-------
Table 5. Listing of STORET, SAS and JCL commands required
to retrieve the data, create and save the SAS data set and print a
copy of the data matrix.
PGM=RET,PURP=205/STA,MORE=SAS,A=112WRD,8=08266000,3=08264970,
P=10,P=61,P=76,P=95,P=300,P=400,P=600,P=665,P=680,P=940,P=945,
P=70301,P=80155,PRT=NO NOECHO,
SASPARMS=BEGIN,
DATA OUTFILE.DALLAS;
I
I
l
I
I
I
I
INCLUDE (FCFREAD)
OPTIONS NOSOURCE? * SUPPRESSES SOURCE LISTING BELOW HERE ; .
RENAME P1=TEMP P2=CFS P3=TURB P4=COND P5=DO P6=PH P7=TN •
P8=TP P9=TOC P10=CHLORIDE P11=SULFATE P12=TDS ™
P13=SUSPSED;YEAR=YEAR(DATE);
IF STATION ='112WRD 08266000 ' THEN STATION = 'CABRESTO CR.;' •
IF STATION ='112WRD 08264970 ' THEN STATION = 'RED RIVER1; •
IF Rl NE '' THEN PI =
IF R2 NE 'l THEN P2 =
IF R3 NE '' THEN P3 =
IF R4 NE '' THEN P4 =
IF R5 NE '' THEN P5 =
IF R6 NS '' THEN P6 =
IF R7 NE '' THEN P7 =
IF R8 NE "' THEN P8 =
IF R9 NE '' THEN P9 =
IF RIO NE '' THEN P10 = .;
IF Rll NE '' THEN Pll = .;
IF R12 NE '' THEN P12 = .;
IF R13 NE '' THEN P13 = .;
DROP P14-P50 R14-R50 AGENCY BEGDATE BEGTIME ENDDATE DEPTH TIME SMK UMK
ENDTIME TYPE CALC NUMBER USGSREMK MORE;
PROC PRINT; •
TITLE1 STORET DATA IN SAS FORMAT;. •
STOPSAS,
./MXM JOB (A755STORP,MMXM),STORET,TIME=(,19}, •
./ MSGLEVEL=(1,1),PRTY=4 |
**ROUTE PRINT HOLD
**JOBPARM LINES=10 m
./OUTFILE DD DSN=MXMA755.SASDATA,DISP=OLD •
/*
PGM=RET is the STORET program that retrieves the data I
from the database and produces a file on disk (MORE=SAS). The •
resulting SAS formatted output file (DATA OUTFILE.DALLAS;) is
saved in a SAS data library for further processing. Note that •
the parameter's remark codes are examined and, if present, the £
corresponding parameter value is set to missing. This insures
that no data are included in the statistical analysis that are not _
strictly quantitative in nature (ie, remark codes k="less than •
value", and , j="estimated"). *
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Table 6. SAS statements necessary for the production of the
descriptive statistics.
PROC SORT DATA=DATA.DALLAS;BY STATION;
PROC MEANS DATA=DATA.DALLAS N MEAN STD MIN MAX SKEWNESS KURTOSIS;
BY STATION;
The dataset is sorted by station which allows this 'variable'
to be used in later procedures as a 'classifying' variable through
which the results are partitioned. The PROC MEANS procedure
produces the descriptive statistics which were desired.
Rather than accepting the default statistics, the specific
statistics of interest are specified.
-------
I
Table 7. SAS statements required to plot the data in the
desired format via the GPLOT procedure.
GOPTIONS CHARACTERS CELLS AXES DEVICE=TEK4662A SYMBOLS; I
SYMBOL1 C=GREEN V=PLUS;
SYMBOL2 C=RED V=DIAMOND;
TITLE1 Figure 1. Turbidity on the Red River and Cabresto Creek.;
PROC GPLOT UNIFORM DATA=DATA.DALLAS;
PLOT TURB*DATE=STATION / VAXIS=0.0 1.0 10.0 100.0 150.0
HAXIS='010CT78'D '010CT79'D 'OlOCTSO'D 'OlOCTSl'D '01OCT82'D
HMINOR=1;
LABEL TURB=TURBIDITY FTU;
TITLE1 Figure 2. Total nitrogen on the Red River and Cabresto Creek.; •
PLOT TN*DATE=STATION / VAXIS=0.0 1.02.03.04.0; •
LABEL TN=TN mg/1;
TITLE1 Figure 3. Total phosphorus on the Red River and Cabresto Creek. ;• •
PLOT TP*DAT£=STATION / VAXIS=0.0 0.05 0.10 0.15 0.20 0.25 0.30; •
LABEL TP=TP mg/1;
The GPLOT procedure produces output that can be displayed •
on a graphics device such as a terminal or on a plotter as is
indicated here (DEVICE=TEK4662A). The variable to be plotted is •
identified (TURB) and, the vertical (VAXIS) and the horizontal |
(HAXIS) are scaled as required. Note that the program is given
the specific dates desired in the "HAXIS=" statement. _
I
I
I
I
I
I
I
I
I
-------
Table 8. SAS statement necessary for the production of the
nonparametrie correlations.
PROC CORE DATA=DATA.DALLAS SPEARMAN NOSIMPLE OUTS=DATA.DALLAS4;
BY STATION;
The PROC CORR procedure produced the correlation
coefficients of interest. The SPEARMAN option was specified in
order to obtain only the rank-order correlations which were
desired; both the Pearson product-moment and Kendall's Tau-b are
also available. The NOSIMPLE option was specified to suppress the
calculation and printing of descriptive statistics which had been
produced by the previous procedure. An output dataset
(OUTS=DATA.DALLAS4) was created to store the data which could then
be used if other SAS procedures should be desired.
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
AN INTRODUCTION TO NORTH CAROLINA'S
BIOMOM70RING PROGRAM: BENTHIC MACROINVERTEBRATES
DEPARTMENT OF NATURAL RESOURCES AND
COMMUNITY DEVELOPMENT
DIVISION OF ENVIRONMENTAL MANAGEMENT
TECHNICAL SERVICES BRANCH
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I1
I
LIST OF TABLES AND FIGURES
Table 1. Taxa Richness of Benthic Macroinvertebr«te» from Seven
Hydrogeologic Regions in North Carolina
Table 2. Biological Classification Criteria: Total Taxa Richness (ST)
Table 3. Biological Classification Criteria: Ephemeroptera +
Plecoptera + Trichoptera (SEPT)
Table 4. Taxa Richness Values Collected at "Overlap" Sites, Summer 1983
Figure 1. Toxicity Testing by Bioassay and Stream Surveys in 18
Compar i sons
Figure 2. Statewide Water Quality Ratings
Figure 3. Statewide Biological Ratings
-------
-------
I
I
I
I
I
1
I
I
I
I
I
I
I
I
I
I
I
I
I
INTRODUCTION
North Carolina is divided into three major physiographic regions: the
coastal plain, piedmont and mountains. Each of these regions have distinct ben-
thic insect faunas. The coastal plain region, which extends inland for approxi-
mately 125 miles (Traver 1932), is characterized by large swamps and several nat-
ural lakes. Rivers form large, broad floodplains vegetated by bottomland hard-
woods or cypress, tupelo, and black gum (Brigham et al. 1982). The soils here
are mainly sandy loam, and there are many darkwater streams. Benke et al. (1984)
found that submerged wooden substrates (or snags) at sites in blackwater rivers
have high production estimates for lotic ecosystems.
From the western edge of the coastal plain to the foothills of the Blue
Ridge Mountains is the piedmont section of North Carolina. Traver (1932)
described this section of the state as having an elevation of 200-500 feet,
extending westward to a plateau attaining an elevation of 1200-1500 feet. The
soils in this section of the state are susceptible to the effects of weathering
and, in areas where the soils are disturbed, streams generally run very turbid.
This section of North Carolina is also the most highly industrialized and urban-
ized section of the state and many of our investigations are subject to the
effects of enrichment and/or toxicants.
The mountainous sections of the state are bordered on the east by the Blue
Ridge escarpment and on the west by the Great Smoky Mountains. Streams in this
section of the state generally are turbulent with higher concentrations of dis-
solved oxygen and lower water temperatures than streams in other sections of the
state. Many of our earlier investigations in the mountain region (Penrose and
Lenat, 1979) dealt with the effects of sedimentation to benthic macroinverte-
brates. These observations illustrated that if the sources of sedimentation were
controlled, the effects of sedimentation are generally short-lived.
-------
I
In unstressed streams in all regions of the slate we would expect taxe rich-
ness of benthic macroinvertebrates to remain high. For example, taxa richness |
values for major benthic invertebrate groups remained high from reference sites _
in several distinct hydrogeologic regions in North Carolina (Table 1). These
hydrogeologic regions, or zones, are roughly aligned west to east with Zone A the I
most western zone and Zone G the most eastern. The data in Table 1 are from
unstressed locations, mostly forested watersheds, and are the results from only •
one collection during the month of August. The single exception is from a sand-
hills stream in March. Taxa richness values for Ephemeroptera and Plecoptera •
were higher in Zones A, 6 and C, reflecting lower water temperatures and higher M
dissolved oxygen values for these mountainous streams. On the other hand, taxa
richness for other groups were higher in other zones. Specifically, Mollusca in 3
limestone rich coastal plain streams (Zone G) and Odonata in Zones F and G.
These diverse environmental conditions offer a unique challenge to aquatic m
biologists in several ways. First, a sample col-lection protocol has to be devel- m
oped that is flexible enough to permit the biologist to collect specimen from
many aquatic habitats (qualitative) and not have to rely on data from one habitat •
type (ie. artificial substrates, Surber). Habitat specific collection techniques
are limited. For example, do you need to know that species X is present at a |
density of 2000/m2 or that species X is present and abundant? Some qualitative
aspects of benthic community structures, especially taxa richness, are directly
related to water quality. A second challenge is to develop assessment criteria '•
relating benthic data to water pollution.
COLLECTION METHODOLOGY |
Qualitative sampling has been employed for many years to collect benthic
macroinvertebrates. Qualitative collections should be a useful tool for environ-
mental assessment. However, there is a need to determine if such qualitative •
sampling techniques will produce consistent and reliable results. Much of the
I
f
I
-------
I
following section has been taken from N.C. Biological Series *106 (N.C. Depart-
i
I
I
1
1
m
I
I
I
merit of Natural Resources and Community Development, 1983).
Biological monitoring groups often ere allocated a fixed amount of money and
personnel, but are assigned a variable work load. Therefore, their usefulness is
directly related to the time it takes to conduct a survey. A lab which utilizes
I
only quantitative sampling might produce excellent data on one investigation, but
might produce jLfi information on several other, equally important, investigations.
Qualitative sampling permits a greater work load.
This sampling methodology requires that a stream or river be wadable. High
water conditions may severely impair sampling efficiency by making critical habi-
tats inaccessible. An important decision in qualitative sampling is when not to
collect samples. Poor data is often worse than no data, as an underestimate of
taxa richness may lead to an incorrect assessment of water quality.
Many labs utilize a "timed" qualitative sampling technique. We feel it is
• better to process a fixed number of samples, usually 10. Different collectors
work at very different speeds, in part due to differences in the level of experi-
• ence. Also, the time necessary to collect a sample will vary. Collecting in a
targe river takes much more time than collecting in a temporary stream.
• The sampling technique outlined here usually takes 4-6 man hours, i.e. 1 1/2
— to 2 hours with three collectors. A sampling team can usually do 3 to 4 stations
" per day. Although more quantitative sites could be sampled in a day, the quali-
W tative technique produces an enormous savings in processing time. Quantitative
sampling is usually limited to a single habitat, and may underestimate species
richness. Allan (1975) found that 12 Surber samples underestimated total species
richness (from a variety of collection methods) by approximately 32%. Our lab
• has obtained similar results (unpublished data) in comparing total taxa richness
fi from qualitative sampling with taxa richness from 3 Hester-Dendy samples.
-------
I
a
i
it is critical that each collection team include one trained biologist. An
expectation of what should be found (and where) should guide the choice of sample
locations. Also, an untrained eye may not see many of the more cryptic insects.
