Inc., New York.
Leliavsky, S. 1966. An Introduction to Fluvial Hydraulics. Third Edition.
General Publishing Company, Ltd., Canada.
Mansue, L.J. and Anderson, P.W. 1974. Effects of Land Use and Retention
Practices on Sediment Yields in the Stony Brook Basin, New Jersey.
United States Geological Survey Water-Supply Paper 1798-L. United
States Government Printing Office, Washington, D.C.
Novak, P. 1959. Vyzkum Funkce A Ucinnosti Pristroju Na Mereni Splavenin
(Study of the Performance and Efficiency of Bedload Meters). No. 99.
Vyzkumny Ustav Vodohospodarsky. Prace A Studie, Prague.
Shen, H.W. (editor and publisher). 1972. Sedimentation. Symposium to
Honor Professor H.A. Einstein, Fort Collins, Colorado.
11-66
-------
Simons, D.B., Richardson, E.V., and Haushild, W.L. 1963. Some Effects of
Fine Sediment on Flow Phenomena. United States Geological Survey Water
Supply Paper 1498-G. Washington, D.C.
Tywoniuk, W. 1972. Sediment Discharge Computation Procedures. Journal of
the Hydraulics Division. Proceedings of the American Society of Civil
Engineers, Vol. 98, No. HY3. March, pp. 521-540.
Waslenchuk, D.G. 1976. New Diver-Operated Bedload Sampler. Journal of the
Hydraulics Division. Procedings of ASCE, Vol. 102, No. HY6, paper
12179. June, pp. 747-757.
Williams, G.P. 1967. Flume Experiments on the Transport of a Coarse Sand.
United States Geological Survey Professional Paper 562-B. Washington,
D.C.
Williams, G.P. 1970. Flume Width and Water Depth Effects in Sediment
Transport Experiments. United States Geological Survey Professional
Paper 562-H. Washington, D.C..
Yalin, M.S. 1972. Mechanics of Sediment Transport. Pergamon Press,
Oxford.
11-67
-------
APPENDIX
FIELD METHODS FOR FLUVIAL BEDLOAD MEASUREMENT
USING THE BOGARDI T-3 BEDLOAD SAMPLER
INTRODUCTION
Knowledge of the movement of fluvial sediment is important to those
involved directly or indirectly in the development and management of water
and land resources. Total sediment load in a stream can be considered to
consist of two types of sediment, the bedload and the suspended. The usual
suspended sediment samplers cannot be used in the lowest 0.09 to 0.12 m
depth of the stream bed. The sediment in transport in this region consists
of the bedload, and some suspended load. If the bedload is measured, the
suspended load in this lower region is still not measured, and is referred
to as the unmeasured load. Normally, the unmeasured load is negligible and
is neglected; however, it can be estimated using empirical or other quanti-
tative methods.
The usual sediment study consists of the measurement of suspended load
and the computation of bedload transport using one or more of several bed-
load formulas. The actual bedload transport may or may not be a sizable
part of the total sediment load. Typical values of bedload as a percentage
of total sediment transport range from 10 to 50 percent, but values outside
of that range are not unusual.
No matter how precise the theoretical prediction of sedimentation pro-
cesses becomes, it is inevitable that man's activities will continue to
cause changes in the many variables affecting sediment transport. Also,
none of the computational methods for determining bedload or total load
that have been developed is universally acceptable for all sediment sizes,
bed configurations and flow regimes (Graf, 1971). Therefore, there is an
increasing need for the direct measurement of sediment movement in streams
including the bedload.
The required knowledge of sediment transport for a large variety of
applications makes it necessary to work in a wide range of hydrological
environments. The complex phenomena of fluvial sedimentation makes the
required measurements relatively expensive in comparison with other kinds
of hydrological data.
The purpose of this appendix is to relate some efficient techniques in
using the Bogardi T-3 Bedload Sampler.
11-68
-------
THE BOGARDI SAMPLER
The Bogardi T-3 Bedload Sampler is extremely lightweight and small,
making it ideal for transportation to and from different sampling sites.
In most instances one person can successfully sample. The sampler front
opening is 10 cm high. Therefore, the flow through the sampler is very
close to that of the unsampled zone in suspended sediment sampling. The
sampler was found to be excellent for use in small streams, even though at
high velocities it is necessary to hold the sampler in place or to add
weights due to the sampler's light weight. In larger rivers it was also
found to perform satisfactorily, as will be discussed later.
GENERAL SAMPLING TECHNIQUES
For information on general sediment sampling techniques, for example,
site selection, the frequency of sampling, location and number of verticals
to be sampled, water discharge measurements, computation of sediment dis-
charge, cold weather sampling suggestions, recording of information, etc.,
the reader is referred to the literature (e.g., Guy, Harold P. and Norman,
Vernon W., 1973).
For this project, samples were taken approximately once a week and
after events such as rainfalls and thaws. It is to be emphasized that it
is extremely important to sample after events since most of the bedload
transport occurs at these times. Other data measured were water tempera-
ture, water stage, water velocity and location of the sample in the cross
section. In practice, the actual data needed will vary with the specific
application.
All equipment should be checked to assure that it is in good working
order before use in the field. To avoid unnecessary delays, a check should
also be made to assure that none of the equipment is left behind. Some
spare equipment such as an extra battery for a current meter (if needed)
should be carried.
For sampling on bridges with traffic, bright clothing should be worn,
and an extra person may be needed to help direct traffic.
Often the sampling personnel will find it necessary to be in preca-
rious positions above a swiftly flowing watercourse. Emphasis must be
placed on safety considerations.
SAMPLING IN SMALL STREAMS WITH LOW VELOCITIES
The most precise and preferred method of sampling is for the sampling
personnel to enter the stream and directly place the sampler on the stream
bed. Naturally, this method can only be utilized in shallow streams flow-
ing at low velocities. Even in shallow streams, it is extremely helpful to
attach short rods to the sides of the sampler to enable placement of the
sampler on the bottom. This keeps the arms of the sampling personnel dry,
a consideration during cold weather.
11-69
-------
The sampling location should be one where the stream bed is fairly
uniform and level and where there is an absence of boulders which would
interfere with the water flow and sampler placement. Care should be taken
during placement to prevent scooping up of any bed material and to insure
that the front of the sampler is sitting on the bottom so that the bedload
does not roll underneath the sampler entrance. If the velocity is rela-
tively high, the sampler may have to be held in place, or a small flat
weight added to the bottom of the sampler catch area to prevent the sampler
from tipping over.
Measurement of the bedload is made by leaving the sampler in place for
a period of time measured by a stopwatch. The sampling time varies depend-
ing on the transport rate. Experience at low velocities indicates that
times from 15 to 60 minutes are adequate. Sampling time should be limited
to that which would fill the sampler to about 33 percent of capacity
(Graf).
Upon removal of the sampler, care should be taken to prevent scooping
up of bed material, especially if sand waves or dunes have formed in front
of the sampler. The front of the sampler should be lifted higher than the
rear to prevent the loss of any sample through the front opening. The
sample plus water retained should be removed through the rear door and
placed in a water-tight container for shipment back to the laboratory.
Note that the rear door should be caulked or sealed before sampling to
prevent any sample loss through the joints between the door and body of the
sampler." Finally, any other measurements such as water temperature, water
stage, water velocity, etc., should be made.
SAMPLING IN SMALL STREAMS WITH HIGH VELOCITIES
When wading is impossible due to high water velocities, another pro-
cedure can be used if a bridge is available. For the sampling done at Mill
Creek in conjunction with this project, 6-foot flat steel bars were bolted
to the sides of the sampler. The tops of the bars were bent into a "u"
shape to facilitate handling. This procedure allows the bars to be easily
removed and transported. The length of the bars depends on the height of
the bridge and, if necessary, they can be made of several attachable
sections. However, this procedure will not work for very high bridges, as
the bars will be too flexible.
Another procedure which has been used by others is the attachment of a
pipe to the top of the sampler, instead of bars. In this project the use
of a pipe to lower and raise the sampler was found to be unsatisfactory.
Due to the flexibility of the top plate of the sampler, buckling of the
plate occurred even at moderate velocities. If a thicker plate is used,
the bolts attaching the plate to the plexiglass sides of the sampler can
themselves bend and shear or split the plexiglass.
The sampling procedure consists of rapidly pushing the sampler down
through the water to the stream bed by means of the bars. The sampler is
held in place by pushing down on the bars while time is measured by a
11-70
-------
stopwatch. For very high water velocities, sampling times on the order of
one minute may be adequate.
For very high velocities, flat lead weights (up to 13.6 kg) can be
placed in the bottom of the sampler for increased stability. Also, a rope
or ropes can be attached to the front of the sampler and held taut by one
person while the other lowers the sampler.
For high water velocities, a certain amount of Judgment must be
exercised relative to the addition of weights and the use of the ropes.
The ultimate measures that can be utilized (that is, for highest veloci-
ties) consist of the use of the maximum amount of weights and the use of
two ropes to stabilize the sampler while it is lowered to the stream bed.
The maximum velocity encountered at Mill Creek was 2.4 mps and it was not
necessary to resort to these ultimate measures. That is why it is felt
samples can be obtained at velocities higher than 2.4 mps.
When the water velocity is high, the stream bed usually cannot be
seen, due to the high sediment load. It is advisable to locate suitable
sampling points when the water level is low and the stream bed can be seen.
If this is not possible, then the sampling must be .accomplished by feel.
Usually it is possible to determine if the sampler is sitting on a boulder;
and if so, a new sampling location should be found. It is also recommended
to push slightly forward on the bars to insure that the front of the
sampler is sitting on the bottom.
Recommendations for placing, removing and sealing the sampler, and for
securing the sample are identical with that of the last section.
SAMPLING IN DEEP RIVERS AND FROM HIGH BRIDGES
In this situation, a system of ropes should be utilized in lowering
the Bogardi Sampler. Two ropes are used which are at about a 30 to 45
degree angle from the vertical, forming a triangle, with the sampler as the
point. A central vertical rope can also be used. The central rope sup-
ports the weight of the sampler and the two side ropes provide stability.
Of course, two or three people are required for sampling in this case.
In this situation, lead weights placed in the bottom catch area of the
sampler are usually required (except at very low velocities) to insure that
the sampler is stable on the stream bed. Lead weights of up to 13.6 kg
have been used.
The sampler Is lowered to the surface of the water, aligned, and
rapidly dropped to the bottom of the river. The best results were achieved
when no weight was supported by the middle rope during the lowering pro-
cess, since tension on the side ropes and water pressure on the sampler
stabilizing fin tends to keep the sampler aligned properly. However, the
central rope is quite useful when raising the sampler. Once the sampler is
on the bottom, tension can be applied to the two sides ropes to assure that
the sampler is aligned upstream. This must be done carefully to avoid the
11-71
-------
possibility of scooping up the bed material. Removal of the sampler should
be accomplished as rapidly as possible, with the front end kept higher than
the rear. Similarly, sealing the sampler, removal of the sample, and
measurements of other quantities are the same as in the previous sections.
When sampling in large rivers, sampling technique is important. The
processes of lowering, raising and aligning the sampler requires experience
and teamwork. It is suggested that a sampling team Inexperienced in the
use of the Bogardi Sampler practice prior to securing actual samples.
Practice is best accomplished in clear water or shallow water, where the
bottom is visible, so that the final position of the sampler can be
verified.
Large rivers are usually very turbid; and since the stream bed cannot
be seen, some judgement is required. If the character of the bottom is un-
known, it is possible for the sampler to land on a boulder or other
obstruction. If no sample is obtained, it can normally be assumed that
something of that nature has occurred. As both sampling experience and
familiarity with the character of the watercourse is obtained, such occur-
rences become obvious.
11-72
-------
REPORT III
NITROGEN AND PHOSPHORUS LOSSES
IN DRAINAGE WATERS FROM ORGANIC SOILS
by
John M. Duxbury
and
John H. Peverly
Department of Agronomy
Cornell University
Prepared for
New York State Department of Environmental Conservation
and
United States Environmental Protection Agency
-------
-------
REPORT III
CONTENTS
Page
Abstract i
Figures ii
Tables iii
Acknowledgements v
1. Introduction 1
2. Methods 4
Study Sites 4
Oak Orchard Creek - Elba Muck 4
Little Conesus Creek - South Lama Muck 6
Analytical Procedures 8
Water 8
Plants 8
Water flow and sanple collection 10
Nitrogen Mineralization and Denitrification in Soil Columns.. 11
Collection of soil columns 11
Nitrogen mineralization studies 12
Denitrification studies 12
Aquatic Plant Nutrient Remobilization 14
3. Results and Discussion 17
Nutrient Output of Mucklands 17
Oak Orchard Creek 23
Little Conesus Creek 34
Metals losses 39
Aquatic plant nutrient dynamics 45
References 55
-------
ABSTRACT
The contrUbution by diffuse sources to water pollution is thought to be
an appreciable fraction of the total. However, the contribution by differ-
ent land types and uses is poorly understood.
Organic or peat soils, which can be very intensively and profitably cul-
tivated after adequate drainage, appear to be an important diffuse source
for N, P and other elements in drainage water. Drainage water from two
organic soil deposits in northwestern New York State was monitored for flow
rates and nutrient concentrations over the period March 1975 to June 1976.
Continuous flow records with composite water sampling for concentration
determinations were used. At other sites, partial flow measurements and
water sampling were performed on a twice-weekly schedule. Rain events were
intensively sampled.
The total annual losses from the soils ranged from 0.9 to 30.7 kg/ha for
ortho-P, 41. to 92.6 kg/ha for NO--N, and <1. to 1.9 kg/ha NH4~N. The P
losses seem to be directly correlated with the depth of organic material.
Concentrations in the drainage water increased as flows increased, so
that by far the greatest losses were during spring and fall events.
Because organic soils constantly undergo decomposition with the simulta-
neous release of N and P, it is not known what the previous fertilizer
practices contribute to these losses. It is clear that organic soils
contribute to N and P in land runoff at a much greater proportion than
indicated by the total area of such soils. Appreciable quantities of metals
may also be lost frcm certain organic soils.
The downstream fate of the elements released include sediment tie up,
biological incorporation and passage into receiving waters. Most N and P
seems to be transported downstream, with substantial quantities incorporated
into aquatic plants during sunmer and fall.