Sampling Techniques
AM sampling equipment must be simple to use, durable and portable. The use I7
of glass and electrical equipment should be minimized. All samples are field
picked in white plastic or enamel trays. *•
Kick Net. A kick net is an easily constructed and versatile sampling __
I
device. It consists of little more than a piece of netting, or window screen, ^
between 2 poles. The net is positioned upright on the stream bed, while the area |jj
upstream is physically disrupted. Many investigators have found that this simple
technique gives very consistent results (Hornig and Pollard 1978, Armitage 1978). "•
If too coarse a mesh is used for the kick net, many animals will not be
retained. If too fine a mesh is employed, the net clogs easily and washout I
becomes a problem. We find that a good compromise is the use of a double layer •
of flexible, nonmetalic door screening. A border of durable cloth material is
used to reduce tears and weights are sewn into the bottom edge. •
Two kicks ere_ taken from riffle areas. The two samples should be collected
from areas of differing current speed. In very small steams, or in sandy areas |
lacking riffles, kicks should be taken from root masses, "snags" or bank areas. M
All types of benthic macroinvertebretes can be collected by this sampling device.
but emphasis is placed on Ephemeroptera, Plecoptera and Trichoptera. M
Sweep Net. A long-handled triangular, or 0-frame, sweep net is another
versatile sampling device. Samples (2-3) are taken by physically disrupting an
area and then vigorously sweeping through the disturbed area. Sweeps are usually
taken from bank areas and macrophyte beds. Bank samples are particularly impor-
tant for the collection of "edge" species which prefer low current environments. •
I
I
I
I
-------
I
Look for Chironomini (red chironomids), 01igochaetes, Odonata, mobile cased Tri-
I choptera, Hemiptera. ? i ft I i s. Crustacea and certain Ephemeroptera.
H A sweep net also can be used to sample gravel riffle areas where stonecased
Trichoptera may be abundant.
• Fine-Mesh Sampler. Since the kick and sweep nets utilize a relatively
coarse mesh size, an alternate sampling technique was devised to sample the
I.
smaller invertebrates (especially the Chironomidae). The resulting sampler is
— known in our lab as a "chironomid-getter". A cylinder (approximate diameter 7
™ cm) is cut from PVC pipe or a plastic bottle. Fine nitex mesh is attached to one
•[ end with glue and a ring clamp. The exact dimensions are not critical, but the
cylinder should fit inside another container, usually a one quart plastic con-
• tainer. This device can be used in a variety of ways.
^ The simplest technique is to wash down rocks or logs in a large plastic
• basin (or bucket) partially filled with water. Rocks are selected which have
• " visible growths of periphyton, Podostcmon or moss. Any large participate mate-
rial (leaves, etc) is washed down and discarded. A single composite sample can
• be made from several rocks and/or logs. The material remaining in the basin is
poured through the fine mesh sampler and the water allowed to drain out com-
P pletely. The residue is quickly preserved in 95% EtoH. This is accomplished by
• placing the fine mesh sampler into another container (see above) which is half
filled with alcohol. The sample is allowed to sit for several minutes and then
'• backwashed into a picking tray. Note that this method of field preservation
requires only a small amount of alcohol, and it may be reused many times. We
| usually bring 2 or 3 of the fine mesh samplers, so that one may be soaking while
• another is being picked.
Field preservation makes small chironomids and oligochaetes more visible,
• and easier to pick up with forceps. This technique is also good for Baetidae.
HydropiiIidae and other grazers. The "pour-and-preserve" technique also can be
I
I
-------
I
used in conjunction with other sampling methods. For example, the elutriate from
a kick or sweep sample can be possessed in this manner. It is also used in con- I
junction with sand samples (see below).
Sand Samples. Sandy habitats often contain a very distinct fauna, but fl
eitraction of this fauna by means of dredge-type sampling can be very tedious. To m
sample sandy substrate, we employ a large bag constructed of fine mesh (300
microns) nitex netting. It can be quickly constructed from a one meter square '•
piece of netting, folded in half and sewn together on the opposite side and the
bottom. This bag is employed like a Surber sampler, but the lack of a rigid |
frame allows for easy storage when folded. M
I
The bag is held (open) near the substrate, and the sand is vigorously dis-
turbed. The material collected (a lot of sand and a few organisms) is emptied I
into a large plastic container half-filled with water. A "stir and pour" elu-
trigtion technique is used in conjunction with the fine mesh sampler. After •
field preservation, look for Chironomitfae (especially flheosmi 11 i a. Har n i sch i a
group, Polvped iI urn spp ), Oligochaeta, Gomphidae and some Ephemeroptera.
Leaf-Pack Samples. Leaf-packs, sticks and small logs should be washed down;
a large bucket sieve is useful for this procedure. Leaf-pack and small log
samples are particularly useful in large sandy rivers. In such habitats, many of
the species are confined to "snags" (Benke et al. 1979, Neuswanger et al. 1982).
Look for "shredders", especially Tipulidae, Plecoptera and Trichoptera. •
Visual Search. Visual inspection of large rocks and logs, (the larger, the
shore (in negligible current) will harbor certain Ephemeroptera, and leaves near
the shore may be the primary habitat for some Gastropoda. In general, look for
I
i
I
better) often adds to the species list. Certain tightly adhering organisms may
be collected only by this technique. Decaying logs should be picked apart to •
look for chironimids, and many taxa can be found under loose bark. Rock near the
I
I
I
-------
I
I
attached cases of Trichoptera. Turbellaria, Coleoptera, Odonata (especially on
P? large logs), Gastropoda. Hirudinea and Megatoptera.
^ Mussel species can be obtained by careful visual inspection of the bottom. A
' mussel search should be conducted if dead shells are evident along the shore;
I look for midden heaps resulting from the feeding of muskrats and other verte-
brates. However, only live specimens should be added to the species list. Dur-
• ing periods of receding water levels, many species will move to deeper water,
. leaving a visible "track". The bases of aquatic weeds (especially water willow)
• may contain many mussel species and must be searched by hand. If possible, mus-
sels should be identified in the field and returned (alive) to the stream.
Sample Processing: Picking and Identification
• Field separation of invertebrates from large amounts of detritus ("picking")
£. can be difficult. However, it is infinitely less tedious than quantitative pro-
• cessing of samples in the lab. This is not a trivial point. The amount of
A -monotony and boredom involved in a task will probably effect both the level of
performance and, eventually, the turnover of personnel. Nobody likes to pick
• samples!
It is a simple matter to add some simple measures of abundance to this samp-
P ling scheme. As invertebrates are identified, we classify them as Abundant
(>10), Common (3-9) or Rare (1-2). Good field notes can also be used to assign
relative abundance values.
I BIOCLASSIFICATION CRITERIA
There ere several different ways to analyze qualitative data. Almost all
procedures require precise taxonomy.
M Comparisons of qualitative data from different sites are only useful if the
* taxonomy and collection techniques were similar. Erman (1881) found poor agree-
I ment between several "baseline" studies when identification and collection meth-
ods were not consistent.
I
I
I
-------
I
Analysis of fauna) assemblages is one way to detect water quality problems.
nities. The taxa associated with organic loading (and tow dissolved oxygen) are ^
well known. More recent studies have begun to identify taxa associated with sed- *
•mentation and toxic stress (Winner et al. 197S, Bode and Simpson 1982). However. •
identification at, or near, the species level is desirable for many genera,
including Pol voed i lutn. Cr icotoous . Hvdroasvche. Ephemere I la. jjtqnonema and •
BaetIs. Note that tolerant species are present in all aquatic habitats (see
Hynes 1960). However, these species will usually become dominant only in pol- •
luted systems. Allowances must also be made for stream size, geographic varia- •
t ion and seasonaIi ty.
The presence of rare or endangered species is often associated with good •
water quality. Again, species level taxonomy will usually be required (Resh and ^
Unzicker 1975).
The simplest method of analyses is the tabulation of species richness.
Species richness is the simplest measure of "diversity", and the only one that
fits a dictionary definition. The association of good water quality with high
species (or taxa) richness is intuitively obvious even to the non-biologist.
Increasing levels of pollution gradually eliminate the more sensitive species, m
leading to lower and lower species richness.
I
i
A primary motivation for quantitative studies has been the calculation of
diversity indices, especially the Shannon-We iner index (Wilhm 1970). While this B
index functions well under conditions of simple organic loading, it often pro-
duces confusing results for other kinds of stress (Hocutt 1975, Mason 1975,
Hughes 1978. Godfrey 1978, Oden 1979. Statzner 1981). Taxa richness values ere
usually more easily related to water quality changes (Dills and Rogers 1974, Win-
ner 1975, Green 1977. Bournaud and Keck 1980). •
I
I
I
I
-------
I
i
i
I
t
i
i
i
t
i
i
I
i
i
i
i
i
i
i
Total taxa richness for unstressed streams and rivers is relatively constant
both temporally and spatially if comparisons are limited to streams of similar
size (Lenat 1963. Patrick 1975). Any differences that occur are predictable.
Seasonally, a maximum is likely to occur in early spring, as over wintering
species which emerge in spring overlap with spring-hatching species. However,
collections from the French Broad River (Penrose et al. 1963) suggest that some
seasonal changes in taxa richness are not inherent, but reflect seasonal changes
in water qua!ity.
Tables 2 and 3 list taxa richness ranges describing water quality conditions
for several major ecoregions in North Carolina. Table 2 is for total taxa rich-
ness and Table 3 is for taxa richness ranges for intolerant species groups (EPT =
Ephemeroptera + Plecoptera + Trichoptera).
We have also conducted overlap studies from which two teams conduct surveys
at a similar location (see Table O. These studies serve as a form of field
quality assurance. Good agreement was obtained al all three locations at which
surveys were conducted during the summer of 1983, particularly for SEPT values.
BIOMONITORING PROJECTS
Several biomonitor ing projects using qualitative collection methods for ben-
thic macroinvertebrates in North Carolina include trend monitoring (Benthic
Macroinvertebrate Ambient Network, or BMAN), instream toxicity testing and
impaired use assessment. Each of these monitoring projects will be briefly dis-
cussed.
The Benthic Macroinvertebrate Ambient Network (BMAN) consists of a subset
(160 sites) of all statewide ambient water quality locations (349). Benthic
mac rotnvertebrates are collections which are staggered from selected sites each
year, for example, data from many locations are collected each year, particu-
larly from sites monitoring specific problem industries or interstate water
quality. However, data is collected from most locations on a 2 or 3 year rota-
-------
I
tion. These rotations allow us to conduct surveys on a much broader statewide
scale without losing much resolution between years. The objectives of the BMAN •
network are 1) determine long term trends in water quality as reflected in taxe
richness and 2) suggest specific locations for intensive investigations. V
Results of the BMAN collections (North Carolina Department of Natural ti
m
Resources and Community Development 1985a) are discussed on a basin by basin for-
mat and include notes on the presence of pollution tolerant taxa (species), wat- •
ershed characteristics and suspected pollution sources. In 1964, seventy-five
I
locations were assigned bioclassifications based on taxa richness criteria (clas-
sification criteria for estuarine locations were provisional). These results are
Iisted below:
B j Qc J ass i f Jcat ion t of Locat ions S of Total
Excellent 6 8
Good 16 21
Good/Fair 30 40
Fair 18 24
Poor 5 7
Benthic macroinvertebrate collections were made tor the first time during
1984 at 22 locations. Data at 28 locations, having two years of data, suggest a
positive trend in water quality at 4 stations and no observable trend at 24 sta-
tions. No negative trends in water quality were noted at stations having two •
years of data.
Data from 25 locations, from which data has been collected for three conse-
cutive years, suggest a positive trend in water quality at 4 stations and no •
observable trends at 21 stations. No negative trends in water quality were noted
at stations having three years of data. I
Data from 97 stations were collected during the 1985 BMAN program.
I
I
f
I
I
I
1
I
I
-------
I
A second monitoring program has been established to supplement toxicity
• testing of municipal or industrial effluents. At facilities either suspected or
m known to lie discharging a toxic effluent (by using toxicity screening procedures)
* benthic macroinvertebrates are collected above and below discharge points to
• assess instream effects. Changes in benthic community composition as well as the
development of a toxic assemblage is used to assess instream effects. Thirty
P such investigations have been conducted. Frequency of agreement between the
I first 18 investigations are illustrated in Figure 1.
Twelve investigations (Column A) illustrate agreement acute between effluent
fl toxicity predictions using fathead minnows, (Pimephales promelasl and instream
biological effects.
• At four of the eighteen facilities (Column B), fathead minnow tests were
m negative even during the 100* effluent test. However, toxicity was detected with
* the benthos. Since effluent samples were taken above disinfection, the results
tt may indicate chlorine toxicity. In each of these instances. The more sensitive
Ceri odaohnia chronic tests were positive (Column C).
• In three of the eighteen tests bioassay results suggested instream toxicity
but it was not detected with benthos (Column D). However, in two of the three
9 tests, poor upstream water quality masked the instream toxic effects. Benthic
community structure suggested toxic conditions both above and below the facility.
In the other of these three facilities, the effluent didn't reach the stream.