Ill-i
-------
FIGURES
Number Page
1 Oak Orchard Swamp 5
2 Soil map of the cultivated muckland near Elba, New York 7
3 Hie South Lima muck near Avon, Livingston County, New York,
drained by Little Conesus Creek 9
4 Apparatus for collecting nitrous oxide from Histosol columns...... 13
5 Diagram of experimental setup for measuring phosphorus movements
between water and sediments 15
6 Pumping time, NO^-N and P concentrations for water flowing from
the muck area served by Site GR-A 24
7 Pumping time, N03-N and P concentrations for water flowing from
the muck area served by Site GR-B 25
8 MRP vs. discharge for Oak Orchard Creek at Oak Orchard Road (1975-
1976) 31
9 MRP vs. discharge for Oak Orchard Creek at Harrison Road (1975-
1976) 32
10 NO^-N vs. discharge for Oak Orchard Creek at Oak Orchard Road
T1975-1976) 32
11 NO,-N vs. discharge for Oak Orchard Creek at Harrison Road (1975-
1976) 33
12 Changes in MRP in the water of control and planted jars over the
experimental period 48
13 Growth of plants in culture jars over the experimental period,
given in both dry weight and total stem length 49
14 Percent and total phosphorus in plant tissue growing in culture
jars 51
Ill-ii
-------
TABLES
Number
1 Soils in Oak Orchard Swamp 6
2 Characteristics of the Histosol columns 11
3 Nutrient output of cultivated mucklands 17
4 Mineralization of nitrogen from Histosol columns 19
5 Denitrification of nitrate-amended Histosol columns 21
6 Nutrient losses from 2000 ha of the Elba muckland in a 24-hour
high flow period (2/19-2/20/76) 22
7 Concentration of molybdate reactive phosphorus in Oak Orchard
Creek 26
8 Concentration of nitrate (NO.,) in Oak Orchard Creek 29
9 Load data for Little Conesus Creek three kilometers downstream
from the muck at the site designated 'F1 near East Avon 35
10 Total annual loadings in Little Conesus Creek near East Avon
CF1), March 6, 1975 to March 5, 1976, calculated by three
different methods 36
11 Flow (floating object method), concentration and load data for
single days upstream of the muck on the main stem of Little
Conesus Creek designated 'U' 38
12 Flow (floating object estimate), concentration and load data for
single days at the east feeder stream to Little Conesus Creek,
entering midway through the muck, designated 'B1 38
13 Metal concentrations in Little Conesus Creek one kilometer below
the muck, designated '0', near South Lima 40
14 Approximate metal loads in Little Conesus Creek, one kilometer
below the muck, designated '0', near South Lima 41
15 Summary of elemental losses from the basin of Little Conesus
Creek 42
-------
Number Page
16 Estimated standing crop for submerged aquatic plants, Little
Conesus Creek. 43
17 Elemental content of submerged aquatic plants, Little Conesus
Creek (dry weight basis) 44
18 Sediment characteristics determined at the beginning of the
experiment 45
19 Soluble phosphorus in sediment or interstial water 45
20 Summary of inorganic phosphorus fractionation data for
sediments 47
21 Summary of data for plants harvested May 15, one month after
planting 50
22 Summary of data for plants harvested May 28, two months after
placement in demineralized water 52
23 Summary of data for plants harvested June 22 after growth in
phosphorus-enriched water 53
24 Summary of entire experiment, representing the means of four
jars including all plant material, not just the samples 54
Ill-iv
-------
ACKNOWLEDGEMENTS
This study was supported in part by contract No. C-81701 fron the
New York State Department of Environmental Conservation. The expert tech-
nical assistance of R. Clayton, E. Goyette and K. Jones is gratefully
acknowledged.
III-v
-------
-------
SECTION I
There are approximately 2.8 million hectares of organic soils
(mucklands) within the Great Lakes basin (262,000 ha in New York). Only a
small percentage of these soils are drained for agricultural production;
there are about 12,000 ha of cultivated mucklands in New York.
The predominant use of the cultivated mucklands is for production of
vegetable crops, which generally receive heavy applications of fertilizer
nitrogen and phosphorus (commonly 100 kg/ha of both N and P_05 in New York).
The soil itself, which consists of the partially decomposed remains of plant
material, is also a source of nutrients as it is constantly being subjected
to mineralization processes which cause release of N and P into soil water.
Most cultivated organic soils contain 2-3% N and 0.1-0.2% P and, based on
long-term subsidence rates, an estimated 1000 kg/ha of N and 60 kg/ha of P
are mineralized annually in cultivated organic soils in the Great Lakes
region (Duxbury, unpublished data). Thus, it is apparent that a high
potential for leaching of nutrients from these soils exists. Also, it can
be noted that additions of fertilizer N are unlikely to influence N losses
because of the large amount of N released through mineralization of soil
organic N. On the other hand, fertilizer P could be an important source of
P to drainage water, depending upon the rate of application.
Organic soils that are not artificially drained can act as either sinks
or sources of nutrients depending upon their state of development and
natural wetness. Indeed, the process of peat formation is one of conserva-
tion of nutrients, and it is only when the agent of conservation, usually
water, is removed that the process is reversed.
In a detailed study of losses of soluble N and P from four shallow
organic soil sites in Ontario, Miller (1974) found that between 91-196
kg/ha/yr of N and 14-26 kg/ha/yr of P were lost in drainage water in the
years of 1972-73. These values contrasted with average values of 25
kg/ha/yr of N and 0.3 kg/ha/yr of P in drainage water from eight mineral
soil sites sampled in the same experiment. Of the annual N losses from the
organic soils, NO--N comprised 80-91%, organic-N was 8-18%, and NH.-N was
<2% of the total. The distribution of phosphorus varied markedly Between
the two years of study, 1972 and 1973; in 1972 ortho-P accounted for 32-48%
of the total, but 63-85% in 1973. Hortenstine and Forbes (1971) found
concentrations of N and P in soil water from cultivated muckland adjacent to
Lake Apopka, Florida, similar to those obtained by Miller, but they did not
measure actual losses.
III-l
-------
Others (Erickson and Ellis, 1971; Nicholls and MacCriitmon, 1974) have
measured considerably smaller losses of both N and P in leachate from
cultivated mucklands. losses of N and P from the Michigan State University
experimental muck farm in 1969 were 18.7 and 1.45 kg/ha, respectively
(Erickson and Ellis, 1971) . The N losses were similar from two farms on
mineral soils but P losses were 10X higher. Only 4.1 kg/ha of N and 1.6
kg/ha of P were contained in drainage water pumped from the Holland Marsh,
Ontario, in the spring of 1971 (Nicholls and MacCrinmon, 1974) . This was
the only period during the year when water was pumped from the soils, but
the authors noted that the 1971 growing season "was characterized by an
unusually dry spring".
All the published data for N losses in organic soil drainage water is
consistent in that losses are well below the potential of plus 1000
kg/ha/yr. This is undoubtedly due to the activity of denitrifying microor-
ganisms. It is clear that the denitrification process is not limited, as it
often is in mineral soils, by lack of a suitable carbon source (energy
source) in the zone where denitrification can take place.
Losses of soluble phosphorus from cultivated mucklands vary consid-
erably, but in all cases appear to be from 10-1000X those from cultivated
mineral soils (cf Hanway and Lablen, 1974; Baker et al. , 1975; Hergert,
1975) . Hie reasons for the variability in P output from the mucklands are
not clear, but may be related to differences in fertilizer practices and the
ability of the different soils to fix phosphorus. The negatively charged
organic colloids have little ability to fix PO.-P (Pox and Kamprath, 1971)
and any PO. fixing capacity of organic soils is usually attributed to
mineral components associated with the soil (Kaila and Missila, 1956; Kaila,
1959; Larsen, et al. , 1959; Okruszko et al., 1962; MacLean et al . , 1967;
Ismirah and Keeney, 1973) . The quantity and composition of mineral material
in organic soils varies widely from soil to soil and it is likely that many
organic soils have a finite capacity for PO.-P fixation. Thus, the length
of time since drainage of an organic soil deposit and the history of fertil-
izer additions could have an important influence on P losses.
In addition to measuring N and P losses directly from organic soils, it
is also important to know how far the nutrients are transported downs team,
how much biological and chemical immobilization may occur, and what factors
effect remobilization.
simplest indication of downstream transport is the difference in
stream load between two sampling sites a reasonable distance apart, down-
stream of the muck. Assuming there to be no appreciable feeder stream
inputs between the two sites, this measurement would give an indication of
inmobilization by stream sediments, aquatic organisms, and perhaps other
losses between the two sites.
Chemical and physical losses of ortho-phosphate to the sediments would
include adsorption to colloid surfaces and possible precipitation with iron,
aluminum or calcium. This P may be remobilized when sediments are resus-
pended, as during high flow (Johnson et al. , 1976) . Similar types of
nitrogen loss may be by ammonium adsorption on colloid surfaces, or by
III-2
-------
volatilization. Nitrate may be denitrifed and lost to the atmosphere as
nitrogen gas or oxides of nitrogen (Bouldin et al., 1974). The reactions of
dissolved organic N and P are not understood.
In addition to sediment losses, N and P may be lost to organisms in the
stream. The major biological component of the stream used in this study in
terms of biomass and, therefore, nutrient content, was submerged, rooted,
vascular plants. These plants can accumulate N and P to levels greater than
that needed for growth as indicated by both field and laboratory studies
(Gerloff and Krombholtz, 1966; and Lathwell et al., 1973). This "luxury
consumption" is likely to occur in the drainage water which probably con-
tains elevated nutrient levels. Therefore, just as these plants are signif-
icant sinks for N and P during the growth season, they could also be signif-
icant sources as they senesce during the fall when flow volume is low.
Booted aquatic plants are also able to absorb N, P, and Ca from sediment
water and transport them to the plant tops (Demarte and Kartman, 1974).
Submerged aquatic plant tissue is also leaky. It is entirely possible for
rooted plants to remobilize appreciable nutrients by absorption from sedi-
ments, transport to the plant tops, and release back into the water while
they are growing. Laboratory and greenhouse experiments were initiated to
see if this does happen and to what extent.
III-3
-------
SECTION 2
METHODS
STUDY SITES
Oak Orchard Creek - Elba Muck
Oak Orchard Swamp lies in an east-vest direction along the Genesee-
Orleans County line in northwestern New York (Figure 1). The swamp is
dominated by organic soils and mucky mineral soils. The eastern section of
the swamp has been drained and approximately 3000 ha of muckland developed
from woody parent material are in vegetable crop production. Two wildlife
refuges, the Oak Orchard Wildlife Management Area and the Iroquois National
Wildlife Refuge, occupy a major portion of the swamp west of the
agricultural area. The swamp is drained by Oak Orchard Creek, which rises
southwest of Elba, New York, and discharges into Lake Ontario at Point
Breeze. The creek channel has been diverted and improved through the
muckland to aid drainage. The drainage basin is about 32,000 ha at the
point where the creek leaves the Iroquois refuge (Harrison Rd).
Three tile-drained sites with defined drainage areas were chosen for
study on the cultivated muckland (Figure 1). At each site, tile drainage
water emptied into a cistern and was then pumped into Oak Orchard Creek or a
main drainage ditch. There was no surface runoff from any of the sites.
The farmers had some control over the depth to water table in that they
could raise or lower the pressure switches which controlled the operation of
the pumps. The fanners changed the depth to water table as they wished
throughout the study period. Two of the sites (48 ha and 32 ha) were
located on the Grinnell farm and the third (4 ha) on the Karas farm. The
48-ha site (GR-A) was on deep muck (>107 cm) and the tile lines were in
organic soil material at a depth of 100 cm. The 32-ha site (GR-B) ranged in
depth from 45 cm to >107 cm with tile lines predominantly in the mineral
layer underlying the organic soil. The mineral substratum under the western
section of the site was sand, with a thin layer (1-2 cm) of silty clay at
the muck-miJieral interface. Marl (CaCOj underlay the center section, and a
calcareous silty clay was under the eastern section adjacent to the cistern.
The 4-ha site was about 90 cm deep and the tile lines were in an underlying
layer. The locations of the sites are shown in Figure 1.
The areas of organic soils and mucky mineral soils within Oak Orchard
swamp are given in Table 1. In brief, there are 3000 ha of cultivated
organic soils, 2500 ha of uncultivated organic soils, and 2500 ha of mucky
mineral soils. About two-fifths of the cultivated muckland is mapped as
being deep and one-fifth of the uncultivated soils are deep. Figure 2 is a
III-4
-------
OAK ORCHARD SWAMP
N
01
CULTIVATED MUCKLANC
RT63
Shollow Mucklond (Ms)
Deep Muck land
Mucky Mineral Soils
Sampling Points
—---Ook Orchord Creek
Feeder Canal
0123
^^^^^^MMB^^MMOT^
km
FIGURE 1. Map of Oak Orchard Swamp.
-------
soils map of the cultivated muckland and the major blocks of uncultivated
organic soils are shown in Figure 1.
TABLE 1. SOILS IN OAK ORCHARD SWAMP.
Soil Type* Cultivated Uncultivated
(ha) (ha)
Organic - deep (>107 cm)
Organic - shallow (300107 on)
Organic - Edwards
Canandaigua mucky silt loam
Alden mucky silt loam
Fonda mucky silt loam
Lamson muck very fine sand loam
1290
1670
150
100
15
100
0
525
1970
65
1220
140
860
90
*Taken from soil survey maps of Genesee County (1969) and Orleans County
(unpublished).
Little Conesus Creek - South Lima Muck
This muck area is 35 km south of Rochester in the Town of Lima,
Livingston County, and is about 230 ha in area. It developed from woody and
reed vegetation and is used primarily for potato production. The surface pH
is about 6.0. It is drained by Little Conesus Creek, which at the muck
outlet has a basin area of about 2040 ha. Drainage from the muck is by open
ditch and tile and no water monitoring of any kind was done in the muck
itself. The basin is generally made up of scattered, drumlin-like features
superimposed upon broad, rolling hills characteristic of the Erie-Ontario
Plains in this part of New York State. The area is underlain by calcareous
till.
III-6
-------
ELBA MUCK
Mi
Ms
Mr
Creek lines
ORLEANS
-County Lint
6ENESEE
• Approx. 1 km
FIGURE 2. Soil map of the cultivated muckland near Elba, New York. Mi is mineral soil, and Ms
and Mr are shallow and deep muck, respectively.