• A final document assesses surface water quality statewide by reviewing
existing data sources (North Carolina Department of Natural Resources and Commu-
• nity Development 1985b). These sources include fishery investigations and water
^ quality studies in addition to benthological studies. The final result is a
series of water basin maps with streams in each basin color coded as to the water
• quality. In addition, differences were noted between biological surveys and
chemical surveys. Figures 2 and 3 illustrate results of water quality rating and
I
I
I
-------
I
biological ratings. Large discrepancies occur between these ratings due largely
to the variable effects of sedimentation and/or complex organic compounds which M
are not analyzed for in most chemical indices and are accounted for in biological
investigations. This report illustrates where there are specific point and non- ™
point source pollution problems and provides a foundation for review and improve- •
ment of the state's water quality management plan.
I
I
I
f
I
I
t
i
I
i
i
i
-------
I
I
1
I
I
t
1
I
I
I
I
I
I
1
I
I
I
I
I
LIST OF REFERENCES
Allan. J.D. 1975. The distributional ecology and diversity of benthic insects
in Cement Creek, Colorado. Ecology 56: 1040-1053.
Armitage, P.O. 1976. Downstream changes in the composition, numbers and biomass
of bottom fauna in the Tees below Cow Green Reservoir and in unregulated
tributary Maize Beck, in the first five years after impoundment. Hydro-
biologia 58; 145-156.
Benke, A.C., T.C. Van Arsdall Jr., and D.M. Gitlespie. 1964. Invertebrate pro-
ductivity in a subtropical blackwater river: The importance of Habitat and
life history. Ecological Monographs. 54(1)25-63.
Benke, A.C., D.M. Gillespie, F.K. Parrish, T.C. Van Arsdall, Jr., R.J. Hunter and
R.L. Henry, III. 1979. Biological basis for assessing impacts of channel
modification: invertebrate production, drift, and fish feeding in a sou-
theastern blackwater river. Georgia Institute Technology, Earch Resources
Center, ERC 06-79. 187 pp.
Bode, R.W. and K.W. Simpson. 1982. Communities in large lotic systems:
Impacted vs. unimpacted. Abstract, Thirtieth Annual Meeting, North American
Benthologieft I Society.
Bournaud, M. and G. Keck. 1980. Diversity specifique et structure des peuple-
ments de macroinvertebr6s benthiques au long d'un cours d'eeu: le Furans
(AinD. Acta Oecologia/Oecologia Gener. 1:131-150.
Brigham. A.R., W.U. Brigham and A. Gnilka. 1982. Aquatic insects and Oligo-
chaetes of North and South Carolina. Midwest Aquatic Enterprises. Mahomet,
I I I inois 61853.
Oitls, G. and D.T. Rogers, Jr. 1974. Macroinvertebrate community structure as
an indicator of acid mine pollution. Environ. Pollut. 6:239-262.
Erman, D.C. 1981. Stream macroinvertebrate baseline surveys: a comparative
analysis from the oil-shale regions of Colorado, U.S.A. Environmental Man-
agement 5: 531-536.
Godfrey, J.J. 1978. Diversity as a measure of benthic macroinvertebrate commu-
nity response to water pollution. Hydrobiologia 57: 111-122.
Green, R.H. 1977. Some methods for hypothesis testing and analysts with bio-
logical monitoring data. Jn: J. Cairns, Jr., K.I. Dickson and G.F. West-
lake teds.). Biological monitoring of water and effluent quality. ASTM STP
607, Amer. Soc. Testing & Materials, p. 200-211.
Harrell, R.C. and T.C. Dorris. 1966. Stream order, morphemetry, physico-
chemical conditions and community structure of benthic macroinvertebrates in
an intermittent stream system. Amer. Midi. Nat. 80: 220-251.
Hocutt, C.H. 1975. Assessment of a stressed macroinvertebrate community. Water
Res. Bull. 11: 820-835.
-------
Horntg, C.E. and J.E. Pollard. 1978. Macroinvertebrate sampling techniques for
streams in semi-arid regions. Comparison of the Surber method with a
unit-effort traveling kick method. EPA-600/4-78-040 28 pp.
Hughes, B.O. 1978. The influence of factors other than pollution on the value
of Shannon's diversity index for bcnthic macroinvertebrates in streams.
Water Res. 12: 359-364.
Hynes, K.B.N. 1960. The biology of polluted waters. Univ. Toronto Press.
202 pp.
Lenat, D.R. 1983. Benthic macroinvertcbretes of Cane Creek,
comparison with other southeastern streams. Brimleyana
Nor th Carolina
9: 53-69.
and
Mason, W.T., Jr. 1975. Chironomidae (Diptera) as biological indicators of water
quality. Jn; Organisms and biological communities as indicators of envi-
ronmental quality. Ohio St. Univ. p. 40-51.
Neuswanger, D.J., W.W. Taylor and J.B. Reynolds. 1982. Comparison of macroin-
vertebrate herptobenthos and haptobenthos in a side channel and slough in
the Upper Mississippi River. Freshwater Invertebr. Biology. 1 (3) 13-24.
North Carolina Department of Natural Resources and Community Development. 1983a.
Qualitative sampling of benthic macroinvertebrates: A reliable, cost-
effective, biomonitor ing technique. Biological Series *108. 11 pp.
North Carolina Department of Natural Resources and Community Development. 1985b.
Benthic macroinvertebrate ambient network (BMAN) data review, 1984. 107 pp.
North Carolina Department of Natural Resources and Community Development.' 1985.
Assessment of surface water quality in North Carolina. 259 pp.
Oden, B.J. 1979. The freshwater
receiving thermal effluents.
littoral meiofauna in a South Carolina reservoir
Freshwat. Biol. 9: 291-304.
Patrick, R. 1975. Chapter 15: Stream Communities. Jjj.: M.S. Cody end J.M.
Diamond
-------
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
Traver, J.H. 1932. Mayflies of North Carolina. Journal of the Elisha Mitchell
Scientific Society. 47(1) 85-161.
Wilhm. J.L. 1970. Range of diversity index in benthic macroiovertebrate popu-
lations. J. Water Poll. Control Fed. 42: R221.
Winner, R.W., M.W. Boesel and M.P. Parrel I. 1975. Response of the macroinver-
tebrate fauna to a copper gradient in an experimentally polluted stream.
Verh. Internat. Verein Limnol. 19: 2121-2127.
-------
(3
H
O
M
•J O
o
W
8
PS
f^| JSSd
s
0 *
OS U
(H
CO
i •
w
H
OH
W O
£
i_l
O
OS
u ^ «
< CO
S Z
0
0 »H
M H
35 0 <
I 3
f* t— J
^H ^ CM o «n
«-l ^^ i-H
CM »A CO ON vO
p-» 1-4
f^ ON ON P^> SO
i— 1
«n . co
»H i-H
CM in r~ ON vo
CM ^H
CM ON r-
Du * *
OH w
co w
w & a -t
ON co —i m r-
CM
m ^H
i 4
d -
o •
I
a •
s 1
W
W * ^
1
1
1
I
1
-------
1
I
I
I
I
t
i
i
i
t
i
i
i
i
i
f
i
i
i
Table 2. Biological classification Criteria: Total Taxa
Richness (S ).
Excellent
Good
Good-Fair
Fair
Poor
Mountain
Rivers
>91
77-91
61-76
46-60
0-45
Piedmont
Rivers
>91
77-91
61-76
46-60
0-45
Coastal A1
Rivers
>84
68-83
52-67
36-51
0-35
Coastal B2
Rivers
?
>60
46-60
31-45
0-30
Shallow, fast-moving
2 Deep, slow-moving, criteria provisional
-------
Table 3. Biological Classification Criteria:
Epheroeroptera + Plecoptera + Trichoptera (S£pT)
>EPT
Mountain Piedmont
Rivers Rivers
Coastal A1 Coastal B'
Excellent
Good
Good-Fair
Fair
Poor
32-41
22-31
12-21
0-11
24-31
16-23
8-15
0-7
Rivers
>27
21-27
14-20
7-13
0-6
Shallow, fast-moving
Deep, slow-moving, criteria 'provisional
Rivers
>11
9-11
6-8
3-5
0-2
1
I
i
I
i
i
i
i
i
t
i
i
i
i
t
i
t
i
-------
I
I
I
i
i
t
i
i
i
t
i
i
i
i
i
i
i
i
i
to
V
•H
CA
«0
•o
o
0)
CO
p*
o w
H H
G
W
H
O
O
CO
M
CO
CO
CO
CO
w
_ - W
O O H
M H PM
: • i
X H H
3 co
ff!
U
>
O
„ S
W H
en to
>J O «
pi
W H
> z
M O
Ptf £
Q
3 W
CQ
CM
GO
<»>
\D
CM O
H
Oi
2
-------
Figure 1. Toxicity Testing by Bioassay and Stream Surveys
in 18 Comparisons.
Toxicity as Detected by Bioassay and Stream Surveys
in 18 Comparisons
Frequency
16
12 -
8 ~
4 -
A: Flow-Thru S Strean Survey Agree
B. Toxicity nissed by Flow-Thru
C. Toxicity Hissed by Flow-Thru and Chronic
0. Toxicity nissed fay Strean Survey
B
Class
-------
I
I
i
I
i
i
i
I
i
t
i
i
i
i
i
i
i
i
i
Figure 2. Statewide Water Quality Indices
STATEWIDE WATER QUALITY INDICES
Fair 6.4% Poor 5.8%
Good-Fair 14.
% of state's total stream mileage.
-------
Figure 3. Statewide Biological Ratings
STATEWIDE BIOLOGICAL RATINGS
Poor 6.7% Excellent 4.07'
Good 16.9%
-Fair 42.9%
% of state's total stream mileage.
I
1
I
I
I
I
I
i
I
t
i
i
i
i
i
i
i
t
i
-------
I
I
I
I
I
1
I
i
i
m BAYOU BUN IDEE WATER QUALITY MONITORING PROJECT
Louis K. C. JOHNSON - LOUISIANA DEPARTMENT OF ENVIRONMENTAL QUALITY
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
1
I
I
1
I
i
i
i
i
i
i
I
i
i
i
-------
BAYOU BONNE IDEE WATER QUALITY MONITORING PROJECT
Presented by
Louis R. C. Johnson
Environmental Program Specialist V
INTRODUCTION
I
I
I
I
I
I
I
The Bayou Bonne Idee Monitoring Project is located in Northeast Louisiana
• completely within Morehouse Parish. The project was originated to monitor the water
quality of Bayou Bonne Idee as the result of a Rural Clean Water Project. Funding for the
I water quality monitoring project is from the Environmental Protection Agency under
« Section 208 of the Federal Water Pollution Control Act (PL 92-500). The Rural Clean
* Water Project was funded because Bayou Bonne Idee was highly affected by agricultural
I runoff.
Agricultural runoff is the leading nonpoint pollution problem in Louisiana. As the
| result of effective weed control programs, fall tillage, skip-row planting, and continuous
— plowing of turnrows and field borders - turbidity, suspended solids, nutrients, and
™ agricultural chemicals had reduced the usability of the bayou. Toxaphene levels in whole
• body samples of fish were recorded as high as 45 mg/1.
The RCWP originally encompassed 220,000 acres and three different watersheds.
• Because of funding problems, the RCWP was revised twice in 1983. The first time, the
area draining into Cypress Bayou was removed from the project. The second reduction
» removed the Bayou Gallon drainage leaving only 66,000 acres adjacent to Bayou Bonne
• Idee in the project. Likewise, the water quality monitoring sites were also relocated to be
* only on Bayou Bonne Idee.
I
-------
The land involved is Mississippi delta farmland with predominate crops being cotton,
soybeans, and rice. According to the U.S. Soil Conservation Service, W,880 acres of
66,000 acres in the RCWP are considered critical erosion acres. The goal of the RCWP is
to have some type of agricultural BMPs on 75 percent of the critical acres. The cost of
the implementation of the BMP's is approximately 3.9 million dollars. Seventeen different
types of BMP's are being implemented on the agricultural lands within the RCWP.
Bayou Bonne Idee is presently a slow moving body of water dotted with cypress
trees. The bayou contains three dams with weirs and drawdown structures. Behind each
dam and weir a lake is formed. The major drainage area is at the upper end (north) of the
bayou. For most of the length of the bayou the highest elevations are near the edge of
the bayou. Bayou Bonne Idee discharges into Beouf River, a major waterway in Northeast
Louisiana.
PROJECT DESCRIPTION
The water quality monitoring program consists of monthly sampling at four (4) sites
located on Bayou Bonne Idee and one (1) site located on a tributary to the bayou. Samples
of both water and sediment are taken monthly. Flow measurements are taken at one of
the sampling sites monthly. Fish samples have been taken twice per year from each of
the three lakes formed by the weirs.