-------
Little Conesus Creek rises in the Town of Livonia to the south and flows
north through the muck in an improved channel, the bottom of which lies in
calcareous material (pH>7). North of the muck outlet it curves to the west
and joins Conesus Creek at Avon. Conesus Creek flows into the Genesee
River, which runs north and empties into Lake Ontario at Rochester.
Two stations on Little Conesus were chosen at which flow measurements
and water samples were taken to determine the contribution muck soils made
to nutrient load. The first was just downstream of the muck, designated as
the outlet or 'O1 site. The second site was about 2 km further downstream
and was designated 'F1. There was no major nutrient or water inputs between
the sites.
Two minor stations were also established, one being above the muck on
the main channel CU") and one on a feeder stream ("B"). These locations
are shown in Figure 3.
ANALYTICAL PROCEDURES
Water
Water samples were centrifuged at 35,000g for 30 minutes before being
analyzed for NH.-N and soluble P in an autoanalyzer, essentially as outlined
in USEPA Methods (1974). Nitrogen as NH. was measured colorimetrically as
the indophenol blue. For total soluble P, samples were digested with
persulfate according to Menzel and Corwin (1965), and the digests reacted
with molybdate to form a molybdophosphoric acid, which was subsequently
reduced with ascorbic acid. Blue intensity was measured on an autoanalyzer.
Ortho-P, or molybdate reactive P (MRP), was determined on the same sample
with no digestion.
Total N03 plus N02-N was determined according to USEPA Methods (1974).
Nitrate was first reduced to N02 with copper in alkaline hydrazine sulfate,
and then the entire NO2 content was measured cholorimetrically as an azodye,
after diazotization of sulfanilimide and coupling with N-(l-Naphtyl)
ethylenediamine dihydrochloride.
Ca, Mg, Na, Fe, Cu, Zn, Mn, Pb, and Cd concentrations were determined by
flame AA spectometry, using lanthanum to minimize anion interference. If
needed, the samples were concentrated 20 times by evaporation. This may
nave produced low readings for Cu, Mn, Pb and Cd as an acid insoluble,
amorphous precipitate formed in these samples. Potassium was determined by
flame emission spectrophotometry.
Plants
Standing crops of submerged rooted aquatic plants plus filamentous algae
were estimated by removing all the live plant material from randomly
selected meter square plots in Little Conesus Creek. The plots selected
ranged 5 to 50 m up and down the creek bed from the 'O1 and 'F1 sites.
Duplicate samples above and below the sites were taken monthly from December
III-8
-------
SOUTH LIMA MUCK
Qwinerol Soils
[3Shallow Muck
£3 Deep Muck
—Creek Bsd
Little Conesus Creek
—^
Livonia
FIGURE 3. The South Lima muck near Avon, New York in Livingston County,
drained by Little Conesus Creek.
III-9
-------
1974 to November 1975. Samples were sorted as to species, dried at 85°C,
and dry weights were recorded.
•total N in the plant tissue was determined by the Kjeldahl method. For
other elements, plant samples were ashed overnight in quartz crucibles at
500°C. The ash was treated with nitric acid, dried, heated to 400-°C,
cooled, and taken to dryness with HCl. The residue was dissolved in normal
HC1 containing 0.5% lanthanum. Phosphorus was determined by the
molybdovanadophosphoric acid method of Kitson and Melton (1944). Cd and Pb
were determined by atomic absorption spectrophotonetry.
Water Flow Measurement and Sample Collection
Elba Muckland: Flow data was calculated from pumping time multiplied by
pumping rate. Pump operation was monitored continuously and pump discharge
was determined by measuring the rate of draw down in a cistern. Water
sample collection was automatic and was based on pumping time. Thus, a
250-mL sample was collected for every 1 or 2 h of pump operation. Samples
from the GR-A site were discretely collected using a Sigma sampler, while
samples were composited at the other two sites. Samples were picked up
every 24 h, then stored at 2-4°C.
Oak Orchard Creek: A continuous recording gauge was located at Oak
Orchard Road (station 2) and partial record stations were located at Watson
Road (station 1) and at Harrison Road (station 11). The partial record
gauges were installed in the fall of 1975. Support for all these gauges was
provided by the Soil Conservation Service. Oak Orchard Creek was sampled
monthly in 1974, bi-weekly in 1975, and on an event basis in the spring of
1976. The sampling locations are shown in Figure 1. Samples were placed in
ice after collection and analyzed within 24 h, except for the period
February 17 to February 29, 1976, when they were analyzed within one week of
collection.
South Lima: Flow measurements at the stations '0' and "F" were made by
wire weight and staff gauges, respectively. The equipment was installed and
the rating curves developed by USGS personnel in Ithaca, New York. Samples
and gauge height readings were routinely recorded twice weekly. Event
periods were monitored more often, and during ice cover only water samples
were taken. At stations 'U1 and 'B1, samples were collected occasionally
during events and flows were estimated by simply measuring the
rate of movement of a barely floating object. The geometry of these sites
was defined by culverts. Water samples were collected frctn March 1975 to
June 1976. The samples were not filtered before storage as they contained
very little particulate mineral material. They were placed on ice immedi-
ately after collection and then stored at 2-4°C until analysis within one
week.
IIJ-10
-------
NITROGEN KQ3ERALIZATION AND DENITRIFICATION IN SOIL COLUMNS
Collection of Soil Columns
Columns (1.2 m) of natural Histosol (muck) profiles were obtained from
the Grinnell farm in August, 1975. PVC waste pipe (1.37 m x 7.6 qn i.d. x
6.4 mm wall) with a sharpened tip was used to obtain the soils column. The
purpose of a lucite insert was to aid in collection of the soil columns by
reducing the diameter of the soil core to slightly less than that of the PVC
tube. Non-compacted soil columns were obtained when the PVC tubes were
driven into the soil using a sledge hammer. The tubes were removed by using
a chain and an automobile bumper jack. Attempts to obtain soil columns
using a hydraulic ram to drive the PVC tubes resulted in collection of about
25 cm of compacted soil in the bottom of the tube. Apparently the vibration
from the sledge hammer enabled the soil to remain in place as the tube was
driven down. After collection, the soil columns were sealed with rubber
stoppers and stored at 2°C until used. Some characteristics of four of the
columns are given in Table 2. The dead volume of a column was determined by
adding 10 mL of solution containing 51.5 mg of NO--N (as Ca(N03)2) to the
column, followed by 50 mL of IN CaCl- and then leaching with distilled
water. The leachate was collected in fractions ranging fron 100-800 mL.
The Cl content of the fractions was determined by a standard Mohr titration
using 0.05 N AgNO_and 5% K2Cr20_ as indicator. The NO--N and NH.-N content
of the fractions were determined by the steam distillation method of Brenner
(1965).
TABLE 2. CHARACTERISTICS OF THE HISTOSOL COLUMNS.
Column No.
15
16
17
18
Flow Rate
PH
(15 cm H_O head) Surface
(raLTh)
1920
5105
280
2010
5.85
5.95
6.05
5.90
pH
Weight of
122 on Drained Column
(kg)
6.15
6.15
5.85
6.15
4.42
4.31
4.54
4.71
III-ll
-------
Nitrogen Mineralization Studies
Four columns, designated 15, 16, 17 and 18 were used. Glass wool plugs
were placed at the top and bottom of each column and a rubber stopper,
fitted with a tygon drainage tube was inserted in the bottom. The columns
were leached free of NH.-N and NCL-N using distilled water and then were
allowed to drain before the drainage tube was closed. The columns" were
incubated uncovered at 23°C for 14 days, then 50 mL of a IN CaCl2 solution
was added to each column followed by leaching with 6.5 L of distilled water.
The leachate was collected in several fractions, which were stored at 2°C
until analysis for Cl, NO--N and NH.-N as previously described. The
chloride analysis was usea to indicate the percent recovery of nitrate-N,
which in all cases was found to exceed 95%. Experiments were conducted in
this manner with incubation periods varying between 7 and 29 days. The
columns were also incubated in a flooded condition by adding 1.5 L of
distilled water to each column. This resulted in a 10 cm depth of water
above the soil.
Denitrification Studies
In the first experiments, 50 mg of NO--N as Ca(NO.)2 (equivalent to 110
kg N/ha) in 10 mL of water was added to the top of each column and rinsed
into soil with 500 mL of distilled water. Following the nitrate addition,
two columns (17 and 18) were flooded by adding another liter of distilled
water, and two columns (15 and 16) were allowed to drain.
A rubber stopper was fitted into the top of each column. A gas
dispersion tube passed through the stopper and extended almost to the soil
surface (Figure 4). Another glass tube which protruded slightly through the
stopper was connected to a nitrous oxide adsorption train which consisted of
a CaCl2-drierite drying tube, an ascarite trap to remove C02, a second
drying tube, and finally two stainless steel tubes (15 cm x 6.4 mm i.d.)
filled with preconditioned molecular sieve 5A to trap N_0. The exit of an
adsorption train was connected to an air pump or water aspirator which
pulled air through the gas dispersion tube and flushed the column head space
to remove any evolved gases. In the case of the flooded columns, the
dispersion tube extended below the level of the water so that any dissolved
nitrous oxide would be forced out of solution. The drained columns were
purged continuously at a slow rate (2-4 mL/min), whereas the flooded columns
were purged every 2-4 days for a 2-h period at about 60 mL/minute. Every
2-4 days, the traps were removed and their nitrous oxide determined by gas
chromatography. A correction was made for the amount of nitrous oxide added
from the air purge. Following analysis, the traps were put back into the
adsorption train, and the air flow was continued. It was found necessary to
replace the first calcium chloride drying tube after each analysis due to
the accumulation of water and the resulting tendency to become plugged.
Twelve days after the nitrate application, the columns were leached with
50 mL of 1 N calcium chloride followed by 6.5 L of distilled water. The
leachate was collected over a two-day period in two fractions of about 3 L
each. Two-hundred mL of distilled water were added initially to the 3.8-L
plastic collection bottle to cover the end of the column drainage tube in
111-12
-------
Flowmeter
Gas Oispersic
Tube
122 cm
Glass Wool
^SU-LsSJ
/ZZEJr
15cm
CoCU
i-*Swogelok Fitting
bleculor Sieve 5A
6.4mm
-Rubber Stopper
FIGURE 4. Apparatus for collecting nitrous oxide from Histosol columns.
order to minimize nitrous oxide loss. From each fraction 1200 mL was saved
and kept at 2°C in tightly capped bottles until analyzed for N~0, NO--N,
NH4-N and Cl.
To analyze the nitrous oxide content of the leachate, a 1000-mL saicple
was placed in a liter suction flask which was sealed with a rubber stopper
through which a fritted gas dispersion tube extended to within 2 cm of the
bottom of the flask. A separate piece of glass tubing, which barely pro-
truded through the bottom of the stopper, was connected to the adsorption
train used to collect gaseous nitrous oxide. Using a water aspirator as a
111-13
-------
pump, air was drawn through the gas dispersion tube and bubbled through the
water sample at about 60 iriL/miriute to remove dissolved nitrous oxide. The
water was simultaneously stirred using a magnetic stirer. The air purge was
continued for 2 h, which removed more than 99% of the dissolved nitrous
oxide. The NO content of the molecular sieve traps was determined by gas
chroraatography, and a correction made for the N00 added from the air.
The same experiment was repeated, except that the two previously flooded
columns (17 and 18) were drained, and the two previously drained columns (15
and 16) were flooded. N?0 evolution was again monitored in the headspace
and in the leachate for a 12-day period.
Following this experiment, the columns were left in a drained condition
without addition of NO. for another 12-day period so that baseline data for
the rate of N_0 evolution without addition of nitrate could be determined.
AQUATIC PLANT NUTRIENT REMOBILIZATION
Experiments were carried out in a greenhouse beginning April 15, 1976
and ending June 19, 1976. Cultures consisted of 12-L battery jars filled
with 3 kg of high fertility soil, and fitted with sediment water testing
devices consisting of gas dispersion tubes set into the sediment, with a
smaller plastic tube inserted from the top with which to withdraw water,
filter it (0.45p), and analyze for MRP (Figure 5). These devices were
placed in the center of the jars at a depth of 5-6 cm below the sediment
surface. Six, 4-5 cm shoots with roots present of Myriophylluro spicatum
were planted in each of four jars. Before this was done, all leaves were
removed and shoots were washed to remove epiphytes. Depth of planting was
1-2 cm, and 300 g of autoclaved soil was spread evenly over the sediment
surface to retard the development of algae. Demineralized water was added
to each culture so that a volume of 10 L bathed the shoots. Four control
jars were prepared in a similar manner, the only difference being that they
were left unplanted. A glass tube (diameter 2 ran) was immersed to a depth
of about 20 on in each culture. Air filtered through carbon and glass wool
was then bubbled through each container at a rate of approximately 2.5 L/h
to provide for mixing and aeration of the water. Each jar was also fitted
with a 4-mm diameter tygon tube for gravitational drawing-off of water
samples. All containers were covered with clear plastic, and black plastic
was taped around the outside of the base of each jar to the level of the
sediment surface, in order to curtail growth of algae at the sediment-glass
interface.
For the first phase of the experiment lasting four weeks, changes in MRP
concentration were monitored for sediment and water compartments. All
samples were filtered through 0.45u Millipore filters before analysis by the
method of Murphy and Riley (1962). After one month of growth, one plant was
harvested from each jar, the length of the harvested material measured, and
the number of nodes recorded. Node counts were made to 1 cm from the tip of
the shoot, with leaves dried at 105°C, and percent phosphorus determined by
the method of Grewling (1966), with Murphy and Riley1s (1962) ascorbic acid
reduction method used for analysis.
111-14
-------
FIGURE 5. Diagram of experimental setup for measuring P movements between water
and sediments, including sampling tubes/ aeration line/ and plants in
battery jars of 10-L capacity.
For the second phase of this experiment (two weeks) the response of
well-developed milfoil plants was monitored in water with a poor nutrient
supply. Sidewall algae was removed from the drained battery jars and
demineralized water was added. Monitoring of ortho-P concentration in the
water compartment was continued/ and a reading for ortho-P in the sediment
water was taken. Six days later (May 21/ 1976)/ cultures were drained, node
counts and shoot length measurements were made, and cultures were refilled
with demineralized water. After eight additional days of growth/ one total
plant, including senescing shoots, was harvested from each jar. All visibly
111-15
-------
senescing shoots were also removed at this time so that only actively
photosynthesizing tissue remained in the cultures. A sample of this live
tissue was also taken. Node counts were made on the plants remaining in the
cultures.