Field analyses consist of dissolved oxygen, temperature, pH, and conductivity
measurements at mid stream. Laboratory analyses consist of two types. Type one is the
normal 25 parameter scan done as part of the DEQ/OWR/WPC monthly Water Quality
Program. The second type consists of a scan for the 26 pesticides found in the 129
priority pollutant list. These scans were done on both water and sediment taken from
each sampling site. Field analyses and samples were taken by the Department of
I
I
I
I
I
I
i
i
i
i
i
i
i
i
i
i
i
i
i
-------
t
I
I
Environmental Quality's Office of Water Resources Northeast Regional Staff. The
laboratory analyses were done by the Department of Environmental Quality's Office of
Water Resources Laboratory on the LSI) campus in Baton Rouge and the Northeast
Louisiana University Soil Laboratory in Monroe, Louisiana. Data resulting from the
project is now stored in the DEQ/OWR Vax 780 digital computer located in Baton Rouge.
This data, after review, will be transferred to the EPA STORET system.
DATA REVIEW
I
I
I
I
i
i
i
At the time of writing this report, 607 samples had been collected. This represents
• 30,350 different chemical analyses and rain measurements, of which only 36 are rain
measurements. This data represents data collected from February 1982 thorugh May
j§ 1985. Even though there has been a large amount of chemical anlayses done, the period of
§ record is small. This short record period makes trend analyses difficult for showing
whether the water quality in Bayou Bonne Idee is improving as a result of the RCWP.
ft Pesticide and water quality data is very scattered on plots. Bar charts of the lower
station versus the upper station for six (6) pesticides shows pesticides found more times at
j| the lower station in both water and sediment than at the upper station. The pesticides
— found present in water and sediment samples most of the time are DDT, DDE, ODD,
™ toxaphene, aldrin, and deldrin. The highest concentrations of these are found in the
ft sediment samples.
Turbidity at one station runs from 36 NTU to 3000 NTU. Turbidity plots are high
one month and low the next. Review of plotted turbidity data for the upper end station
(127) versus the lower end station (122) shows the concentrations higher at the lowest
station. But, the turbidity data for station 128 contains very high results. This station is
located on a tributary just one quarter mile south of the northernmost station and seems
-------
I
I
to be the largest source of pollutants to the bayou.
Rainfall has been a factor during this monitoring program. The years 1982 and 1983 • .
were wet, while 1984 and 1985 were somewhat dry. No flow was recorded at the flow ||
measuring site during the monitoring period due to lack of rainfall on numerous occasions. m
I
OTHER FACTORS AFFECTING THE PROJECT •
The level of the bayou when the project began was low. The dams were cut, and the ||
bayou was only a channel. Approximately six months after the monitoring began, the cuts ||
in the dams were filled, and the lakes began to fill. At the end of the project, the lower
lake was lowered in order to build a new bridge. •
During the life of this project, two people primarily responsible for the sampling
were injured on the job - both due to back injury. Because of their injuries, supervisory £ II
personnel had to take over the sampling program. A permanent replacement did not ^
take over until the project had only four months remaining. ™
I
SUMMARY
I
I
After reviewing the data and the scope of the RCWP, this water quality monitoring
project on Bayou Bonne Idee was too short. In order to properly evaluate a RCWP of this M
size and duration, the water quality monitoring program should last at least the length of
the RCWP, in this case, 15 years. •
I
I
I
-------
-
"' " "x-f
;/ i t
FIGURE 3
KONITORDG STATICS
BAYOU BOWE IDEE
RURAL CLEAN WATER PROJECT
HOREHOUSE PARISH. LOUlSL*uNA
O - Monthly Water Quality
-------
24 -
22 -
20 -
^
c IS -
6
u 16 -
E
E 1 J -
H-
•5 1-2-
ci in -
jj
E
T i7j --
^
e -
4 -
O -
Station
/
y.
,>
,.'
t'
<•'
i
.1
, •'
i1'
ji
.1
y'
>
jf
/'
S3
H t><
ri
3-
1 PI
CM
>
H
K
f
CO
id
f
CO
?i
m
/•
,•
•-
.»•"'
,••'
•-•;;
..•
.*'
„
,-•'
.r
.•
y
i i i
0 C3 0
000
HMO
127
H
J
ri in
71
y
i
>o
t- M
O Pl
to r
M O
z
?0
M
Z
"
T
,»'
/
/
f.'
/
/
..-'
/
,..•
y
*"
.1'''
,-•
f
,-•''
.«'
..-•
«_
P]
••'•
/
;,
''<,
/
/
f
's
.'
,.-'
,••'
r..'
,
.''
.••''
t'-
••'
,.••
,i
.*'
>
/
/
f.-
r./
/
,.•'
/
t.x
,
'"
/'
,•
V
t--'
'•1
'i
n
.<
/
/
y.
y
/
/"
r'*
fy
r*
•••
,.-•'
.r
~r
P""1
0ii
Station 122
"r
r^
V
/
.
/
/
f-
/
/
,.'
,.''
f/
_,.
'"
.•"'
/
/'
..'
\
•
|
17-1
•7
,-•*
,.
/
s
/
,.•'
/
/
' ,
p'
y
«•'
/
,.'
V
y
,•"
/
r*
,-•'
71
1;
..-•
',
^
/
,.i
.,••
/
^
/
/
f'
/
y
>
•'
.•'
.-•'
.r'
f.-'
...'
rf
7"
...-'
,.•
-•'"
X
r-
/
^i
/
/
f-
/
.-'
/
/
,/
y
'
/
,•1
,«'"
^V
.,<
./
/
^
••'
/
>
/
/
y-'
/
''
/
/
/
/
•'
9 i
/
.'
/
,«
/
/'
/
'./
/
,-•''
,1
.•i
,4
•i
' fk1 f
,-••]
;;
/i
.••i nn rr
-J ,--U-
k'fi/u/ii.-1
.•*1l A I A \ A
.*i
•jnn
••JKIK
J1 I~I1 1 1 1 1 1 ~l ~l 1
HHOOO>C H OOO>O HHOOO>C
OOOOCt-'M O OOOfM OOOOOfM
HXHP3OOKI H HPJOCR) HXHPlOOK
>> jc r >
r TJ MO r
:r z PO
t tn MI
z z
co PJ C
K >
O H
M M
K ?0
p:
Z CO
Crt >X3
<5 r1
r*
S 5 P% Sb
Z !» rr z pc
M 1 PJ M
Z Z Z
co m
K
o
M
3
re
z
CO
g
r
M
CO
Pesticide Analysis - Water and Sediment - Stations 127 & 122
I
I
I
I
I
I
I
i
i
I
i
i
i
i
i
i
i
i
i
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
•6
II
^^\f
-------
5 i
g a
>i -
pa
u>
09
(0
CD
CO
•
CO
|
a
u.
u>
CO
(SI
o
I
I
I
I
I
i
i
i
i
8 5
Q
"
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
I
w
o
155
u
§
o.
y
PQ
CJ
(VJ
00
• u.
oo
a
§
UJ
eo »-
oo <
8
O
UJ
MM
• 00
-------
%y
»
CQ
I
u>
U>
u>
CM
U)
00
00
tD
00
u_
CO
I
I
I
I
I
I
i
Ul
Z> 03
- O
00
o
I
I
I
I
I
I
i
i
i
i
-------
~o
PQ v> m
£> I ?
O
CM
CSJ
(NJ
W
00
L/5
CM
LA
II I I
V)
00
I
i
oo
CD
i
00
CD
UJ
U.
CO >-
CO <
^ -3
CNJ
CO
O
UJ
O
I
-3
rs
LO
OJ
-------
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
u
-------
™
PQ
PQ
S
i
CO
•8
I
I
I
I
I
I
I
i
i
i
i
i
i
i
i
i
i
i
i
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
BAYOU BONNE IDEE
MONTH
Ja nuary
February
March
Apri 1
May
June
July
August
September
Octomber
November
December
January
February
March
April
May
June
July
September
Octomber
November
December
January
February
March
Apri 1
Kay
June
July
August
September
Octomber
November
December
WATER QUALITY MONITORING PROGRAM
RAINFALL DATA
ATTACHMENT SIX
YEAR
1982
1982
1982
1982
1982
1982
1982
1982
1982
1982
1982
1982
1983
1983
1983
1983
1983
1983
1983
1983
1983
1983
1983
1984
1984
1984
1984
1984
1984
1984
1984
1984
1984
1984
1984
RAINFALL
6,
4,
2,
5,
1,
6,
2,
5
3,
10,
7
2
10
2
6
11
5
1
2
1
8
10
3
7
9
6
3
7
1
6
2
7
5
34"
37"
67"
29"
88"
81"
46"
21"
00"
27"
36"
18.71
,00"
16"
,35"
,05"
,19"
,01"
,67"
,14"
,04"
,33"
,15"
.88"
.93"
,21"
.22"
,00"
.97"
.70"
.09"
.09"
,77"
.73"
1.68"
-------
BAYOU BONNE IDEE WATER QUALITY MONITORING PROJECT
FISH TISSUE RESULTS AND STATISTICS
IN PPB
ATTACHMENT FIVE
DATE SAMPLED
09/22/79
08/20/79
08/20/79
03/19/80
03/19/80
09/22/79
01/20/84
01/20/84
01/20/84
01/20/84
01/20/84
01/20/84
01/20/84
01/20/84
01/20/84
01/20/84
01/20/84
01/20/84
10/25/83
10/25/83
DDT
300.00
.65
1.07
2.35
.26
300.00
85.21
127.62
32.35
35.28
255.65
10.56
201.23
18.05
230.78
143.15
866.52
1016.83
.00
.00
ODD
.00
.68
.71
4.17
.56
.00
288.15
179.49
44.36
46.22
255.65
10.91
269.09
36.02
410.18
249.52
1398.30
1825.59
16.15
8.54
ODE
3321.00
5.20
3.32
13.26
1.50
3321.00
761.62
319.80
272.38
366.43
1115.38
101.70
546.92
344.56
470.53
821.92
4715.00
1726.61
95.47
50.48
TOXAPHENE
1439
1
2
3
1439
.00
.58
.98
.27
.18
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
.00
I
I
I
I
I
I
I
i
I
f
I
I
I
I
I
i
i
i
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
TEXAS WATER CUIfllSSlUN FISH KILL REPURT1NG SYSTEM
PATRICK ROQUES - TEXAS WATER COMMISSION
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
THE TEXAS HATER COMMISSION
FISH KILL REPORTING SYSTEM
November, 1985
One of the most visible results of poor water quality and pollution are fish
kills. The Texas Water Commission and the Texas Parks and Wildlife
Department share the responsibility of investigating and reporting fish kill
events in Texas. The immediate purpose of the investigation is to identify
and eliminate the causes of the kill. This may involve the elimination of a
discharge, the clean-up of a spill or just advice to a farmer about avoiding
dissolved oxygen depletion in his stockpond. We also attempt to gather data
that will be useful in understanding the causes of the kill and preventing
future occurrences. Some of this information could be used for litigation
such as the counts of dead fish and their monetary value, a chain-of-custody
water and tissue sample, and the names of witnesses.
Information that is gathered during an investigation is passed on through
reports, in particular through the reporting system that I am going to
describe below, but also in discussions with appropriate parties in the
Commission and with those responsible for the pollution. The fish kill
reporting system is intended to record and adequately document all fish
kills that occur in the State of Texas. The data base currently records all
fish kills investigated by TWC and TPWD since 1970. Investigations made
before 1970 were reported in memos and stored in the agency's central
files. Since 1970 a formated report has been used. Data recorded includes:
Date of Kill
County
River Basin
Stream Name
Number of Fish Killed
Game or Nongame Types
Cause of Kill
In the mid-1970's this data was put into an automated data base that could
produce summary reports.
Three years ago we enhanced the automated system, adding the capability to
record more information and formalized the procedure for field investigation
by our district personnel. The new information includes:
TWC segment
Pollution source, i.e., industrial, railroad, oil and gas
Water body type, i.e., pond, estuary
Monetary value
Areal extent of kill, i.e., river miles, acres
Duration of the kill in days
Latitude and Longitude
Names of complaintants and investigator
Day and time of investigation
-------
FISH KILL INVESTIGATION
The reports that are entered into the automated file are generated by field
personnel from the Texas Water Commission or Texas Parks and Wildlife. The
first notice of a fish kill is usually reported by a citizen to the local
field office of one of our agencies. Rapid response is essential to a
successful investigation and fish kills take precedence over other
activities in the district offices such as stream monitoring and routine
municipal and industrial inspections. The polluting agents dissiplate
of the cause
of sight or be
rapidly, tissue
characteristics of
information can be
most familiar with
related fish kill
events. They may also be able to describe the chronology of events before
the investigator arrived at the site.
rapidly or conditions
impossible to document.
scavenged by predators.
may change, making confirmation
Fish may wash downstream, sink out
In warm weather fish decompose
samples loose traces of organic pollutants, morphological
injury, and evidence of disease organisms. Often useful
gathered by interviewing local citizens. They are
typical conditions in the waterbody and historical or
When the field personnel make an investigation, they are prepared to, 1)
interview local observers, 2} collect biological information such as number,
size and species killed, 3) make physicochemical measurements in the field,
and 4) collect water samples and tissue samples for laboratory analysis. We
have prepared the "Fish Kill Investigation Guidelines" and a "Fish Kill
Investigation Checklist" which follow, as Appendices 1 and 2. Information
is coded by the investigators onto a form that can be directly keypunched
into the computer. An example of the computer report follows, below.
NUMBER OF FISH KILLS REPORTED PER YEAR
The number of fish kills reported each year has been fairly constant since
the data base was computerized in 1970 until four years ago. It is likely
that some of this increase has resulted from more effective reporting rather
than an increase in the actual number of kills. Many factors influence the
frequency of reporting. Some kills are not observed by anyone. Many
citizens are not concerned about localized fish kills perhaps because they
have accepted them as an isolated or natural event. Even when citizens are
concerned, they may not know which public officials to contact. In the past
few years public awareness of water quality problems has increased. When
state personnel appear on television during an investigation, the public is
informed of their role in reporting fish kills and pollution events. Fish
kills are now more likely to be reported, particularly in urban areas.
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
I
I
I
I
I
I
I
I
I
I
I
I
Year Number Reported
1971 74
1972 66
1973 67
1974 61
1975 52
1976 55
1977 52
1978 69
1979 44
1980 64
1981 43
1982 161
1983 139
1984 >61
NUMBER OF THE FISH KILLED
More than two-thirds of the fish kills record that less than 5000 fish were
killed. This is certainly a conservative estimate since the number of dead
fish observed is less than the number killed. Dead fish are washed away,
sink to the bottom and are taken by scavengers. About two percent of the
fish kill events resulted in the loss of more than half of the total number
of fish reported killed.
Number Killed Cumulative %
< 5 14%
< 500 40%
< 5000 71%
< 1 million 97%
-------
NUMBER OF TAXA PER KILL
We have data for only the last two years.
Taxa Count
Cumulative %
1
2 or less
3 or less
10 or less
30%
44%
61%
97%
DURATION OF THE KILL
About 15%'of the kills have this information.
Number of Days Percent
less than 1 4.4
1 to 2 48
2 to 3 20
3 to 4 9.4
4 or more 18
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
RIVER BASIN WHERE THE KILL OCCURRED
One-third of the fish kills were reported from the Brazos River Basin and
San Jacinto/Brazos Coastal Basin.
Basin % of Reports
Canadian River 1.9
Red River 1.8
Sulphur River .5
Cypress Creek .8
Sabine River 3.0
Neches River 3.0
Neches-Trinity Coastal 5.0
Trinity River 7.8
Trinity-San Jacinto Coastal 3.8
San Jacinto River 7.6
San Jacinto-Brazos Coastal 17.7
Brazos River 14.3
Brazos-Colorado Coastal 1.5
Colorado River 8.1
Colorado-Lavaca Coastal .5
Lavaca River .4
Lavaca-Guadalupe Coastal .6
Guadalupe River .7
San .Antonio River 7.7
San Antonio-Nueces Coastal 1.0
Nueces River 1.5
Nueces-Rio Grande Coastal ' 3.2
Rio Grande 1.8
Bays and Estuaries 5.2
Gulf of Mexico .9
WATER BODY TYPE
About 59% of the fish kills were reported from streams and 24% in ponds and
reservoirs. The remaining 17% occurred in the estuaries and Gulf of Mexico.
AREAL EXTENT OF THE KILL
More than 70% of the fish kills on streams and rivers were restricted to
less than two river miles. Most kills reported from ponds and open water
were limited to less than ten acres in area.
-------
CAUSES OF FISH KILLS
Dissolved oxygen depletion is listed when the cause is known for about 90%
of the fish kills reported each year. This cause may be over estimated
since dissolved oxygen delpletion is often suspected based on descriptions
by witnesses when the investigation is made after water quality conditions
have returned to normal. Dissolved oxygen depletion is caused by microbial
respiration, often at night by dense populations of algae that are supported
by nutrient-rich conditions. Microbial oxidation of inadequately treated
wastewater that is illegally or accidently discharged demands oxygen and may
exceed the capacity of a waterway to reaerate. Since the capacity of water
to hold oxygen is lower at warmer temperatures and respiration rates are
higher, about two-thirds of the fish kills caused by low oxygen occur during
the months of May through September.
Other causes reported include:
Industrial discharges of toxic or high BOD substances
Runoff from feedlots and industrial sites
Spills of toxic substances
Pesticide drift or runoff
Disease
Excessive chlorine in STP effluent
Fish culled from commercial fishing operations
Cold weather
Illegal fishing with rotenone or electroshock
In recent years kills related to oil and gas exploration are more common.
Pesticides and disease organisms are specifically identified in more
reports. Fish kills resulting from industrial discharges are reported less
often, although dissolved oxygen depletion resulting from municipal
discharges and by passes are still as common.
The Texas Water Commission Fish Kill Reporting System has been useful for
documenting fish kill events, identifying areas where water quality is a
persistent treat to fish and wildlife, and characterizing causes of fish
kills. Accuracy and completeness of the reports depends upon the skill and
cooperation of field personnel. Fish kill investigations are of particular
interest to the stream monitoring personnel and investigations are conducted
with enthusiasm. Their skill depends primarily on a good instruction manual
and training. We are continually evaluating the information that we
request, and clarifying and updating instructions. We hold annual stream
monitoring workshops that have proven useful in motivating the field
personnel, communicating procedural matters for making and reporting
investigations, increasing skills of biological identification and
measurement of field parameters. Our objective is to continue to build a
long term record and improve the system's usefulness as a tool for water
quality management.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Appendix 1
Fish Kill Investigation Guidelines
I. Objectives of Fish Kill Investigations
A. Identify and eliminate man-made sources of pollution causing fish kills.
B. Determine the number of dead fish, size distributions and weights so
that the monetary replacement values of the fish lost can be calculated,
C. Gather information that could be used to prevent or lessen the impact
of future kills.
II. Materials
A, References
1. Monetary Values of Freshwater Fish and Fish-Kill Counting Giiide-
lines.American Fisheries Society Special Publication No. 13. 1982.
2. Checklist of Texas Freshwater Fishes. Clark Hubbs. Texas Parks
and Wildlife Department Technical Series No. 11 1982.
3. Taxonomic keys to fish identification.
a. How to Know the Freshwater Fishes. Samuel Eddy. Second or
third edition.
b. Key to the Estuarine and Marine Fishes of Texas. Second
edition.Parker, J. C., D. R. Moore and B. J. Galloway. 1976
B. TWC Fishkill Forms-0129A and 0129B
C. Fish Kill Investigation Checklist Form (TWC 0563)
D. Maps
E. Instruments for measuring dissolved oxygen, temperature, pH,
conductivity, sec.chi disc and chlorine residual.
F. Sample containers
1. Aluminum foil and plastic bags for tissues
2. Acetone-rinsed glass jars with teflon liners for
water, sediment and/or tissue samples.
3. Cubitainers
G. Scale and ruler for measuring fish weight and length
H. Jon boat, motor and dipnets may be useful for some investigations.
III. Fish Kill Investigation Procedures
A. Notification of fish kill by telephone to district office
1. Person receiving phone call at district office should try to obtain
as much information as possible from the person that is reporting
the fish kill. Use the "Fish Kill Investigation Checklist" as a
guide to what types of questions to ask.
-------
Appendix 1, con't
B. Once notification of a fish kill has been received, the district office
should notify the Texas Parks and Wildlife Department Regional Chemist"
that serves the area where the kill occurred. —-
C. The district office should investigate the fish kill as soon as
reasonably possible after notifying TPWD.
1. The district representative investigating the fish kill should take
as a minimum of the following equipment:
a. "Fish Kill Investigation Checklist".
b. Map-preferably a USGS 1:24000 scale topographic map.
c. Instruments for measuring dissolved oxygen, temperature, pH,
conductivity, secchi disc and chlorine residual.
d. Sample containers: Ice chest, aluminum foil and plastic bags
for tissue, acetone-rinsed jars with teflon liners and
cubitainers.
e. Scales and ruler for measuring fish.
f. Monetary Values of Freshwater Fish and Fish-Kill Counting
Guidelines^
2. The fish kill investigator should collect as much information
asked for on the "Fish Kill Investigation Checklist" as is
possible. The fish-kill counting guidelines in the Monetary Values
of Freshwater Fish and Fish-Kill Counting Guidelines should be
followed in obtaining a count of all dead organisms including fish,
birds, mammals, reptiles, amphibians and significant
macroinvertebrates. Ensure that the count is a valid random count
conducted in an objective manner. The manner in which the count is
actually conducted should be well-documented in the investigator's
field notes.
3. Water quality data and water quality samples should be collected
from a minimum of one site and preferably three sites.
a. The site at which dead or dying fish are counted.
b. The source of pollution causing the kill.
c. A control point upstream or in an area unaffected by the cause
of the fish kill.
Water quality samples could be submitted using Stream Monitoring
Program forms or Chain-of-Custody tags. Chain-of-Custody tags
should be used whenever there is chance that litigation will
result from the fish kill.
4. Tissue samples if collected (usually a whole fish), should be
wrapped in new, heavy-duty aluminum foil with the tissue next to
the dull side of the foil. Data collected from tissue specimens
should include weight of individual specimens, total weight of
sample, length of specimens, species, sex and physical condition of
specimens. Foil-wrapped samples should be placed on ice in the
field and frozen as soon as possible.
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Appendix 1, con't
Data collected during the fish-kill investigation should be
submitted to the TWC Stream Monitoring Unit within 30 days of the
district's receipt of water quality analyses from the laboratory.
The data from the site of the kill should be reported on TWC forms
0129A and 0129B {See attached example). Data from additional sites
should be reported on regular Stream Monitoring forms. A copy of
completed TWC forms 0129A and 0129B should be mailed to the TPWD
Regional Chemist that serves the area where the kill occurred.
All questions concerning fish-kill investigations and forms should
be addressed to the TWC Stream Monitoring Unit.
-------
Appendix 2
Fish Kill Investigation Checklist
1, Investigator , Hate Time .
2_ Complainant Address
Phone -
TDWR District
3. County Segment (TPWD Region)
Rasin
4. Field conductivity less than 3000 /imhos/cm: Yes NO
5. Kill location:
6. Cause of kill: *
7. Pertinent Observations:
Appearance of dead or dying organisims:
Behavior of dying organisims:
Weather conditions:
Water appearance:
Comments of residents or fishermen:
8. What proposed action is there to prevent future kills?
9 . -Date kill began: Mnuu tnng ki|| has tasted' .
10. What was the source of the pollution that caused the kill?
11. Estimated number killed: (Check one)
<10 D 10.000 - 100,000 D
10-100 D 100,000-1,000,000 D
100 - 1000 D >1, 000,000 D
1000 • 10,000 D
Over
{Complete back of form)
1
1
I
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
-------
I
I
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
Appendix 2, con't
12. Species counts
Species
Total counted dead
Total estimated dead
13. Total length of river (feet) or surface area (acres) in which you actually counted dead organisims:
14. Water Quality Measurements
Date
.Time.
. Depth.
Parameter
Value
TDWR-OS63
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
AQUIRE: AQUATIC TOXICITY INFORMATION RETRIEVAL
AMBIENT TOXICITY SURVEYS
MICHAEL BAST IAN - US EPA REGION 6
-------
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
Presentation Summaries
Michael Bastian
AQUIRE: Aquatic Information Retrieval Toxicity Data Base
The staff of the Analysis Section within the Environmental Services
Division retrieves information from AQUIRE through a contract with Chemical
Information Systems, Inc. The data base contains summaries of papers
about the toxicity of single chemicals to freshwater and salt water
plants and animals. AQUIRE excludes studies about toxicity to birds,
bacteria, adult amphibians and mammals. It does not contain information
about the toxicity of complex effluents, oils or mixtures.
Data can be retrieved which pertains to (1) acute toxicity (2) chronic
toxicity (3) bioconcentration studies (4) field studies (5) algal studies
and (6) sediment studies. Papers entered into the system are reviewed
for the quality of the test methods and the amount of supporting information •
such as chemical analysis of the sample. The summaries can be retrieved •
in several formats which correspond to increasing levels of detail.
There are approximately 2,000 chemical and 40,000 records in AQUIRE.
The data base is frequently updated.