For the final experiment (three weeks), sidewall algae was removed, and
culture and control water was exchanged with demineralized water. Suffi-
cient KH2PO. was added to each jar to raise the phosphorus level to about
500 ppb. Monitoring of water and sediment-water was continued for 25 days,
at which time all plants were harvested. Epiphytes and necrotic tissue were
separated by dipping the harvested material in distilled water and shaking
lightly. As before, node counts, stem lengths, dry weights and tissue P
analysis data were obtained. In addition, dry weight and percent P was
established for a composite of epiphytes and necrotic tissue which had been
separated in distilled water.
111-16
-------
SECTION 3
RESULTS AND DISCUSSION
NUTRIENT OUTPUT OF MUCKLANDS
The losses of nutrients from the cultivated muckland sites are given in
Table 3. The annual per hectare loss of NO.-N did not vary greatly among
the sites studied and was on the order of 10% or less of the estimated
potential. The NO.-N losses measured were nevertheless higher than those
expected from most cultivated mineral soils. The variation in the depth of
the organic soil sites studied did not appreciably affect NO--N losses.
Mineralization of organic N, which is the major source of N and largely
restricted to the aerobic soil zone, was probably comparable in the soils
studied. In general, one might expect that the denitrification process
which converts NO- to N_O and/or N2 would be progressively reduced as the
depth of an organic soil decreases since carbon sources may become limiting.
Counteracting factors are (1) the possibility that the drainage water
TABLE 3. NUTRIENT OUTPUT OF CULTIVATED MUCKLANDS
Site Period NO,-N NH.-N MRP TSP
- * - (kg7ha)
Elba muck
GR-A 2/1/75 - 7/31/75 36.0 <0.5 12. ND*
8/1/75 - 7/31/76 87.5 <1 30.7 ND
GR-B 2/20/75 - 7/31/76 41. <0.5 0.9 ND
8/1/75 - 7/31/76 69.5 <1 0.9 ND
South Lima 3/6/75 - 3/5/76 92.6 1.9 11.0 17.9
*Not determined
111-17
-------
contains sufficient available carbon so that denitrification occurs in the
mineral substratum; and (2) the quantity of N mineralized may be less in the
shallower soils. Comparison of our data with that obtained by Miller (1974)
for organic soils shallower than the New York sites suggests that any
influence of organic soil depth on NO,-N losses is minimal. It may be that
the most important factor is the deptn to water table maintained by the
farmer since this affects both mineralization of organic N and the creation
of a favorable environment for denitrification.
The amount of N mineralized in the soil column studies corresponded to
500-600 kg N/ha on an annual basis for columns 15, 16, and 18 incubated in a
drained condition (Table 4) . These values are without correction for the
amount of N denitrified. Based on the denitrification studies, an estimated
one-third of the N mineralized under these conditions would be denitrified,
so that between 750-900 kg N/ha was likely mineralized. The amount of N
mineralized per unit time was independent of incubation time and was main-
tained over successive incubation periods. This result would be expected
for soils that have a large reserve of organic nitrogen.
Under flooded conditions, the amount of N mineralized decreased to
170-190 kg N/ha for columns 15, 16 and 18, and almost one-half of this was
NH.-N. Column 17 was poorly drained compared to the others and the
N-mineralization data obtained reflect its wetter condition in the drained
state.
Denitrification in soils has long been of interest because the process
causes loss of fertilizer nitrogen. In organic soils, denitrification
appears to be beneficial in that it dramatically reduces the output of
NCL-N. However, the question of whether NJ3, one of the denitrification
products, has a detrimental effect on the earth's protective ozone layer has
recently been raised. This question is a long way from being resolved, but
is conceivable that pollution from NO- would be preferred to pollution from
N2°'
The denitrification sequence is:
nitric nitrous
nitrate nitrite oxide oxide nitrogen
The biochemistry of denitrification is complicated and is not fully under-
stood. Nitric oxide appears to be an enzyme-bound intermediate and is not
usually released. Nitrous oxide and nitrogen are the cannon products of
denitrification, but the N20:N2 ratio is extremely variable and cannot be
predicted. This fact, and the ubiquity of nitrogen make it extremely
difficult to study denitrification directly.
The usual approach is nitrogen balance studies is to attribute to
denitrification the difference between the initial NO,-N and the sum of that
remaining plus the amount lost by leaching. This approach was adopted in
111-18
-------
TABt£ 4. MINERALIZATION OF NITRQGBt FROM HISTOSOL COUMJS.
K
VO
Column
No.
15
16
•
17
18
IncUbation
Time
Days
7
14
20
29
14
7
14
20
29
13
7
14
20
29
14
7
14
20
29
14
Soil
Condition
Drained
Drained
Drained
Drained
Flooded
Drained
Drained
Drained
Drained
Flooded
Drained
Drained
Drained
Drained
Flooded
Drained
Drained
Drained
Drained
Flooded
NO,-N in
Leachate
4.75
9.88
13.90
19.92
1.90
5.22
8.85
13.82
16.26
2.08
1.91
3.59
6.73
10.99
1.16
4.68
8.93
11.78
16.91
1.74
NH.-N in
Leachate
rag
0
0.15
0.45
0.31
1.38
0.04
0.32
0.20
0.08
1.04
0.48
1.56
0.53
0.73
1.02
0
0.49
0.15
0.07
1.28
Total N
in Leachate
4.75
10.03
14.35
20.23
3.28
5.26
9.17
14.02
16.34
3.12
2.39
5.45
7.26
11.72
2.18
4.68
9.42
11.93
16.98
3.02
Annual
rag/column
248
261
262
255
85.5
274
239
256
206
87.6
125
142
132
148
56.8
244
246
218
214
78.7
Release of N
kg/ha
545
575
576
560
188
604
526
563
453
193
274
313
291
325
125
537
540
479
470
173
-------
the present study of denitrification in soil columns, with additional
monitoring of N20 production. The procedure for collection of N2O was
developed in this study. The results, which are shown in Table 5, should be
interpreted in terms of trends and generalizations rather than accurate or
absolute quantities. A substantial amount of the added NO was denitrified
in the soil columns when they were flooded for a 12-day period and
denitrification occurred even when the soils were kept drained for this
period. These two conditions often exist together in an organic soil
profile in the field, e.g. when the water table is maintained above the
drainage tile. The data for the soil columns incubated in a drained condi-
tion also support the hypothesis that denitrification can occur at anaerobic
raicrosites within a well aerated soil. Thus, both mineralization of nitro-
gen and denitification can occur simultaneously within the soil.
The data clearly show that large amounts of nitrogen are being
mineralized annually in histosols and that denitrification removes about 90%
of this nitrogen before the drainage waters enter streams. It is possible
that the soluble nitrogen output of histosols could be further reduced by
maintaining the water table as high as possible consistent with the use of
the soils.
There was a great variation in the loss of MRP (nominally orthphosphate)
from the sites studied (Table 3). At the Grinnell farm, the A and B sites
are deep and shallow soils, respectively. The tile lines are soley in
organic material at the former site but are in a mixture of organic and
mineral materials at the latter site (see site description). The smaller
loss of MRP from the shallow site is attributed to fixation of phosphorus
through interaction with the mineral material, perhaps mostly the marl.
Some idea of how representative the data is fron the Grinnell A and B
sites can be obtained by comparing their nutrient output with the nutrient
load in Oak Orchard Creek at Oak Orchard Poad (station 2). For this pur-
pose, the assumption is made that the nutrient load in the creek at Oak
Orchard Poad is derived soley from the 2000 ha of cultivated muckland in the
drainage basin at this point. Data for a 24-h high flow period (discharge
7.1 m3/s at Oak Orchard Road) is presented in Table 6.
The 4.0 kg N/ha value derived from loading in the creek is slightly
higher than either of the two Grinnell sites. This is because the creek
contains a significant NO--N load before it enters the muckland. We cannot
determine this loading precisely until an accurate rating curve is estab-
lished for the staff gauge at station 1 but it will likely be less than 20%
of that at Oak Orchard Road.
The data for MRP indicates that the Grinnell A site is more representa-
tive of the 2000 ha of muckland than is the Grinnell B site. This is not
unexpected as much of the muckland is deep. The MRP loading in Oak Orchard
Creek at station I is insignificant.
The output of MRP from the South Lima site was intermediate between the
two Grinnell sites. Although the Lima deposit is underlain by calcareous
material it is not extensively tile drained and water may not move through
111-20
-------
5. DDHTOIFICfiTION OF NITWaE-flMQJDH) HXSTOSOL CXUMIS.
H
Oolum
No.
15
16
17
18
15
16
17
18
Soil
Condition
Drained
Drained
Drained
Drained
Flooded
Flooded
Flooded
Flooded
Total
(ng)
0.894
0.582
1.866
0.470
7.636
5.786
3.462
6.357
Leachate
% Total
52.0
35.6
11.3
43.0
97.6
95. 5
79.9
78.1
Control*
N2O-N
0.326
0.345
0.239
0.255
Net Total
u riifii
n»\^n
0.568
0.237
1.627
0.215
7.310
5.441
3.223
6.102
mineralized
6.15
6.17
2.68
5.25
1.62
1.92
0.99
1.49
Total NO,-*!
in soil"
m
56.15
56.17
52.68
55.25
51.62
50.99
50.99
51.49
C SO,-N
recovered
49.95
48.98
32.40
46.71
11.09
10.97
11.95
16.92
NO.-N
denitrified
6.20
7.19
20.28
8.54
40.53
40.95
39.04
34.57
t Total
N03-N
11.0
12.8
38.5
15.5
78.5
78.9
76.6
67.1
Tbtal N.O-N
% of Nof-M
denitrified
9.2
3.3
8.0
2.5
18.0
13.3
8.3
17.7
Value* derived from unfertilized, drained ooluma.
' Values derived from previous mineralization studies fron each oolunt.
1 Values include 50 ag NOj-N added to each colun.
-------
TABLE 6. NUTRIENT LOSS FROM 2000 HA OF THE ELBA MUCKLAND IN A 24-HOUR HIGH
FLOW PERIOD (2/19/76 TO 2/20/76).
Site NO.-N MRP
J (kg/ha)
Grinnell A 3.8 0.89
Grinnell B 2.2 0.023
Calculated from load
in Oak Orchard Creek 4.0 0.43
the calcareous material, but rather over it until it intercepts a drainage
ditch. There may be limited contact between the drainage water and the
underlying mineral strata in the creek bed. The data from the Karas site,
where the tile lines are in marl, could not be quantified because of diffi-
culty in measuring the pumping rate and unreliable sample collection.
However, it can be deduced from the concentration data obtained that the
output of MRP from this site was similar to the shallow Grinnell B site.
Considerable trouble was encountered with measurements of total soluble
phosphorus (TSP) in samples from the Elba muckland sites. The USEPA
persulfate digestion procedure was used for digestion of samples. Sample
TSP values at least 1 ppm less than the corresponding MRP values were
obtained on a number of occasions, and on others there was essentially no
difference between the TSP and and MRP values. If the digested samples were
diluted by various amounts, but always working within the linear range of
the method, values differing by as much as 10X were obtained. The TSP
values increased with increasing dilution which suggested that digestion of
the samples released an interfering substance; however, none could be
detected. Some samples from Little Conesus Creek were analyzed for TSP
after a dry ashing procedure, but dilution of one set of samples by various
amounts prior to ashings of a standard volume gave values 10X those for TSP
in the original samples.
The manipulations mentioned in the preceding paragraphs did not affect
the values obtained for MRP, which was also unchanged when an iso-butanol
extraction procedure was used. After much effort it was concluded that the
MRP values were accurate and that there probably was little organic
phosphorus in the samples studied.
111-22
-------
From a survey of the literature/ it can be seen that generalizations on
the leaching of phosphorus from organic shallow soils cannot be made.
Certainly losses of phosphorus from shallow soils can be as high as those
found from the deep Grinnell A site. On the other hand, it is not clear
whether the losses of phosphorus from deep histosols are necessarily high.
Even so, it is apparent that the losses of phosphorus from histosols are
considerably higher than those for mineral soils (300 to 1000X in the
present study). In this connection it should be noted that the phosphorus
leached from histosols is soluble, whereas much of that derived from mineral
landscapes may be in a particulate form that is biologically inert.
It is difficult to assess possible management practices for reducing
phosphorus losses from histosols because of uncertainty in which factors
control phosphorus release. At this time, the effects of reducing or
eliminating fertilizer phosphorus are unknown and treatment of the effluent
by conventional sewage treatment methods would be expensive.
The output of nutrients (N,P) and nutrient concentrations were
positively correlated with flow at all sites (Figures 6 and 7). Thus, as
others have indicated, the importance of orienting a sampling program
towards flow events cannot be overemphasized. The fact that a major portion
of the nutrient loss occurs during flow events complicates management of the
soils to reduce nutrient losses. The major flow events on the New York
mucklands occur in the winter and spring months and are most intense at a
time when removal of water from the land is of prime importance to the
farmer. They also occur during a time period when there is no agricultural
use for nutrient rich water.
Oak Orchard Creek
Oak Orchard Creek was sampled monthly in 1974, bi-weekly in 1975, and on
an event basis in 1976. The sampling locations are shown in Figure 1. MRP
and N03 concentration data are presented in Tables 7 and 8, respectively.
Where available, discharge data are also presented. The water status of the
creek can be inferred from the concentration of MRP at Oak Orchard Road
(station 2) for dates where discharge data were not available. Thus, when
the concentration of MRP is 1 ppm or more flow is in the moderate-high
range.
Plots of nutrient concentrations against discharge at Oak Orchard Road
(station 2) and Harrison Road (station 11) are shown in Figures 8-11. From
these, it is evident that the creek load of both MRP and NO--N is positively
correlated with flow. This is expected as nutrient losses from the muckland
increase as flow increases. At Harrison Road, the concentration of MRP is
fairly constant between 2.8 and 21. m3/s, but increases where the discharge
is below 1.4 m3/s (Figure 9). This appears to be a seasonal effect which is
not related to the cultivated muckland.