-------
Ambient Toxicity Surveys
The Environmental Services Division in cooperation with the Water Division
has allocated some laboratory and personnel resources for ambient toxlcity
surveys. The objective of these studies is to screen ambient waters for
toxlcity and, preferably, to bracket reaches of a receiving water where
toxidty occurs.
The studies are designed to follow the guidance 1n the Technical Support
Document for Water Quality-based Toxics Control. Ambient toxicity studies
are appropriate for low flow conditions and multiple discharge streams.
They provide a measure of the combined instream toxicity and persistence
of toxlcity from all sources. It is recommended that ambient toxicity tests
be Interpreted by comparison with the results of effluent toxicity tests
from the nearby area.
Generally, chronic tests are the recommended endpoints for ambient studies.
The EPA Houston laboratory can conduct toxicity tests with fresh and salt
water organisms (Table 1). Samples may be grab, replicate grab or composite
collections. If the receiving water characteristics are variable, replicate
grab or composite samples are recommended.
We ask that state agencies that request ambient studies consider how the
results can be used. It is important that state agencies commit to
follow up studies to identify sources of toxicity if the initial results
warrant'this action.
Table 1. Aquatic Toxicity Test conducted at the EPA Houston Laboratory
Type
Chronic
Chronic
Chronl c
Acute
Acute
Acute
Organism
Ceriodaphm'a
Fathead minnow
Fathead minnow
Fathead minnow
Daphnl a
Mysid
Endpoint
reproduction
hatching
and
Survival
Survival
and
growth
LC50 •
LC50
LC50
Test-period
7 days
7 days
7 days
48h
48h
48h
Note: The sheepshead minnow, the salt water cousin of the fathead, is
being cultured at Houston and will be used for salt water chronic tests
In the near future.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
• WATER QUALITY TRENDS ANALYSIS
" - ANALYSIS PLAN FOR WATER QUALITY TRENDS
- DECISION TREE FOR TREND/CHANGE ANALYSIS
• - EXAMPLES OF WATER QUALITY TRENDS ANALYSIS
LISA LAVANGE - STATISTICIAN (RESEARCH TRINGLE PARK)
I
I
I
I
I
I
I
I
I
-------
-------
1
1
•1
1
•
1
1
1
1
1
1
1
1
ANALYSIS PLAN FOR WATER QUALITY TRENDS
1. Identify data for analysis. Data consist of:
a.
b.
c.
d.
e.
Station(s);
Parameter! s) - raw data or water quality indices;
Time periods - sampling units (e.g. months) and beginning and ending
dates of series being analyzed;
Auxin lary data - flow, depth precipitation, etc.; and
Episodic data - plant or dam construction, etc.
2. Print dataset and edit data.
a.
b.
c.
d.
e.
f.
Observations (rows) correspond to sampling units (time intervals).
Variables (columns) correspond to water quality parameters (or
indices) under investigation.
Verify missing data witn source.
Verify units for each variable.
Check obvious outliers with source.
Can use STORET Retrieval or SAS PROC PRINT.
3. Generate Summary Statistics.
a.
b.
c.
Include mean, median, quartiles, minimum value, maximum value,
number of observations with missing values, and number of observations
below the limit of detection.
Produce frequency distribution - bar chart, stem leaf, or box plot.
Can use SAS PROC UNIVARIATE or PROC CHART.
4. Decide how to handle missing data and data below limits of detection.
a.
b.
c.
Delete observations.
Substitute mean values assuming an appropriate distribution.
Substitute regression estimates assuming an appropriate model.
5. Plot data.
a.
b.
c.
d.
e.
f.
9-
Plot water quality variables versus time for entire period being
studied.
Add reference line (vertical) for external events.
Add reference line (horizontal) for median.
Plot runs with reference to the median (+,-).
Plot runs with reference to preceding values (+,-).
Plot water quality variables versus seasons (e.g. months, quarters).
Can use STORET or SAS PROC PLOT.
-------
6. Assess seasonal!ty.
a. Examine plots (3.f.)
b. Subtract seasonal means from raw values and plot to see if cyclical
patterns nave been removed.
c. Can use SAS PROC MEANS or PROC SUMMARY.
7. Formulate hypotheses.
a. Is a trend expected?
b. Is a change expected?
c. What magnitude and direction are expected for trends/changes?
d. Are probabilities of stream standard violations to be estimated?
e. What families of probability distributions are likely to fit
the data?
8. Perform statistical analyses for each hypothesis.
a. Probabilities of violation/compliance:
i) Generate Q-plots.
ii) Conduct goodness of fit tests - skewness, Kolmogorov, and/or
Chi-square tests.
iii) Compute probabilities based on estimated distribution functions.
b. Trends and changes analysis - reference decision tree.
9. Present data and interpret results.
a. Resolve conflicting results.
b. Adjust probability levels for multiple tests, if necessary.
c. Append plots, data listings, and summary statistics to aid reader in
interpretation of results.
d. Identify possible problems with analyses needing further work.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
1
I
I
I
I
I
I
I
I
1
I
I
I
DECISION TREE FOR TREND/CHANGE ANALYSIS
Change Only?
I
yes
I
Seasonal
no
Seasonal?
_J
yes
regression
with seasonal
and change
variables
I
no
t-test
yes
regression with
seasonal, trend,
and change
variables
no
regression with
trend and
change
variables
I
residuals normal
& homogenous?
residuals normal
& homogenous?
I
yes
I .
Stop
no
Rank Sum
test (aligned
If seasonal)
I
yes
Stop
Stop
No
Kendalls' tau,
Spearman's rho,
or Sen test
(adjusted for
seasonal 1ty
If necessary)
Stop
-------
I
I
I
I
1
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
• EXAMPLES FROM THE WATER QUALITY TRENDS WORKSHOP
| Example 1 Summary Statistics using SAS UNIVORIATE
•Example £ SAS Plots for Seasonality and Trend Determination
Example 3 SAS Plots of Deseasonalized Data
V Example 4 SAS Regression Analysis
Example 5 SAS Tests for Trend Analysis
i
i
i
i
i
i
i
i
i
i
i
i
i
-------
I
I
1
I
I
I
I
I
I
I
I
1
I
I
I
I
I
I
I
-------
I
I
i
I
i
i
i
i
i
i
i
i
i
I
i
i
i
i
Example 1 Summary Statistics using SAS UNIVARIATE
o
IU
ii n
in 03
» o
« in
i
II II II
.» o — o
• o o o
n 10 ro »
o a. a. a.
E252
SjEpjE
P o o o
O A A A
A O " O
O O O
2£££
O.
a o, ii
zzSlo;
n n H n u i
co r- .3- tsi -*'
o cv) ro c- in i
O O O O O I
A A F r- I
in co o o <
X O
P4«4H4V V****~<
> a ii -in
IU II UJ O CO
P a. £ a. a.
iu o o o o
K o o o o
' O O O O O O O
o •» o
oooo
_ K! KI -J-
IA CO V V
V4 O
•c in IA o
•• " W IO
!2£
V V V
V V
in o in
•a o o to «
££2££
• IU -< •• •* •* ft
« • V
•« V V V V V
V O
V W N * CJ •< O
o o CM to * in o-
omooooo"
in iu
M e
M •
. Ik u. U. I
oooo
oooo
000
o o o
'OOOOOOOOOOOOOO
oo—
-------
eooooooooooo
„ •*• O -4
o o -< o CO
M !*• w Kl M
II CO
— . — in ii
O -> t£ IU «X>
U 19 I- l-l E
ooooooooo
OOOOOOOOOOOOOOOOOO'
OOOOOOOOOOOOOOOOOOi
oooooooooooooooooo
OOOOOOOOOOOOOOOOOO'
oooooooo^oooo
oooooooooooo
OOOOOOOGOOOO
oooooooooooo
u.
M
in
M
(-
in
1
<
M
o
in
in
I
I
I
i
i
1
i
i
i
i
i
i
i
i
i
t
i
i
i
-------
I
I
I
1
1
1
1
1
1
1
I
•
•
•
1
1
1
1
1
I
IA
ft
i
IU
i
IA
H
£
in
tit 14-
H U
IU
IA 0
3 H
H W
® £
S a
o u.
o
^ <••
X O
K 0
VI -» IA
> * K>
£ 5 i
O fO
CM
g~ |j
M H 1-
k. Ul |
^ H IX K.
• «So»J
6 J UJ N
in ik it x
« ii:.*
8tt Z OIA .
< IU Kl -4 O
a. _ » v » o
o ** < E
«aP uj-w
O IU 3 Ik — O- UJ
Z t <9 O H
-1 O Z IU <
H i ik tt a. <0
«n «t i z z z <3
m
S
Or
CO 10
DC M
IA •»
CM * CO
DC CJ
O
• * O 0
co u a. a.
*« . •
(•« 0 * •»
e IA co m
*" 2 0 IA
• f» O-
* • a. o.
O GO CO
0 IA Of O!
a. O lA
DC
LA
O
Q.
(M —
O >—
a w
u
SG^
a: — uJ
S S °
- « Z
Q. iu
•9 03
0 II H
w o- >- a.
o — o
^ l^ ii.ee
<
I
Vt It II
E •* O •
• ••I^KI*«. .1. o
ik ii ii ii B. a, n n n u ii n o
uj -OOOZZCJM~O • • Ki -« o -a
8«3ao.o.xxiLQ.o.ooo.Q.a.Lnz
o -« HP r» to o- ui
azzz z z z a. a. z z z a z
if) luujiuooiuujul iijiuuj H
< Z XXX XXXZZXXXZ
tniUHHPA AHHHUJUHI-HllJO
•^ O
OAAAoir>AAAOOAAAON
CO « — 0
iuAo-OOIAIAIAIAO.
(A eoeiiooo o o o <-<
|U>*> *4»|UU*..|U«-«
U VVVIACOVVV VVV
>V «O VV V *l
iD.a.o.aQ.aa.a.a.a.a.a.aao.o.
|
1 lklklkU.u.u.u.u.lkU.U.u.|kU.U.Ik
«,
II
tJ M •* IA (s
O O 0 O Q.
IU
Z Z Z Z X
IU UJ Ul U H
H-HH^O
O O O O A
r» * cj -< o-
0 0 0 0 h>
tv a. a. a
o
o
a. a a. a. a.
u, u. u. u. u.
0
Z
li»
M
w iij
< 0.
a o
M
Ul H
E
Pi
H^
53
c£
ss
CONTAINS P
CONTAINS P
gg
LLj
to in
i i
i i
i i
§
u.
o
Ul
E
M
Ik
O
Hi
_j
o
*• ^y
O U
E a
I ss
3 lA
O Of U
U
IU Q. CO
in o
i g£
.- PC
o a —
iu •o
a. o S)
3 -« Q.
o o
iu a
S-52
X -* IU O
^ O cj P
0 H O IU
N ce o
11 ' U>
•"K^E
H 1
Ik < 1
Ll
2
o
o
N
a.
IU
U
H
a
IU
•;
o
IU O.
H
CO 3
a
0 Z o
» U (O
II •* O.
0 »-
0 O H
o iu in
o?" S0
O ii H &
O — i fJ O
a a a. Q.
1
Ul
o
ji §
(0 Uj
a -(
3 IA
o - o
D: U - -
* 3 E Or
n O 3 ui
IA Z M O.
•i «J E 0.
IA O -
a- - it
- 0. II CJ
H P- *
«j eu o
< a
RESIDUE TOT
RESIDUE TOT
•ARSENIC'
•CHROMIUM1
= 'MERCURY*;
M II W ^ O
M o e o 1-4
«r;
PROBABILITY PLOTS;
3$665
Z oo
H E o.
^ Q; •» Is*
H O -JLAtJ
V) z • is M ct a.
SSSn'S*
r: H H 3 -t-t
SB ii B.
•• 6 E aaiA
-< u E O-
IU H V> HOO
a. o iu MQ.O.
5 H 5 goo
x z a >-r-
IU M Ul MtkO.
- gSBa
-J U 1 <-»
H O 1 O
M DC 1 tt
H B> * O-
^ lA *6 t** CC^
I'
-------
»- o o o e o
«n * e o o o
I «1 CM CM C:
o
M
S
r»
(0
P*
•
§
HI
H
M
g
u.