At moderate to high flow there is a gradual decrease in MRP concen-
tration with increasing distance downstream from the cultivated muckland
(Table 7). At low flow, which occurs during the summer and fall months, the
concentration of MRP decreases between Oak Orchard Road (station 2) and
111-23
-------
K
V
ro
*>.
90r
PUMPING TIME
MRP
NOj-N
1/30 Z/IO Z/ZO 3/1 3/10 3/20 4/IO 4/ZO 4/30 S/IO S/20 5/30
DATE
FIGURE 6. Pumping time, NO_~-N and P concentration for water flowing fron the muck area served by
Site GR-A. Pump time is directly proportional to flow volume.
-------
90r
en
PUMPING TIME
MRP
NOj-N
12/1 12/10 12/20 12/30
1975
5/30
FIGURE 7. Pumping time, NO- -N and P concentrations for water flowing from the muck area served by
GR-B. Punp time is directly proportional to flow volume.
-------
TABLE 7. CXDNCENTRATIOJ OF MDLYBDATE REACTIVE PHOSPHORUS IN OAK ORCHARD CREEK.
Date
1
2
Station
98 7
8 9
11
EO
ppn P (discharge m3/s)
12/7/73
1/4/74
2/6
3/5
4/3
5/1
6/5
7/1
9/15
3/24/75
4/8
4/17
4/25
5/21
6/3
6/18
7/2
7/16
7/30
0.04
0.03
0.02
0.08
0.08
0.01
0.08
0.03
0.01
0.08
0.02
0.0
0.02
0.02
0.02
0.04
0.04
0.02
0.02
1.5
1.3
1.3
1.4
1.8
0.36
0.66
0.77
0.42
1.50
(2.5)
0.53
(1.3)
0.97
(1.6)
0.43
(0.4)
0.38
(0.1)
0.48
(0.1)
0.68
(0.2)
0.57
(<0.1)
0.51
0.57
0.49
0.40
0.35
0.32
0.41
0.20 0.12
0.24 0.16
0.39 0.31
0.21
-
- -
_
- -
_ _
-
- -
>
-
_ _
0.38 0.29
0.31 0.21
0.27 0.23
0.29 0.18
0.38 0.29
0.08 0.06
0.14 0.09
0.24 0.15
0.21 0.18
-
- -
-
- -
- -
-
-
- -
-
v —
0.16
0.16
0.17
0.12
0.15
0.12
0.16
0.29
0.32
0.26
0.16
0.07
(14.)
0.12
(6.9)
0.25
(<0.1)
0.37
(<0.1)
0.45
(1.4)
0.58
(1.2)
0.38
0.38
-
-
-
-
-
-
-
-
-
-
-
0.02
0.07
0.15
0.12
0.19
0.10
0.13
0.10
111-26
-------
Date
8/14
8/28
9/11
9/29
11/24
12/18
12/30
1/14/76
1/26
1/27
1/28
2/17
2/18
2/19
2/20
2/21
1
0.02
0.01
0.06
0.06
0.01
0.07
0.06
0.07
0.05
0.28
0.22
0.08
0.10
0.11
0.06
0.05
2 98 7 8 9
ppn P (discharge m3/s)
0.44 -
t^ f\ 1 t
i ^y ill
0.27 -
0.22 - 0.18 0.21
0.30 - 0.15 0,18
0.39 -
1.4 - 0.45 0.31
(0.9)
0.73 -
(0.3)
0.57 - 0.15 0.16
(0.3)
0.53 - 0.13 0.14
0.56 - 0.30 0.17
0.75 - 0.37 0.32
1.3 - 0.41 0.33
(7.2)
1.6 - 0.43 0.35
(6.8)
1.4 - 0.47 0.36
(7.1)
1.8 - -
(4.4)
2.1 - -
(3.5)
11
0.41
0.43
0.29
0.24
0.16
0.17
(4.7)
0.17
(0.9)
0.14
0.12
(3.3)
0.11
(3.5)
0.12
(3.3)
0.10
(17.)
0.09
(18.)
0.09
(18.)
0.11
(20.)
0.14
(20.)
EO
0.48
0.08
0.35
0.18
0.38
0.08
0.07
0.06
0.03
0.07
0.06
0.10
0.08
0.10
0.06
-
111-27
-------
Date 12 98 7 8 9 11 BO
2/22
2/23
2/24
2/25
2/26
2/27
2/29
ppn
0.10 1.5
(7.4)
0.07 1.9
(4.7)
0.07 2.1
(3.4)
0.05 2.1
(3.1)
0.11. 2.2
(2.4)
0.11 2.1
(2.0)
0.06 1.9
(1.2)
P (discharge m3/s)
- 0.15
(21.)
- 0.16
(21.)
0.52 0.43 - 0.17
(21.)
0.57 0.46 - 0.16
(20.)
0.55 0.45 - 0.18
(20.)
0.49 - 0.17
(19.)
0.59 0.46 - 0.15
(17.)
0.09
0.03
0.04
0.06
0.04
0.03
111-28
-------
TABLE 8. CONCENTRATION OF N03 IN OAK ORCHARD CREEK.
Date
12/17/73
1/4/74
2/6
3/5
4/3
5/1
6/5
7/1
9/15
3/24/75
4/17
4/18
4/25
5/21
6/3
6/18
7/2
7/30
8/14
8/28
9/11
9/29
11/24
12/18
12/30
1/26/76
1/27
1
2.3
10.5
12.0
2.2
5.2
8.0
5.0
5.5
0.2
4.2
6.9
4.1
8.2
4.9
5.1
0.5
3.3
0.3
0.1
0.1
0.06
0.06
0.0
4.7
1.1
4.5
3.2
2
15.3
12.6
13.3
6.6
8.8
3.7
2.3
2.9
0.2
7.3
6.3
6.1
5.2
3.6
2.0
0.4
0.4
0.3
0.0
0.0
0.2
0.3
0.1
23.9
7.6
8.1
7.8
Station
98 7
NO,-N ppn
8.9
7.0
4.8
3.4
3.7
3.4 2.6
1.6 1.2
2.4 1.7
0.4
-
-
- -
- -
- -
-
- -
- -
-
-
- -
0.2
0.2
-
7.4
- -
3.9
5.6
111-29
8
9.8
4,0
3.6
3.0
3.4
2.6
1.1
1.8
0.3
-
-
-
-
-
-
-
-
-
-
-
0.2
0.2
-
5.5
-
3.1
4.1
9
5.7
3.3
3.2
2.2
3.3
1.1
0.7
1.1
0.1
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
11
5.5
2.8
3.3
1.9
2.4
0.1
0.1
0.2
0.0
2.5
3.5
0.4
0.2
0.0
0.0
0.0
0.0
0.0
0.0
0.0
0.3
0.2
0.1
3.4
2.7
2.3
2.1
EO
-
-
-
-
-
-
-
-
-
-
-
0.7
0.5
0.3
0.4
0.4
0.4
0.1
0.2
0.0
0.4
0.2
0.1
2.1
0.2
2.3
3.0
-------
Station
Date 12 98 7 8 9 11 BO
NO,-N ppn
1/28
2/17
2/18
2/19
2/20
2/21
2/22
2/23
2/24
2/25
2/26
2/27
2/29
4.0
3.5
4.1
4.5
6.1
7.5
3.5
6.8
8.5
8.4
9.1
9.0
10.4
9.3
11.0
12.8
13.2
15.0
15.8
10.6
12.3
14.5
14.5
14.5
14.5
12.5
w -
4.9
6.0
6.5
5.9
-
-
-
-
7.0
7.1
7.2
-
7.3
5.3
5.3
2.7
2.7
-
- -
_ _
- -
6.0
6.2
6.3
6.8
7.1
2.0
2.4
2.5
2.6
3.1
3.3
3.3
3.7
3.8
4.3
3.9
3.8
3.6
3.1
3.0
2.0
2.7
2.8
-
2.7
-
3.1
3.0
3.4
3.5
3.7
111-30
-------
Knowlesville Road (station 9), then increases between this point and
Harrison Road (station 11) (Table 7). The reason for this increase is
unknown; there are no obvious point sources. It is possible that MRP is
picked up as the creek flows slowly through the swamp in the Iroquois
wildlife refuge. At Oak Orchard Road, the concentration of MRP increases as
flow increases, but appears to plateau between 1.5-2.0 ppm at flows of 2.8
m3/s or greater (Figure 8).
The concentration of nitrate in Oak Orchard Creek at Oak Orchard Road is
generally high at high flow (Figure 10), which again reflects the behavior
of the cultivated muckland. The highest concentrations are found during the
first event after the summer. This is presumably related to mineralization
of soil organic nitrogen and its accumulation in the soil profile during the
warm summer months when little or no water is draining from the soil.
3 -
0.
cc
5
• DEC- APR
O MAY-NOV
2 4
DISCHARGE (m3/s)
6
FIGURE 8. Oak Orchard Creek at Oak Orchard Road (station 2): MRP vs.
discharge, 1975-76.
111-31
-------
.5
1.3
a
a.
tr
5
.2
• DEC-APR
O MAY-NOV
1O
DISCHARGE (m3/s)
20
FIGURE 9. Oak Orchard Creek at Harrison Road (station 11): MRP vs. dis-
charge, 1975-76.
25
2O-
15
101-
£
O
O
00
• Main spring event, 1976
O Fall, winter 1975
2 4
DISCHARGE (m?/s)
FIGURE 10. Oak Orchard Creek at Oak Orchard Road (station 2): N03~-N vs.
discharge, 1975-76.
111-32
-------
Relationships between nitrate concentration and discharge at Harrison
Road can be seen if the data are separated into two seasons (Figure 11).
Between December and the middle of April, concentrations are in the range
2-4.5 ppm NO--N, with the higher values at high flow. For the rest of the
year, nitrate concentrations are always low, even though there may be
considerable inputs into the creek from the cultivated muckland. During
this period, nitrate may be removed iron the creek by plant intake, by
denitrification, or by both processes.
Two observations worthy of note here are:
1) the maximum discharge at Harrison Road is about 3X that at Oak
Orchard Road, although there is a seven-fold increase in the area of the
drainage basin between these points. Oak Orchard Swamp, which is flooded at
high flow, buffers the impact of a flow event. The spreading out of water
in the swamp appears to reduce the load of MRP in the creek, but not the
load of nitrate.
2) The loss of nitrate and MRP from the cultivated muckland always
increases as flow increases. This leads to an increase in the concentration
of these nutrients in the creek at Oak Orchard Road during high flow
compared to low flow. However, within an event the concentration of nitrate
and MRP does not always increase as flow increases. For example, for the
principal event of spring 1976, the concentration of IKL-N and MRP increased
from 13 to 16 ppm and 1.3 to 2.1 ppm, respectively, over a 24-h period as
discharge decreased from 7.6 to 3.3 ma/s. Twenty-four hours later when
discharge was back up to 7.4 m3/s, NOv-N and MRP were down to 11 and 1.4
ppm, respectively. A similar pattern was also noted for Little Conesus
Creek. An obvious explanation for this observation is that the muckland
contributed a greater percentage of the water in the creek as discharge
decreased. However, the nutrient load in the creek was highest at peak
flow.
• DEC-APR 15
O APR 15-NOV
!T
to
DISCHARGE (m3/s)
20
FIGURE 11. Oak Orchard Creek at Harrison Road (station 11): NO,~-N vs. dis-
charge, 1975-76.
111-33
-------
Little Conesus Creek
For the period when Little Conesus Creek was monitored at site 'F1 for
flow and N and P concentrations, the total annual losses from this muck
basin were about 18 kg/ha for TSP, 11 kg/ha for MRP, about 93 kg/ha for
N03-N, and 1.9 kg/ha for NH.-N (Table 9). The P loss is 10 to 100 times
that usually measured for a cultivated mineral soil, and the losses for N
are high but less spectacular.
Flow was calculated by Equation 1, the regression equation (r=0.944,
n=102) correlating instantaneous flow readings based on the staff gage at
'F' (excluding December, January and February readings) and the average flow
on the same days established by a stage height recorder on Oak Creek at
Warsaw, New York:
Q - .99357 (W) (K) - 295.303 [1]
where Q = m3/km2/day at 'F1
W = average daily flow at Warsaw in ft3/s
K = 22.447, a constant converting ft3/s to
m3/km2/day for the basin
The correlation coefficient between flow at site 'O1 and Warsaw was
0.875, so site 'F1 was chosen as the primary downstream site. There are no
substantial inputs into the main stem between the muck and site 'F' and flow
did not increase appreciably between sites '0* and 'F'.
From Table 9 it appeared that concentrations may decrease between site
'O1 and 'F1. This was especially looked for in the case of NO--N, because
the streambed should be an ideal environment (anaerobic, warm and high in
organic carbon) for denitrification. No such trend was established, and
there are uncertainties about apparent concentration differences between 'O1
and 'F1 in Table 9 because of the large error terms. Therefore, it was
assumed that sites 'O' and 'F1 do not differ in flow or load, and that flow
at 'F1 is calculated with minimal error.
Three methods were initially compared for calculating N and P loads
(Table 10). The first method uses the mean annual concentrations with a
sizable standard error term. In fact, these numbers appear to be low
compared to the other two methods, and the error terms range up to over
100%. The second method, using mean concentrations and flows for shorter
time periods, and the third method, using both instantaneous values and
means, produce similar numbers for the total annual N and P loading in
Little Conesus Creek. The second method was used to calculate the final
loss values as outlined in Table 9. The error terms are relatively less
than those by the first method. According to Treunert et al. (1974),
sampling once every two weeks with discontinuous flow measurements produces
a 20-35% error. In this study, a total of 57 samples were collected over
the year for an average of one every six days. Therefore, it is felt that
the error terms associated with the second method (Table 9) represent the
maximum. However, they serve well to emphasize the problems in sampling
111-34
-------
se-in
,
1
tff
s
s
*
9-
S
S
S
S
1
1
i
1
i
1
i
§
•
*i
s
1 1 1 i I ! 1 1 ! |
5* C K
S 1 1 8 s s 2 t * §
• • • • b in ' i, ii '
Ss!28~BI3SS
e * B *" ""*
iSSSscSsSS
• • • • i i ' u • '
S53SC..S8.H
- k - 8
tsspsKoSSc*
" " "S""""
£ -.4 • fc o C •» J - *
s B i I g « a i i i
«. s
5 2 i g s „ i * i 1
*
.S§»~SSBS
S • j. w v •» k
' » J» • !• * e e
£
R
f
R
S
R
f
R
*
R
V<5*>*VSS^KSiC
5 5
"s i S * 1 i "s s i a
52SS5SS6IS
§ 1 S 5 S § s 8 S S
; £ S s 5 si £ 2 5 i £'
s 5 8 E § k 88 § k
S215St'cSSsa
g i s s i s 2 k k k
|hssggbkbs
gggkJisbhKS
. • n k 1 1 g s i
• • k k k k k i k k
S g S S S S i S S 2
5 s i s k k k s s i
t g £ g g S E 2 i S
S5g5S?*I§l
g i s i 1 k § i e k
g s 3 S g | s e « s
§ S S » S S • 8 S i
- • g s § s i s i k
- ' s § i k k s k k
•ja
til
i
a
•
n
?
a
1
N
H
H
R
1
N
1
11
i
H
1
R
4
H
1
,
I
i
|
J
'
8
I
?!
i
^s
j
j
i
i
s
\
-
-------
TABLE 10. TOTAL ANNUAL LOADINGS IN LITTLE CCNESUS CREEK NEAR E. AVON ('F'), MARCH 6, 1975 TO MARCH 5, 1976, CALCULATED BY
THREE DIFFERENT METHODS.
u>
Methods TSP MRP DOT NO~-N
1. Product of the annual mean
concentration, C, and total
flow, Q.
EQ » 4666666 m'/KTO". - — — --- — — — — •— - -- — — — ~ -- — — [411^-— -- — — — — -- — ~
0.330 0.236 0.095 1.73
c - SB* ±0.309 ±0.172 ±0.198 ±1.65
_____ Vn
*9
. 3210.9 2296.3 924.4 16833
cEQ - SE* ±3006.6 ±1673.6 ±1926.5 ±16054
NH*-N
0.054
±0.039
525.4
±379.5
2.
Sum of products of mean
concentration and flows
for specified intervals
delineated by shifts in
concentrations (see
Table 9).
4076
494
1572
21,087
425
3.
Sun of products of the
instantaneous concen-
trations and flow for
sample days, and the
products of mean con-
centration and flow
between sample days.
3911
2555
1363
23,117
479
*SE, standard errors of the mean.
-------
frequency and timing which exist when attempting load calculations for small
streams, and the error that must be considered when using these numbers.
Concentrations of TSP at the gauging station (site 'F') generally ranged
from about 0.1-1.5 ppm, MRP from 0.05-1.0 ppm, and dissolved unreactive
phosphorus (DUP) from 0-1 ppm. Nitrate-N ranged widely from less than
0.003-7 ppm, and NH.-N ranged from 0.003-0.2 ppm. There was no definite
relationship between flow and concentration as in Oak Orchard Creek,
although TSP and MRP concentration generally increased with flow, indicating
the nonpoint character of the P inputs as described by Johnson et al. (1976)
for a mineral stream.
DUP, NO.-N and NH.-N showed a closer relationship to season than flow,
with highest concentrations in winter and spring and lowest in the summer
and fall months. In fact, total phosphorus concentrations and loads were
higher than those for N during September and October. Nitrate-N especially
seemed to follow this seasonal concentration pattern described by Johnson et
al. (1976) by a sinusoidal regression equation. They suggested that NO.,-N
concentration is controlled by water movement through the profile with
little or no movement during the summer but increased leaching during fall
and winter from water recharge and plant decomposition. This would seem to
be the case here. Concentration of NH.-N and DUP may be influenced posi-
tively by temperature effects on biological processes and senescence as
evident by higher late spring and fall values (Table 9). These influences
are not precise, but are only suggested as explanations for the obser-
vations.
Flow varied widely at 'F1, the highest flow occurring in March (5.4
m3/s) and the lowest (negligible) during August and September. The creek
always had water in it and movement was discernible, but reliable flow
information was usually not obtained during these two months. The same was
true of December and January because of ice problems.
The greatest losses occurred during a relatively short period of high
flow. Over 75% of P and N lost during the year was measured during a 60-day
period from the middle of February to the middle of April.
The muck area (2.28 km2) makes up about 10% of the entire basin area
(20.85 km2), so major surface water inputs to Little Conesus Creek above the
muck were estimated at flow extremes and summarized in Tables 11 and 12.
Above the muck, the main stem drains predominantly old fields of mineral
soils with grass and brush cover, and some cultivated fields. There was no
livestock associated with the main stem. As indicated in Table 11, the main
stem carried into the muck 20% or less of the N and P measured at site 'F1
on March 4, 1976, while flow was about 11%.
The east feeder (site 'B') drains less total area but there is a small
pocket of muck and at least two barnyards close to or on the branch. It
carried into the muck greater than 25% of the N03-N and less than 20% of the
P measured at 'F1 during the March 4 event. During low flow, N and P loads
coming into the muck were less than 10% of those at 'F1, as were the flows.
The muck occupies only 10% of the area, but contributes at least 80% of the
111-37
-------
11.
FLOW (FLOATING OBJECT ESTIMATE), CONCENTRATION, AND LOAD DATA FOR SINGLE DAYS UPSTREAM
OF THE MUCK ON THE MUM SIGH OF LITTLE CGNESUS CHEEK DESIGNATED *U'.
V
U)
CO
Average
estimated
flow
Date m'/s
3/2/76 0.16
3/4/76 0.64
3/16/76 0.02
3/22/76 0.02
4/8/76 0.02
4/16/76 0.02
4/26/76 0.15
5/4/76 0.01
Concentrations
upstream of muck
NOj-N
1.30
1.58
0.88
0.68
0.041
0.021
0.283
"
mg/L
TSP MRP
0.017 0.015
0.160 0.038
0.038 0.008
0.041 0.004
0.083 0.002
0.042 0.005
0.078 0.035
0.035 0.008
Load
upstream of muck
kg/day
DUP NO--N
0.002 18.4
0.122 87.3
0.030 1.70
0.037 1.46
0.081 0.071
0.037 0.043
0.043 3.58
0.027
TABLE 12. FUM (FLOATING OBJECT ESTIMATE), CONCENTRATION, AND
MIDWAY THROUGH THE
Average
estimated
flow
Date mVs
3/4/76 0.90
3/16/76 0.02
3/22/76 0.02
4/8/76 0.01
5/4/76 0.01
NU3-N
2.80
1.41
1.13
0.032
-
MUCK, DESIGNATED
Concentrations
East feeder
mg/L
•tse rm?
0.133 0.059
0.030 0.004
0.044 0.013
0.110 0.011
0.050
"'
ISfr
0.241
8.84
0.073
0.088
0.144
0.085
0.988
0.035
MHP DUP
0.213 0.028
2.10 6.74
0.015 0.058
0.099 0.079
0.003 0.141
0.010 0.075
0.443 0.545
0.008 0.027
HOj-N
209.
809.
56.4
94.7
3.95
2.85
239.
"
LOAD DMA FOR SINGLE DAYS NT THE
Load at
East feeder
kg/day
DUP NO--N
0.074 217.
0.026 2.07
0.031 2.38
•J.W
10.3
0.044
0.092
0.099 0.027 0.091
-
0.048
MRP BUP
4.57 5.74
0.006 0.038
0.027 0.065
0.009 0.083
-
NO--N
809.
56.4
94.7
3.95
-
Load
downstream of muck
site '¥'
kg/day
TSP MRP
43.5 34.1
68.8 34.1
10.1 9.78
21.4 16.2
2.44 1.75
1.31 0.720
29.1 31.7
4.17 3.32
EAST iJsnxH STREAM
Load
Upstream load
%of
,
load at 'F'
DUP NOj-N
9.34 8.8
34.6 10.8
0.380 3.0
5.18 1.5
0.680 1.8
0.590 1.5
0.0 1.5
0.850
TO LITTLE CONESO
lw
0.6
12.8
0.7
0.4
5.9
6.5
3.4
0.8
3 CREEK,
downstream of Buck East ft
site 'F1
kg/day
'k&f MRP
68.8 34.1
10.1 9.78
21.4 16.2
2.44 1.75
4.17 3.32
HW-
0.6
6.1
0.2
0.6
0.2
1.4
1.4
0.2
DUP
0.3
19.5
15.3
1.5
20.7
12.7
-
3.2
ENTERING
seder load
% of
DUP NO,-4)
34.6 26.8
0.38 3.7
5.18 2.5
0.68 0.7
0.85
load
rar
15.0
0.4
0.4
3.7
1.2
at 'F"
****'
13.4
0.1
0.2
0.5
-
DUP
16.6
10.0
1.2
12.2
-
-------
dissolved N and P carried downstream by Little Conesus Creek at site 'F'.
The fate of these muck-derived, dissolved nutrients downstream of 'F1
could not be monitored because of substantial farming operations, including
mineral soil cultivation and dairying.
Metals Losses
Approximations for metal concentrations are presented in Table 13. It
must be stressed that these numbers and those in Table 14 are approximations
only, based on one analysis of composited water samples.
Calcium was high, indicative of the calcareous till underlying the basin
and the marl layers (calcium carbonate) at a depth of 1-1.5 m under the
muck. The creek itself flows across these marl layers in the muck.
Magnesium was about one-quarter the concentration of calcium, and seemed
to increase as flow decreased. Potassium concentration was appreciable and
increased with flow, as did iron, zinc, and manganese. This may reflect
increased quantities of colloidal-sized particles containing these elements,
which were not filtered out during the sample pretreatment. Sodium, lead,
copper and cadmium concentrations were not especially elevated and
apparently bear little relationship to flow rate.
Table 14 gives approximate dissolved metal losses from the muck area to
Little Conesus Creek. Substantial quantities of calcium, magnesium,
potassium, sodium, and iron were lost from the basin even though these
elements probably do not affect water quality. Those elements (Cu, Zn, Mn,
Pb, Cd) which may be harmful or interfere with water uses showed quite low
losses. A summary of all measured losses from the basin by way of Little
Conesus Creek for the year March 1975 to February 1976 is presented in Table
15.
Aquatic plant growth was estimated at two sites in the creek before,
during and after the 1975 growing season. Site '0' was dominated by the
growth of elodea (Elodea canadensis), coontail (Ceratophyllum demersum) and
pondweed (Potomageton crispus). In contrast, milfoil (Myriophyllum
spicatum) dominated at site 'F1 to the point of exclusion of elodea and
pondweed except in the winter and early spring. This dominance of milfoil
during the warm months has been observed in experimental ponds at Cornell
and the north end of Cayuga Lake (Peverly et al., 1974).
The differences between sites 'O' and 'F1 that explain the predominance
of milfoil at 'F1 are few and hard to precisely define. The rooting medium
at '0' is soft, has high organic matter (12% on a dry weight basis) and is
very fine in texture. At 'F1 the sediment is firm, high in sand but low in
fine-textured particles and organic matter. In addition, the channel is
heavily shaded during the summer at '0' but is not at 'P1. Milfoil seems to
prefer the greater light and coarse, firm bottom at 'F1. However, in Cayuga
Lake, milfoil grows on soft, organic-rich sediments, but will not (or
cannot) colonized the sandier areas (Peverly et al., 1974), especially a
large sandbar left from early dredging operations. This suggests that light
or temperature may be the deciding factor in the growth of milfoil at 'F'.
111-39
-------
TABLE 13. METAL CONCENTRATIONS IN LITTLE OONESUS CREEK ONE KM BELOW THE MUCK, DESIGNATED 'O1 , NEAR SOUTH LIMA.
Composite sanples made up of sanples grouped by flow as indicated.
H
0
O
Date Flow
1975 mVs
6/5
6/9
6/20 0.28-2.8
3/20
3/13
3/17
Metals, ppn
Ca Mg K Na Fe Cu Zn Mn It» Cd
86 21.0 4.4 8.7 0.533 0.004 0.006 0.019 0.014 0.002
..... 0.03-0.28 86 22.4 3.2 9.6 0.238 0.004 0.003 0.011 0.010 0.002
31 t.\l
3/28
7/17
7/27 0-0.03 70 27.6 2.2 8.0 0.039 0.004 0.001 0.002 0.012 0.002
8/2
-------
8£0*0
6££*£I
0£t*0
09*t
Z£*t
68*t
Z£*»
EO*E
06*1
Zt>*0
»££'0
9*0*0
ZU'O
9EZ*0
E£E*0
089-0
6WO
££E*0
iWO
9S£*0
E6t*0
9£'l
»9Z*0
iO
£99*0
6£*t9l
866*0
f£t
fOt
s*9t
£*8E
6*»Z
9*St
Z**Z
9S*»
69Z-0
£S*Z
99*t
£0*£
t9*9
6K'0
££*Z
6Z£'0
SZ'9
Eft
s-n
9S*I
'8
SZ't
9 '61
ti't
UH
8SZ*0
609 *8S
66Z*0
»9'9
£0*C
£9*9
9*91
£*0t
£9*9
9Z£*0
t£*t
180*0
Oft
6EfO
t£*t
0»*Z
6ZO-0
6ft
iZO'O
89 *Z
0*£'0
IZ'9
£9»*0
uz
*8t*0
££6*t»
66£*0
9£'£
»0*»
Z***
ftt
Zf£
9»*»
696*0
Z8*t
80fO
K£'0
SS9*0
££8*0
O9't
9tfO
JJ6i*0
OTT *0
8£'t
>9»'0
»f»
ZZ9*0
no
9V
£f£Z
££ZS
8*EZ
109
OK
68S
»£»t
6»6
»69
9*£S
sot
0»'9
8*£6
tf'S
ill
HZ
Eft
901
£0*t
BCZ
0'£Z
tss
0*£E
aa
peoi
mipoxiddtf
6*SOI>
Z6CZ6
8S6
Z£t8
E696
6196
»90»Z
Z6»9t
9696
SZEZ
9££»
89Z
969t
Ottl
£061
98ȣ
ZCZ
»Z£t
OZZ
609»
0601
£668
£6H
»N
9 •|'6I
tO£»»
6t£
EEtf
t£Z£
998»
0£tZt
S£8£
E06»
S££
89»t
f 98
£08
90E
••96
'E9£t
6*C9
Z£8
f'09
£961
E9E
OSS*
86*
y
Z*9£6
»96tZZ
SCZZ
»Z£6t
£t9ZZ
£IZ£Z
9808S
»6E££
ZOKZ
*Z*S
£OZOt
£09
Z98E
OE8E
Z09»
*t*8
ZOB
t9t»
£9£
69£6
Z*SZ
8UTZ
98*C
&H
»*OUE
Z6*»»8
C8S8
£££08
S£898
I80S6
t£8£EZ
6EtESl
8E8S6
SZ80Z
68I6E
*UZ
£££9t
El£6
8*881
9S»>£
»EOZ
6EO£t
tZ6t
69C8E
t9£6
8C688
6£EEt
*0
H'O
Z'Z
»Z'0
9't
9*t
ȣ'0
BE'O
80 '0
EZ'O
£0*0
Z»'0
zo-o
ZC*0
EE'O
zo-o
8£'0
zo-o
Z£*0
SfO
£*l
SfO
B/,a
8£*JaAV
:
(fee) P
8
9
8*
8
ZZ
8Z
*£
EC
EZ
Ot
S
£6
8
ft
ZZ
9
it
91
8
£
Zt
•Ana
PIM '»U/fa(
9£/S/£-9£/9/E
S/S-8Z/*
£Z/»-£Z/»
ZZ/»-9/E
9/E-9Z/Z
SZ/Z-t/Z
£/Z-£/t
9£/9/l-*/Zt
£/Zt-I/tl
IE/OI-6/Ot
8/OI-OE/6
6Z/6-SZ/6
*Z/6-»Z/9
CZ/9-9t/9
9t/9-Z/9
t/9-tt/S
Ot/9-S/S
»/S-8T/»
£I/»-Z/*
t/»-SZ/E
»Z/£-8t/£
£t/£-S£/9/E
SS^CQ
H
"El »iqm UT BaidiBS K)Tf»iu3D sif» jo
HUUS avsN ,o. oaLwoisaa 'xxw am tcnaa HX aw> XXIID
uoT3BX)uaauao auo Atuo uo paeeq are
snsaeo snun HI sown TOUM aiwtxoiddv
'ft
-------
TABLE 15. SUMMARY OF ELEMENTAL LOSSES FROM THE BASIN OF LITTLE CONESUS
CREEK. Nitrogen and phosphorus losses were measured three km
below the muck, at 'F1. Metals were measured one km below the
muck at '0*, and are approximations.