S
M M
5 S
? >
m M
>. §
1
*•
*-*
EL
UI
*.
lU
5< H » r. o- in
*
N X X X X X
•» o> in o o in «H
* o> -T> o- — i
ii
IL
u
o
<_ • o —> r-
u •
M 0000*
H- C3 CM -H -T C>
Ofi:0'r SSui S 8
XXXXX Sid 10
oirtomo x
or>-incM tt o r r
v-f
Ifteco**-.-- -
p> y) f*» ^ r*» rt o o o
o (^ o &* ** o o •
CM CM « O
S CM
CL
n
w IA e a-t o -i
CD lu UJ Z < H !C
< D 2 < c r a
M Z S UJ O
K •• O D ••
> zccr>.nr)L)i- CM
KI CM
KI CM
ft CM
<0 ••* CW
^ ••* •*
O «^ KI CM ••* O
DCT*.fKICM»w*O
CM
•>
*H«
+
O
•»
••H
1
CM
1
•
•
O
^
*
»
o
IP*
CD
u.
w
r
UJ
u>
^
D.
h"
i
1
I
I
I
I
I
I
I
I
I
i
i
i
i
i
i
i
i
i
-------
1
1
"tr
*"4
^H ift
-
1 2"
• &
| j
8
|f"<
|
CO
1
I
<
1
(A
^ a u,
I | |
& £
1
M
• i
1
1
1
1
1
1
*- •« O IA «0 IV
« •* IA in in f-
W IA e o in «* V
•» o in IA • o o
-• in !•- K>
h- y M •< H in
is z in uj — — o
S < O J- A A A
IH H (0 ffl CD
CD EEtKCKMOOO O
a 5 5 < 5 in I- o: a K
5 55 «> ^ u in o. D. a
U
£ "A *^ "A ^ ^ ^ *"* ^ ^ ^0
• <0 9* ^ * » sD *< «
e o
E:
§ >S 1?^0<
CD iu uj z < " r
3t a Z jtf o: r a
at M
-------
g
S
IU
g
E
IU
<1
•m
z"*
M OC
IU
IU D
. If
o i-5
M\ 2
SM iu
en iu
in < Q
in in
0 M
8 -I
-I U a
i *-3
grS
ui iu
- II
Example 2- SAS Plots for seasonality and trend determination
g;
:e o o
CO o o o
o o o o
o o o o
o o o o
o o o o
o o o o o
o o o o o
o o o o o
o o o o o
o o o o o
oooooooooooooooooooooooooooo
r«»-iw4»Hcsjc\jcJcacMcMC>jcMc.jc>jKitOKitOtOfiroroFOto<>3>.^^
oooooooooooooooooooooooooooo
oooooooooooooooooooooooooooo
oooooooooooooooooooooooooooo
oooooooooooooooooooooooooooo
oooooooooooooooooooooooooooo
II
§
O O
i*> in
e> e-
a (0
M CSl
il II
«J -1
22 ..
K cr ui
|U UJ
in in
g
«n
S
M
O:
3 i
<0 •• E
tl M
in x
II
II II
•* «
Ul
u
•r 2
*8
1*1 II
20
20
ED AR
oo a
gz>
a.
u u
tff
o t-
in o
< If)
Ul
I
Es
DC Ul
il
*i
•• II
p
in
§ S
il
IU
in <&
o
m
ui
^tt w
H. i g
•a ":
u
lu
in
•o
o
o
o
Ul
in
i
S
-J Ul
S 8
- g
Ul
UI
o in
o —
>- • 10
^ >• in
o 05 ea
o >
to in o io
?"oS
og^O
Ii OK
in o n
M ii in o
gfeSSgiE
II O * O E Z
«t c in n » o
K. * •< * cv c
^ o «£ in «* *
o o •» o o o
Q Q
1
co
o
CO
o •
o o
o
Kl
§
Kl
CO
a
• § «o
CO Z Z
£ £5S
CD Z Z
'if
o - -
-ao
p p 0 . .
e- in ii o in
II 111
x Q o
Ul - - <-•
t- CO J- O CO
W "1 -<
3- -
<:??
LU CO ^
t- Ix
-------
1
1
1
1
1
1
1
I
1
1
1
i
i
i
i
i
i
i
i
«•* ift
P*
*-«
S
*
5
?
£
p*
i J
M P
N* Ui
i »
i 8
Ul
H M
Ul
0 II
e to
E S
p 8
u
E <
_i ••
i §
o w
in 19
< IU
u _i
en
B
1^ Z
.J *
a. o
e
n
.. a.
N
U.
IU O
a t
"*
-------
r- in
§
iHI
Ik
*
. " <
i - :<
§5 <
to u) <
< IU
UJ -J
«" . < <
g <
U. UJ <
w < ^
6 8 < 5
t S %
*
.. S
w a.
UJ U.
-J 0
1 s
£5 E: <
^
s -
<
<
<
* <
<
*t
ooooooooooo
?% jj* ^ ~\ ^ i& c\j O* •£ l^j
r^ CM cj cj -H — ' —•
U. U> (J L> O -J M
1
<£
<0
Z
n
•H<
O
1A
o
->
r*
0
1JANS3
0
CV
§
<
->
•^
o
CO
2
<
*^
•^
0
-+ — — —
JANSO
r*
0
1
<
->
«M
a
CO
i r-
1 Z
+ <
1 ->
1 p<
1 O
UJ
Q
Ul
2
<
ne
u.
O
i
UJ
CL
tu
3
g
at
UJ
-j
i
S
V)
c
i
•H
tJ
UJ
|
r
1
1
1
1
1
f
1
1
i
i
i
1
i
i
i
a
i
i
V
-------
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
e in
••< co
»4
•S
ft
ee
i
> <
i < •*
\ - .--•
? - <
^
*••* .
5
§ . *
P 2
< u -
1 . <
£8 * <
iy O
G ft! < ^
o •* <
II ^
i * <
UJ < ^
P rf *
§8 < <
* <
t "
3 * «"
i <
I i
si
s a „
s 2
u. *
w pj <
1 * 2
a. o
o.
~ S
ui • < -
IS *
t ^
< 0.
X —
<
<
<
: -
*c
<
3
^ <
<
omeinetnoinomo
m*»tOMCJCJ«-«'-«
i- 3 a a>
I
§
«-«
.A
(O
1
P4
o
*
cs
Z
4
•-»
«^
o
t^
cc
i
n
0
1JAN32
o
*M
40
<
1
t^
O
0
Z
•1
P4
0
£
5
-^
o
OIJAN78
UJ
H
<(
O
UJ
I
a:
u.
O
w
UJ
S
K
g
en
UJ
<
13
w
tn
in
•H
^
^
|
in
S
Kl
•^
UJ
I
IJ
-------
Example 3 SAS Plots of Deseasonalized Data
*
»4
oe
i
I
Ift
IA
o o o o
eo cr- o —
o o o o
o o o o
o o o o
o o o o
o o o o
0 0
CJ Cl
0 0
0 O
0 0
0 0
o o
ooo
•x in -o
ooo
ooo
ooo
ooo
000
o o o o o
rv co o o -•
o o o o o
O 0 0 O O
O O 0 O 0
o o o o o
o o o o o
0 O 0 0 0 0
CM to -
O 0 O 0 O 0
O O 0 0 O 0
O 0 0 O O O
o o o o o o
o o o o o o
o o
CO «>
0 O
o o
o o
0 0
o o
o o o o ooo
0 O O 0 OOO
o o o o ooo
o o o o ooo
O O 0 O OCO
O O O 0 OOO
in
g
0.
^v
w
in
IU
p»
(d
g
<
?J
£
*
CO
ui
<
in
(9
o
^
Be
in
in
0
BC
M
CO
•* en
CO <
t> X
w<
m
A UJ
He
•1
1- 19
§ CD
M O
o: i
>"
O. UJ
0 X
t> p
« «
e o
o o
IO fO
II II
Ui U
£c
£ «
0 0
!•- »»•
M CM
II II
-1 u
K or
in in
e o
CM CM
U II
||
in in
ix o:
m n^
> 5»
O O
IH M
£2
u u
»»
UJ
51
o
111
M
U.
M
Q
Ul
0.
to
Q
M
HIT
0 K
l/> §
in
58
in
co o ui
uj z
in in o
< 3 n
5 g
i i- in
; in
a:
D.
»- UJ
in
g
M
IU
tt
5o
$2
^H
i§
Is
> Ul
Q: in
IU
in r»
8^
o
in
co o
in
in in
3 O
iu
I- O
IU O
in a
a.
< X
O 1-
5 fc
o o
Z Z Lf) vO
!§ i,
a
E
<
«O
s
S
a
ui
a
0
Z
Ul
o
a.
o
a
a
g
If)
>- iu
5 *:
in o
i s
w s
o
u ui
p y
in
ui
o
CM
O
••4
in
.- —ul
O OQ
O. ClA
•O
fi f-i-O
O GUI
s=p
If) 1
>^ o. c«
a in
< tfl UtUJ
E -• - a etna.
iu 2: s tu
a o t~ o. Coo:
• r; ui
— ul O >- V>->
Z O E « «O
M t^ n: ttC:
UJ >- nr ctC-O
< z to < «o
o — <\j m ,j- o o in so r- eo o o
.^_*«*.*_zz: — ^ -< « ^ CM
i p
-> f J O
tsi rj Z
rj eg tjtjcj
I
I
I
I
I
I
I
I
i
i
i
i
i
i
i
i
i
i
i
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
o o e
I* CO O~
f\ n o
o a o
o o o
o e o
o o e
o o o
S3
o o
CM tO *
•*•*•»•
o e o
o o o
o o o
o o o
oooooooooooooooooooooo
oooooooooooooooooooooo
oooooooooooooooooooooo
oooooooooooooooooooooo
GOOOOQOOOOOOOOOOOOOOOO
oooooooooooooooooooooo
NG
S
in
S r
a.
ii
to
BE
5)
<
en
26
AND 57 VARIAB
568K.
P*
ii
lu O
to u
SIU
OT
in r>.
CO O
to d
So
IU
§in
«D
25
to ui
Ul
a
is
in
IU
AM
DA
3£ g
« H w
M z < in
£ •< u in
IU I-
to <
a
t- ui
< X
e> i-
O O
z z
A
0
e
t>
en
Ul
19
<
a
o
Ul
t
M
£
g
<
*
O
•o
h-
|
en
Ul
V)
&
M
o
O
Ul
in
Ul
<
M
|
M
Ul
g
Ul
U
o
8
IU
H
0
K»
CO
1
V4
-
p
CM
->
•PH
»
O
^,
z
->
™ Bo
11
-> i
M M
•o a o
to £ <§
S 53
en -> ->
Ul MM
0 V. - -
M t- O O
z < - •
•H Q CO *
u * r^ co
?o55
«< »o -» ->
o in •« M
IU - -
t- O II
o in
_I 1- M
IJ O, 3;
o
°-
O
M
I
1
M
-
O
CM
-»
i-t
"
a
•*
S
<
n
^
o a
ii
n ->
v4 **
p p
» in
N co
V ^ ^
Ul M M
H • -
o o a
» • -
in co *
M r» o
M?§
ro -> ->
10 M — .
UJ . .
Q II
m
23
a x
o
K>
CO
Z
<
•>
v^
•
o
tJ
-)
vri
•
o
M
z
1
•«l
e o
o >o
CO CO
-> ->
o o
Kg
55
1 -»
M _4
Ul
h- O Q
< - -
O CO •*
» f- O
Ift Z Z
«0 < <
.e •> i
tn M — c
Ul * -
O II
to
1- M
Q. X
0
Kl
1
-1
•H
•
O
CM
CO
1
p4
"
O
^4
z
-»
" ..
O 0
o <£
5 ^
1 1
M M
O Q
o> in
!*• CO
55
grr
< o o
o - -
» CO *
CM r» co
» Z z
0 < <
5522
IU . .
O II
10
H M
23
D. X
Q
J«1
I
->
v4
-
O
CM
CO
1
*H
•
Q
M
Z
•V
a^
a o
CO CO
55
"1 -»
«p* **
pp
tV U)
t^ CO
zz
3 3
S " "
Ul Q O
t- . .
•< CO •»
O f>- CO
£55
En-22
Ul * -
O It
to
K >1
35
Q- X
.
m+
0
M
£
S
£
S
1
£
Q
i
1
en
g
w
u
UJ
to
. «^
IU O
So
O Ul •
tn y-
SSI
0 O UJ
a.
< UJ 0
00 —
u. UJ
U 0
O ui
a in
tu in
x <
t- to
O
Z
ui
o
CO
in
CM
Ul Ul
H* Ul O *
M -I O U
(_ U tO .
11 Cr Z
Z MX
M O O *
CO >-
in in o:
< < o <
v> 10 a u
ui
ro 10 z n
in -o
10
• o
l Z
CO O-
ro M
C* O M CM Kl itf1 1/1 •
coefo-.fjf>.»i/>>or-'
•a-^inininmminuAini
-------
BE
i
8
§ o
? £
OOOOlflKlOONOKlO
eeo'eeeooiooo
OOPJOOOOOKlOin
Q
Q
IU
g
UJ
: 2
* s •„.
y. r -41
o
12
3 8:
a. o
-------
UN
O ft
a is. •+ o> o o
a. x e •# eg
a. «4 o KI •*
a. •*o ft r«.