kg/ha/yr
kg based on
Substance 3/6/75 - 3/5/76 228 ha muck
NO--N
NH^-N
TSP
MRP
DUP
21,087
425
4,076
2,494
1,572
92.6
1.87
18.8
11.0
6.91
Ca 844,492 3,710.
Mg 221,964 975.
Na 92,392 406.
K 44,301 195.
Fe 5,273 23.2
Cu 41.98 0.184
Zn 58.61 0.258
Mn 142.0 0.624
Pb 151.8 0.667
Cd 17.78 0.078
Elodea and pondweed may be inhibited by higher summer light intensities and
temperatures, and may grow better during the cooler months. This is
indicated in Table 16.
The composition of the aquatic plant tissue, and total quantities of N,
P, Zn, Cd and Pb in the tissue are presented in Table 17. Total N and P
seem to peak in late May at site 'O', where there was no milfoil. At site
'F1 where milfoil dominates, the peak was in September. In both cases, this
corresponds to the maxima in standing crops for the two sites (Table 16).
In general, this is also the case for Zn, Cd, and Pb, although in the last
column of Table 17, where average composition and standing crop are used to
calculate totals between 'O1 and 'F1, the two heavy metals Cd and Pb are at
higher total levels in the plants in late sunnier.
111-42
-------
TABLE 16. ESTIMATED STANDING CROP FOR SUBMERGED AQUATIC PLANTS, LITTLE OONESUS CREEK.
U»
Date
12/17/74
4/11/75
5/27/75
7/9/75
9/10/75
10/15/75
11/14/75
Species
Elodea
Coontail
Pbndweed
Elodea
Coontail
Pondweed
Milfoil
Coontail
Pondweed
Milfoil
Elodea
Coontail
Milfoil
Elodea
Coontail
Milfoil
Elodea
Milfoil
Elodea
Milfoil
Near S. Lima, 'O'
g dry wt/ma
by species total
3.04
0.44
1.18
4.66
11.14
6.70
13.93
-
31.77
76.79
149.22
-
226.01
21.00
124.00
-
145.00
62.83
23.16
-
95.99
62.0
-
62.0
196.3
-
Near E. Avon, 'F'
g dry wt/m*
by species total
18.88
10.60
11.80
41.28
0.13
9.03
0.20
2.07
11.43
3.49
-
139.62
143.11
-
63.00
228.67
291.67
88.
369.98
370.86
-
387.35
387.35
-
316.8
Average total
kg
between 'O' and 'F'
(7559 m»)
173.6
163.3
1395.
1650.
1764.
1698.
196.3
316.8
1939.
-------
TABLE 17. ELBBBAL CONTEOT OF SUBCRGED AQUATIC PUNTS, UTTI£ CCNESUS CREEK (DRY WT. BASIS)
Date
12/17/7*
4/11/75
5/27/75
7/9/75
9/10/75
10/15/75
11/14/75
Species
Elodea
Ooontail
Pondweed
mg/m»
Elodea
Ooontail
Pondueed
Milfoil
ng/m*
Ooontail
Pondueed
Milfoil
ng/m»
Elodea
Ooontail
Milfoil
mg/ra*
Elodea
Ooontail
Milfoil
ma/"1
Elodea
Milfoil
ttg/m*
Elodea
Milfoil
mg/m1
Near S. Una,
Percent
N P Zn
3.45
1.43
122.
2.68
3.35
2.58
882
3.73
3.53
1.02
0.88
0.74
43.6
0.62
0.80
0.60
206
0.77
0.75
8132 1711
3.99
3.44
5104
3.68
3.08
3625
2.93
1817
2.48
4868
0.72
0.59
883
0.75
0.73
640
0.57
353
0.48
942
-
-
84
100
90
2.84
133
118
27.8
125
169
236
168
58
11. §
75
4.65
84
16.5
•0'
ppn
Cd Pb
-
••
1.0 13
1.0 10
1.0 13
6.032 0.3$
0.8 11
0.6 11
0.151 1.69
0.6 7
0.6 8
0.08? 1.14
0.6 11
0.5 7
6.649 6.8$
0.9 10
0.056 0.62
1.0 12
0.196 2.36
Near E.
PQEOGftt
N P
3.20
2.60
3.35
1275
2.39
2.13
260
3.48
4859
2.08
2.15
6227
2.64
2.39
8&3
1.85
7166
2.60
8237
1.02
0.65
1.10
391.
0.60
0.38
61.1
0.70
977.
0.45
0.51
1450
0.48
0.45
16"6*
0.36
1394
0.36
1140
Avon, 'F'
Zn
-
~
81
100
0.936
107
14.9
81
70
21.1
90
45
i6.1
65
25.2
68
21.5
ppn
Cd
™
~
0.8
1.0
0.009
0.07
0.098
1.6
1.6
0.467
0.7
1.5
0.556
1.9
0.736
1.3
0.412
Pb
"
«r
10
13
o.iis
12
1.68
13
12
3.56
8
2.97
12
4.65
13
4.12
Total grams between sites 'O' and 'f*
based on average dry wt. and composition
N P Zn Od Pb
5280 1643 -
4316 1013 14.4 0.155 1.92
49,099 10,159 161 0.941 12.7
42,826 8818 . 169 2.09 17.8
44.351 8727 108 2.29 14.4
33,951 6603 113 2.99 19.9
49,530 7869 144 2.30 24.5
-------
If a comparison is made between the total quantity of N and P in aquatic
plant tissue in Table 17 (1975) and the amount being transported downstream
at any specific time, it is obviously that much greater quantities were
contained in the tissue in late summer and fall (45 kg N, 8 kg P in July and
September) than were carried by the stream water (average of 0.06 kg N and
0.17 kg P per day during July, August, and September, Table 9). It seems
entirely possible that as the plants senesce in the water during late fall,
they may contribute a relatively large fraction of the total dissolved N and
P to the stream.
Of the heavy metals in Table 17, Zn is the only one which was
accumulated in quantity (0.16 kg in July) by aquatic plant tissue compared
to the quantity transported in the stream at the same period (0.001 kg per
day, Table 14). Cadmium and Pb do not seem to be accumulated by the plants
compared to N, P, and Zn, or the quantities moving down the creek.
Aquatic Plant Nutrient Dynamics
Phase one, four weeks: In the growth chamber experiments with milfoil
growing in 10-L jars, analysis of the sediment by extraction with 10% sodium
acetate at pH 4.8 (Morgan's solution) showed it to have adequate amounts of
the primary plant nutrients N, P, and K (Greweling and Peech, 1965). The
results of the entire analysis is presented in Table 18.
The P nutrient status of the soil did not change over the period of the
experiment, as indicated by constant phosphate levels in sediment water.
These did not vary appreciably either among treatments or over time. As
illustrated in Table 19, soluble phosphate levels were more than adequate
throughout the experiment for good root growth and phosphate absorption by
the milfoil plants (Bistow and Whitcombe, 1971).
TABLE 18. SEDIMENT CHARACTERISTICS DETERMINED AT THE BEGINNING OF THE EXPERI-
MENT. Nutrients were extracted with 10% sodium acetate at pH 4.8
(Morgan's solution) and are presented on a dry weight basis.
POM pH P K Mg Ca Mn Fe NO--N
(ppm) i-
9.1 7.0 3.5 37 387 3250 12 2.5 16
111-45
-------
TABLE 19. SOLUBLE P IN SEDIMENT OF INTERSTITIAL WATER, WHERE THE ROOTS
ACTUALLY GROW.
Control Jars Planted Jars
Date Mean Range Mean Range
May 15
April 29
June 3
June 16
June 19
_
0.826
1.05
0.958
1.120
0.583
0.944
0.819
1.000
*»
- 0.987
- 1.200
- 1.070
- 1.410
_
0.911
1.100
0.917
1.060
_
0.768 -
0.974 -
0.728 -
0.880 -
1.073
1.160
1.120
1.180
Inorganic P fractionation of the sediments before and after the experi-
ments according to the scheme of Chang and Jackson (1957) showed an increase
in iron phosphate with a corresponding decrease in adsorbed P. These
differences were a result of flooding, not the presence of growing milfoil.
These data are presented in Table 20, and further support the idea that
there was little change in the adequate supply of available P in the sedi-
ments during the experiments.
With the release of P upon rewetting of the fertile sediment at the
start of the experiment, the level of phosphate in the overlying water rose
rapidly to about 300 ppb (Figure 12). The subsequent decrease in P concen-
tration was more rapid in the planted jars compared to the unplanted con-
trols. This is probably a result of absorption by the growing milfoil
shoots (Bristow and Whitcombe, 1971). Within three weeks, P concentration
had fallen to similar levels in all jars and by the end of the month were at
5-10 ppb. In a similar control jar which had a 2-cm layer of washed white
quartz sand on top of the sediment, P concentration in the jar was still 180
ppb after three weeks (Figure 12). The sediments were obviously important
in removing P from the overlying solution, and were a bit slower than plants
in effecting this removal. It should be noted that both plants and sedi-
ments removed P to a very low level even when sediment-water P concentration
was 100 to 1000X that in the overlying water.
In the first part of the experiment, there was no evidence of leakage of
P from the sediment-water by virtue of the plants' presence. There were
good sources of P in the sediments and water initially, and it was not
111-46
-------
TABLE 20. SUMMARY OF INORGANIC P FRACTIONATICIN DATA FOR SEDIMENTS (Chang and
Jackson, 1957).
Sediment Control Jars Planted Jars
before after after
P fraction a b c mean 246 mean 1357 mean
adsorbed P
(.5 N NaHC03) 239 245 226 237 194 190 174 186 184 191 171 159 176
Fe-P 156 167 180 168 224 219 256 233 202 214 259 248 231
Al-P 147 153 149 150 146 143 135 141 137 141 132 129 135
occluded Al-P 3- 4 4 454 4 4- 44 4
Ca-P 204 399 313 305 245 346 443 345 296 335 260 285 294
occluded FeP 45 - 58 51 53 61 52 55 51 - 55 52 53
surprising to see no apparent leakage during this period of rapid plant
growth (Figure 13). During this phase the young plants grew vigorously,
producing auxiliary shoots and exhibiting no dead or dying leaves or tissue.
Therefore, when the cultures were sampled May 15, only live, photosynthetic
tissue was removed. A summary of the data for the sampled milfoil plants is
given in Table 21.
Percent P in all samples was high, indicating the presence in the jars
of excess, readily available P. Both percent P and total plant P increased
during this phase, as indicated in Figure 14.
Phase two, two weeks: The next experiment, involving the growth of
well-developed milfoil shoots in water with poor nutrient supplying
capacity, was started on May 15. By this time a filamentous algal mat had
formed over the sediment surface in all jars. Epiphytes on the milfoil were
not yet conspicous. Monitoring of phosphate concentration in the water
compartment showed no significant difference between controls and planted
cultures, with phosphorus levels equilibrating at about 10 ppb (Figure 12).
During the two weeks of growth in demineralized water, parts of the
milfoil plants began to exhibit signs of senescence. Typically, the termi-
nal segment of each shoot continued to photosynthesize and elongate, while
leaves yellowed and died. At the completion of this phase, nearly one-half
111-47
-------
400 -
PPb
P
200-
0
4/13
20 40
5/15 5/28
DAYS and DATES
60
6/19
FIGURE 12. Changes in MRP in the water of control and planted jars over the
experimental period. Note the high level after three weeks in
the jar with white sand. Each point represents the mean of four
jars.
of the nodes present in the jars had dead or senescing leaves, and sane
shoots had died. Data for the harvested tissues after two weeks on low P
water are presented in Table 22.
Growth during this phase continued in terms of elongation, but dry
weight as indicated by whole plant sampling decreased (Figure 13), probably
as a result of leaf senescence. Phosphate percentage and total P content of
tissues both decreased, but this P did not appear in the water. Even if P
was lost from senescing tissues, in this experimental setup it was not
detected. This is in contrast to leakage of P from leaves of milfoil
(Myriophyllum exalbescens) detected by Demarte and Hartman (1974). Not only
did the plants in the experiment reported here lose from their tissues
during this phase, but they were also unable to absorb enough P from the
high P root medium to maintain the ambient tissue concentration. This
111-48
-------
600
DRY
WEIGHT
400
200-
0ry Wtighf
wig
ppm P
total)
0
4/15
20 40
5/15
(Mrs and DATES
5/28
€0
200
LENGTH
cm
100
6/19
FIOTRE 13. Growth of plants in culture jars over the experiment, given in both
dry weight and total stem length. The dashed lines indicate
removal of senescent tissue at the beginning of the third phase.
Each point represents the total for one jar, averaged over four jars.
contrast to DeMarte and Hartman's data may be a result of differences in
experimental setup (their's was an axenic culture and they used T?) or
physiological differences in absorption characteristics between M.
exalbescens and M. spicatum. It appears that the plant M. spicatum used in
this experiment would not act as a pipeline for P movement from sediments to
overlying water to any extent during the normal growing season. This is in
agreement with findings by Brostgw and Whitcome (1971) for M. spicatum in
translocation experiments with T?.
Phase three, three weeks: The final experiment ran 25 days starting on
May 28 and ending June 22. After harvesting at the end of the second phase,
only live photosynthetic tissue remained in the culture. Four plants
remained in each of four jars at that time when 500 ppb P was added. The P
left the solution very rapidly (Figure 12) and again fell to a level of 5
111-49
-------
TABLE 21. SIMMS? OF DATA FOR PLANTS HARVESTED MAY 15, ONE MONTH AFTER PLANT-
ING. One plant (of six) was harvested fron each of four jars.
Tissue Stem Dry Percent P in
Jar harvested length weight P plant tissue
No. each jar (on) (ing) (ing)
1 one whole 22 59.5 1.0 0.60
3 plant of 35 85.5 0.85 0.73
5 the six 54 210.8 0.74 1.56
7 planted 38 144.0 0.94 1.35
mean 37 125.0 0.85 1.06
ppb after only two weeks. There was no difference in the rate of decrease
between planted and control jars. Again, the role of this particular
sediment seemed to dominate P removal, and adsorption was at a rate which
masked any leakage from plants if there was any way.
The plants grew in length and dry weight as indicated in Figure 13, and
epiphytes also appeared in appreciable quantity. The harvest data are
presented in Table 23 for milfoil.
Even though there was adequate P in both the overlying water and the
root medium in this third phase, the plants grew but continued to lose P
from the tissue. Total P in the milfoil tissue increased because of the 231
mg increase in dry weight.
An experimental summary is presented in Table 24. The plants grew well
and in a pattern that would be expected. When soluble P was present in the
water at concentrations above about 20 ppb, it was absorbed and utilized as
evidenced by growth in the first and third phase. When less than 10 ppb was
present however, the plants did not grow, but senesced and lost P from their
tissues, as in the second phase. This happened even when soluble P concen-
tration in the root medium was 1000 ppb. The P lost from plant tissues was
not measured in the water and was either transported to the roots, adsorbed
by the sediment particles, or reabsorbed in unsampled plant material such as
algae.
111-50
-------
% in Ephiphytes
Totol in EphiphytetA
"So
%
P
.4
4/15
20
40
5/19
DAYS and DATES
3/28
60
6/19
14. Percent and total P in plant tissue growing in culture jars.
Dashed lines indicate removal of senescent tissue at the beginning
of the third phase. Each point represents the total for one jar/
averaged over four jars.
Translocation to roots from leaves has been measured in Heteranthera
dubia and Myriophyllum brasilense (Funderburk and Lawrence, 1963) and in M.
exalbescens (DeMarte and Hartrnan, 1974), but not in M. spicatum (Bristow and
Whitcombe, 1971). Tliis argues against the first suggestion above and points
out another difference between M. spicatum and M. eyaltescens for P dynam-
ics. It is more likely that in this experiment, excess or leaked P was
adsorbed by sediment or absorbed by algae, as in the third phase.
The roots showed little or no capability to absorb P for translocation
to the plant tops, at least not when tissue concentrations were above 0.5%.
According to Gerloff and Krombholtz (1966) this is not a growth-limiting P
concentration, and indeed did not hinder milfoil growth in the third phase.
From Table -24, it is obvious that most of the added 500 ppb P was adsorbed
111-51
-------
TABLE 22. SUMMARY OF DATA FOR PLANTS HARVESTED MAY 28, TWD WEEKS AFTER PLACEMENT
IN DEMINERALIZED WATER
Jar
No.
1
3
5
7
mean
1
3
5
7
mean
1
3
5
7
mean
Tissue
harvested,
each jar
One whole
plant of
5 remaining
plants
Live tissue
only: subsample
removed for
analysis
All
senescent
material
remaining
Stem
length
(on)
27
37
50
75
47
16
5
14
32
17
119
71
133
173
124
Dry Percent P in
weight P plant tissue
(mg) (mg)
64.7 0.72
54.6 0.66
97.5 0.75
199.5 0.82
104.1 0.74
35.6 1.07
14.6 0.85
50.5 0.80
79.5 1.12
45.1 0.96
219.8 0.72
90.6 0.50
285.6 0.78
389.1 0.75
246.3 0.69
0.46
0.36
0.73
1.64
0.77
0.38
0.12
0.40
0.89
0.43
1.58
0.48
2.23
2.92
1.70
111-52
-------
TABLE 23. SUMMARY OF DATA FOR PLANTS HARVESTED JUNE 22 AFTER GROWffl IN
WATER.
Jar
No.
1
3
5
7
mean
1
3
5
7
mean
1
3
5
7
mean
Tissue
harvested,
each jar
One whole
plant of
four remain-
ing plants
Live tissue
only: subsample
removed from
analysis
Rest of
plant mater-
ial: three
plants plus
epiphytes
Stem
length
(on)
40
44
39
34
39
7
5
6
10
7
75
43
74
147
85
Dry
Weight
(rag)
63.9
93.6
119.2
78.7
88.9
23.3
15.6
17.1
34.1
22.5
137.6
87.9
205.2
319.1
187.4
Percent
P
0.60
0.53
0.38
0.55
0.52
0.68
0.39
0.42
0.63
0.43
0.60
0.46
0.32
0.60
0.50
P in
plant tissue
(mg)
0.39
0.50
0.46
0.44
0.46
0.16
0.061
0.071
0.213
0.12
0.82
0.40
0.65
1.92
0.93
by the sediments and what was left was quickly absorbed by the algae. If
algae had not been present, tissue P concentration for milfoil may have
increased again, as in the first phase.
The roots of M. spicatum may become more functional when the P concen-
tration in the tops falls below 0.5%, but from the data presented here, in
terms of P removal from overlying water in planted jars and decreases in P
concentration of tissues, it appears that M. spicatum depends upon its
leaves to absorb most of the P needed for normal growth, with little or no
leakage even when P concentration in the sediment water is 100 or 1000X
greater than in the overlying water from which the leaves must absorb.
111-53
-------
H
V
TABLE 24. SUMMARY OF ENTIRE EXPERIMENT, REPRESENTING THE MEANS OF FOUR JARS INCLUDING ALL PLANT
MATERIAL, NOT ONLY THE SAMPLES.
Date
April 15
May 15
May 28
June 19
Experimental
stage
Water added,
planted
Drained, sampled;
water added
Drained, sampled;
Live tissue sub-
sampled, added
0.5 ppm P
All harvested
Change between
5/28 - 6/19
Dry
weight
(mg)
109
50
625
520 total
125 alive
356 total
90 alive
(471 epiphytes)
-1-231 total
(+471)
Percent
P
0.60
0.85
0.85
0.74
0.96
0.52
0.53
(0.20)
-0.44
Mg P in
plant
material
0.66
6.38
5.31
3.85
1.20
1.85
0.48
(0.94)
+0.65
(+0.94)
Mg total P
control planted
3.11 3.36
0.05 0.12
0 0
0.1 0.1
5.16 5.21
0.04 0.04
-5.12 -5.17
-------
REFERENCES
Baker, J.L., K.L. Campbell, H.P. Johnson, and J.J. Hanway. 1975. Nitrate,
phosphorus, and sulfate in subsurface drainage water. J. Environ.
Quality 4:406-412.
Bouldin, D.R., R.L. Johnson, C. Burda, and C. Kao. 1974. Losses of
inorganic nitrogen from aquatic systems. J. Environ. Qual.
3:107-114.
Bristow, J.M. and M. Whitcotibe. 1971. The role of roots in the nutrition
of aquatic vascular plants. Am. J. Bot. 58:8-13.
Chang, S.C., and M.C. Jackson. 1957. Fractionation of soil phosphorus.
Soil Sci. 84:133-144.
32_ 59
DeMarte,.J.A. and R.T. Hartman. 1974. Studies on absorption of T>, Fe
and Ca by water milfoil (Myriophyllum exalbescens Fernald). Ecology
55:188-194.
Erickson, A.E. and B.G. Ellis. 1971. The nutrient content of drainage
water from agricultural land. Mich. State Univ. Res. Bull. 31.
Fox, R.L. and E.J. Kamprath. 1971. Adsorption and leaching of P in acid
organic soils and high organic matter sand. Soil Sci. Sec. Amer. Proc.
35:154-156.
Funderburk, H.W., Jr. and J.M. Lawrence. 1963. Absorption and
translocation of radioactive herbicides in submersed and emersed aquatic
weeds. Weed Res. 3:304-311.
Gerloff, G.C. and P.H. Krombholtz. 1966. Tissue analysis as a measure of
nutrient availability for the growth of angiosperm aquatic plants.
Limnol. and Oceanog. 11:529-537.
Greweling, H.T. and M. Peech. 1965. Chemical Soil Tests. Cornell Univ.
Agr. Expt. Sta. Bui. 960.
Hanway, J.J. and J.M. Lablen. 1974. Plant nutrient losses from tile outlet
terraces. J. Environ. Quality 3:351-356.
Hergert, G.W. 1975. Effects of manuring on phosphorus movement in soil.
Ph.D. thesis, Cornell Univ., Ithaca, N.Y. 14853.
111-55
-------
Hortinstine, C.C. and R.B. Forbes. 1972. Concentrations of nitrogen,
phosphorus, potassium and total soluble salts in soil solution samples
from fertilized and unfertilized histosols. J. Environ. Quality
1:446-449.
Ismirah, N.O. and D.R. Keeney. 1973. Contribution of developed and natural
marshland to surface and subsurface water quality. Technical Report on
project OWRR A-049-WIS, Water Resources Center, Univ. of Wisconsin,
Madison, WI 53706.
Johnson, A.H., D.R. Bouldin, E. Goyette, and A.M. Hedges. 1976.
Phosphorus loss by stream transport from a rural watershed: quantities,
processes, and sources. J. Environ. Qual. 5:148-157.
Kaila, A. 1959. Effect of superphosphate on the retention of phosphorus by
a peat soil. Maatalons Aik. 31:259-267.
Kaila, A. and H. Missila. 1956. Accumulation of fertilizer P in peat
soils. Maatalons Aik. 28:168-178.
Kitson, R.E. and M.G. Mellon. 1944. Colorimetric determination of phospho-
rus as molybdovanadophosphoric acid. Ind. and Eng. Chem. (Anal. Ed.)
16:379-383.
Larsen, J.E., G.F. Warren, and R. Langston. 1959. Effect of iron, aluminum
and humic acid on phosphorus fixation by organic soils. Proc. SSSA
23:438-440.
Lathwell, D.J., D.R. Bouldin, and E.A. Goyetta. 1973. Growth and chemical
composition of aquatic plants in twenty artificial wildlife marshes.
N.Y. Fish and Game Jour. 20:108-128.
MacLean, A.J., J.J. Jasmin, and R.L. Halstead. 1967. Effect of lime on
potato crops and on properties of a sphagnum peat soil. Can. J. Soil
Sci. 47:89-94.
Menzel, D.W. and N. Corwin. 1965. The measurement of total phosphorus in
seawater based on the liberation of organically bound fractions by
persulfate oxidation. Limnol. and Coeanog. 10:280-282.
Miller, M.H. 1974. The contribution of plant nutrients and pesticides from
agricultural lands to drainage water. Report of research conducted for
Environmental Protection Service, Environment Canada, under DSS Contract
No. OSR3-0031.
Murphy, J., and J.P. Riley. 1962. A modified single solution method for
the determination of phosphate in natural waters. Analytica chimica
Acta 27:31-36.
Nicholls, K.H. and H.R. MacCrimmon. 1974. Nutrients in subsurface and
runoff waters of the Holland marsh, Ontario. J. Environ. Quality
3:31-34.
111-56
-------
Okruszko, H., G.F. Warren, and G.E. Wilcox. 1962. The influence of Ca on P
availability in muck soil. Proc. SSSA 26:68-71.
Perverly, J.H., G. Miller, W.H. Brown, and R.L. Johnson. 1974. Aquatic
weed management in th Finger Lakes. Cornell Univ. Water Res. and Marine
Sciences Center Tech. Report No. 90.
Reference Manual of Methods for Chemical Analyses of Water and Wastes. U.S.
EPA-625/6-74-003, 1974, Washington, D.C.
Treunert, E., A. Wilhelms, and H. Bernhardt. 1974. Effect of the sampling
frequency on the determination of the annual phosphorus load of average
streams. Hodrochem. hydrogeol. Mitte 1:175-198.
111-57
-------
TECHNICAL REPORT DATA
(Please read Inunctions on the rtvenc before completing)
1. REPORT NO.
EPA-905/9-91-005C
2.
3. RECIPIENT'S ACCESSION NO.
4. TITLE AND SUBTITLE
Genesee River Watershed Study
Volume 3 - Special Studies - Rensselaer
Polytechnic Institute & Cornell University
6. REPORT DATE
. March 1978
6. PERFORMING ORGANIZATION CODE
. AUTHOR(S)
Hasson El-Baroudi
Thomas F. Zimmie
B. PERFORMING ORGANIZATION REPORT NO
John M. Ouxbury
John H. Peverly
9. PERFORMING ORGANIZATION NAME AND ADDRESS
New York State Department of Environmental
, . , . . Conservation
Bureau of Technical Services and Research
50 Wolf Road
Albany, New York 12233
10. PROGRAM ELEMENT NO.
A42B2A
11. CONTRACT/GRANT NO.
R005144
12. SPONSORING AGENCY NAME AND ADDRESS
Great Lakes National Program Office
U.S. Environmental Protection Agency
230 South Dearborn Street
Chicago, Illinois 60604
1J. 1 tr
-------