(L -4 O O CM
n.o> in
a in o o
a. in •-< in
a. in KI e
a o o
CJ
is"
co •" •*• ~
in
CO
in
o o
in in
in KI
00
eg
i
o o o
o o o
cj in g£O IO OJ ^OCMO^**
nut •* & inrs.com-*
15
g>
Kt
' d
Kt CM •« .
K» l«> * •
••» ft m i
v- o M CM
' in * * co
Kl O CO Kl •
a ui w
CM
O
o
o
eg
,0 fs
is. r»
e o
to o
js IS.
in in
CO CO
CM ^*
O O
eg« ft
eg co f>
ill
O ft O Q-
eg ts eg co
* in •*• •*•
•4 CO ft ft
—1 CJ .3- O
o o o o
o> in
~ Q
—
o o
o o CM m
O O O O
* *
g> g>
in M
o o
m -o
o o
> o o
o o o o o o
I I I
e e
i
o e o
I i
O IA
•f *
3
s
o o
eg in
-«in
cj r» o co
-« «M
N
i CM -O • »
CO CO
co co
> CM in
' o e a e
cj 10 r~ -»
o gj gj in KI o
>»• ft eg i? KI in
co eg o g> gj.eg
o gj o rs CD g)
O 4> CO
KI in KI
co co eo — •
1A co r*> KI i
III
i in
• •»
KI o o
o
o o
* * m i
o o
-3-
'3
eg
o
CO
i o
i O I
I O I
I -O i
o o
o o
*o -o
O ft
o o
o o
o o
o o
i O O
I O O
I O ft CO O i
I I
i e o o o
i O O O O
o o m in
o o fs rs
o o
o o
m in
o
CO
ooooooo o o o o
* * co co -o
CJ CJ
fj CJ
00
o o
in Kt -o •*
CO CJ « «
lA fs rs
ft ft CM
eg co eg eg i
eg — .-i eg
o. gj> -o m
a. -a Kt o
o o
0 0
o o
0 O
CM
e
o
o
o
oeooooooooooo
ooooooooooooo
^inej-»og)OKivj-^cjcoeo
ooo-*-*o~*ooooorj
ooooooooooooo
O >
O <
. gj o
• eg co
o o
o o
o o
o o
CO CJ
Kl Kl
o o
in in
o o'
cj eg i
i eg eg cj co
i ^ *y ^ ^
in in
§5
o o
• & •&• o o i
i *< —• O O i
rs fs o o
in in o o
co co m in
o o o o
o o
o o
in in
o co -»
l « eg in
2'.
oooooooooooooi
a - •* -•
«-• eg. •« eg
OOOOONO
" Kl IO I
M 1*1 (O Kl Kl I
.
>o -o •& -t> -a -a -o
CKIOO KIK>KIK>KIK»M
O O
o
•a -o
CO CO
•C -O •
t*. t*
"* *
CO CO
rspvrstscoco*H«4v**4COCOC3CO
rs.£in-£<£to-4>*oinoj>in
' '$*
o co o o
• Kl •
CO I
IA
oeooMoboo
a. -»o
it- M r iu
O < K U
O CO
O IA IA
CO O- in v> in
O to co f i
^r^-^-sTa-sTT^rsr^^^^
ooooooooooooo
ooooooooooooo
mu)tomtnv>inininininin(/i
Iu. o: u Or |
r o z »- x
U eg M
U eg O
U -« t>
U -4 CO
U -« r«.
U -< <«
u M m'
u ••« •*
o-< KI
U — eg
O — —
U -• O
O fcis|sJf%fs.cocococo
icgejegegegejCMinKiKiKi
O eo en -i <\> hi
-------
ft. i-» ** o- o o
in
U «M
U t
<* CM CM
O O O
-i CO O- O CM (OCfCO Kl KIO
EL •« o in « •< n "CM ' •» "CM ' K! -4 K) ' m " KI CM
ooooooooooooo
u
g
IU
a -i
V)
O UJ 10 •-• O * CM
cocococo
CMCMCvlCJ
o o in •*
in co •»
III
-O CM Kl CM II) US 4 in K>
CM X Kl C\J « IA i-l icTCM
o o ooo o o oo
a. -< o CM r^ o
O in i i
eoooooooooooo
i i i i i lit i
i r* KI ** •
I -C Kl |v I
o> >« o> -*
CM r» CM r>-
< o-
in
CM
lUitnKioo cMcOcor-oooooooooooooo
~ co co i
•^ 4 «O i
O. in o o ocoo^o • • o co -o co w CM • o •* o co
O CO CO O ^ O CO IV CO -• ~ O CM — O
M CM ~» <-• *-• -H -H
O. in >* m CM o CM eo o • . * ~
oooooooooooooooooo
MOOCMOOOOOOOOOOOOOO
I I I I I I I I I
inininintninintninininmcocococococo
CMCMCMCMOOOOOOOOCMCMC'JCMCMCM
OOOOCMfJCMCJCMCtJCjmOOCPOOO
ooooooooooooo ooo
• « M M —i Kl CM Kl 1 CMCMin —
CM CM CM CM O <> O I
* ^ * •* KI KI m i
KI K>
I^IIA oooooooooooooooooo
m
KI -< -o <-> in
oooooooooooooooooo
oooooooooooooooooo
oooooooooooooooooo
oooominminininintnoooooo
oooooooooooooooooo
oooooooooooooooooo
O. >O CM in Or-IOOrtOOOOOO^-HOOO
(L K> -* in e co oooooo
* CO CM CM M Kl
M v CM U) O O O O I
i ICMCMCM
O CO >£
o r>
O -C -I
o in x ^
u -s-
O Kl
M
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
•
>.
If
1^
5
l!
o
IU
• «
II
s
I*
I
O.
I:
c.
I
I
I
I
I
I
I
MM M N
iLNlChOO O O OO
^ *i cO i*^ o o o *if *o ^O
O. — o * «g Ift -H •» *egcgg)K>"gJK»roin' ^ooc\r^
&9-IA in -o ro Ki K)o-*tg-jf--Kiing5ifi• o • •
MOO • • eg o- eO rj w • eg * o « rt ^ eg « eg «
ooooooeoooooooo•o
en
O 1»l V> -H O * M
o in m iin* o o
E « o
OOOO O^OOOOOOOOO
our^o to r^ 10 to cocoo^i^coo^cooxh.
l^t^*M(M v«tO i^7^
eg
OOOOOO-*OOOKIOOO«4O O
CL >« « in ooooodoooooooooci o
IO OO OOOOOOOOOOOOOOO
CL -c eg in oooooooooooooooo o
O. K> rt in O CO OOOO >OOOOOOO lOOOOO
co in o eg o j- o o -T o o~ e\io>rrjO
MKI^O* OCMh*t>»CI't>e'KIKlK»KIIOtO«i-.^
^^>MF4M<-ii-i-4'4tMegc>icMtMrgcvltvitvi
oooooooooeoooooooo
oooooooooooooooooo
eO
0. p.o-o>O>o>OP-o-cgr >o
OO>f^-3'>r«4'4'OOOOOO>C in -*• tj
»oo>oKioinmoNin^r<«f^m«o<4>^^g3<^>^
o o o o o o
&KIO<-I ifto ••• o * * * ' in «j "
O>KIOOOOOOOOOOOO^OOO
a. KI o o ...
eiAooppopooooaptfinop
& -« o eg ej eg i
II- M c ui -• o » l
Q < 1- IU
oooooooooooooooooo
oooooooooooooooooo
eooooooooooooooooo
oooooooooooooooooo
oooooooooooooooooo
O -.-.-.-H — OOOOOO — -.-<
oof^^*h»r^^(s*hfcOOOoooh-Nf*»
acAinrtininininino-c^w-o^oo-ininin
~t~i^
o —
O (0
U N
O •* J -J
u in ^
u *
U KI
-------
in
OKI
Q- U <*
•* OL f^ »4 G* O O
LU 0. *+ O 1ft ^
£
0 CL *-• O •£• *l
2
M
^
< O. -H O K)
8
5 B. •* o eg f»
f
h> a -< o o cj
m
a. o in
in
M
O
CO Kl CO
CJ CJ O
O * •*
K* eg ^
e o o
*-* <9 it
in o o
eg r* K
tn * KI
CJ
0
<6 O .» r- in •*
••> KI •» o eg KI
M M eg » »> O
o -4 cj -4 in -o
**
Kl Kl f^o CO O
eg •« eg eg -o
** eg « to in o
o o o o o o
o o * KI in
>4 Kl •* O 0
• toooocococ^o
•OCJOO^l-ClAO
eg co
eg-
Kl
M
•* •»•
eg eg
•«
o
in
f.
OOO
in o o
* CJ
0 O
•>4
o
- «
eg «
eg
<0 O-
•4 Kl
eg
o
•4 »
O O
0 — Kl
iv -o in
o ui tn r> o
** o -9 -a o
a * -o m
(L •£ Kl O
eg co •
>«O«oeg^4COc» »cjcg>Mg>cj
eg eg in
a. o. KI — in o tt1«O«ocjooooo
ooooo oooooooooooo
oooooooooooooooooo
o o <~< cj o ooooooeooooo
oobooooooooooooooo
CJ r4
-o tn
n K> -• •* •* in
oooino
r»u>''in
eg -< KI -<
oooomr»ooooo-*r~
O. * o o
a. KI o -4
a. KI o o
II- M C IU
a < i-
o- o
•O P»
o o
CO sO
ooooooeootoooooooo
co N -4 a- —
S
oooooooooooooooooo
oooooooooooooooooo
oooooooooooooooooo
oooooooooooooooooo
O CO
lu. a uj cr I
E O 2 I- X
U CJ «
u eg o
o — c-
o -H Kl Kl
« Kl Kl K)
^) K> M M
^O K> Kt Kl
O Kl *H o
U) Kl Kt O •«
CO
I
•
->cM-» — -j-
OOOOO
OxOOOOO
I ^ 10 ^ M ^
i »4 co •« * m
K f«. r. r«. N *»
in in in m m eg N in
T o cj
^ ^4 ^ *y in •& cj
-• eg KI -4 r» co o
i cj cj eg co oo
^ «J CJ O CJ *4
h« ^ •* Kl cj I
K> « -4 in m <
i _ n t ~>
e*» v* -4 f> e> o
r^ cj cj r* />• «o
Ki co KI KI eg eg
OOOOOOOOOO-MOOOO
i
l l l
oooocjcjcgcgcgcjcjo
mininm-»'*-*--wf*»-'a-'.)-p
ininmmooooooot^
o o o o o o
o o o o o o
o o o o o o
OOOOOOOOOOOItKIKIKlKlKlKl
CJCJCJCJGOOOC3OOK>K>K>K1K1KIK>
oo
oa
ooooo
ooooo
oooooooo
oooooooo
in in i
CO CO
cj cj
cj cj
* * •*
> m
l *£
in m in in
m m
vo >o -
-»•<«•
tn in
i Kl Kl Kl Kl Kl Kl Kl
****•*•*••»
It Kl K> Kl Kl Kl
Kl
<»•
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
tft IV|tlOK*f*»f^OOOOO
S^J *3h ^ t 1*^ t ^ ^ O. O O O O
«O I*. It 1*1 *fl *O OOOOO
x uit o lii 01 r- o
O" O O . . O f-~ fv m (0 N Ol if)
Soo x o iu in x o ^ «j
co in * x r» o> o x
IU «\l x O X K> O CJ I I I I
o ui in • MCO^N ooooooooooo
»4~< ^< ..,.«...».«.*
OOO OOOOOOOO
in -4 ** o e-.« o> ill till
a«oK>* . . oo
IL <-> o N r^ >. oitJinitxioxin ooooooo^ -H ^- in -»
o oo ooooo incur-NinN-*-* m f- m —
II II I -« I Kl PH
*0 x « ' r^ ' 'd o o o o o c; r; r-. f. r; »o 5
OLOIfl *oNco<>.f»ininh-o ^ odddddococdcdtococo
CJ MrtlOXXXCM rtXXXXXX
en
oooooooooooo ooooooooooooo
&(L«ltO ••
oooooooooooo ooooooooooooo
-J ooooooooooo
•Xlt*^tf)OCO OOOOOOOOOOOOO OC^O^ChOOOCOcCCOCOCOCO
Kio-tf^f-tinocjeo^-or** ^^.^.^f^p^.^
OOOOOOOMKinKlKIM
m ^ 10 *^ ^ x m ^ o o ^ t^ o o o * i*^ o o o in in in in in m in FO tt fv 1*1 i^ fo
a«oo IU.KUJOI
tt.l«)0x . cu K CO .... E O Z (-I
U CM x
oooooxoeoooo
CLltOO '****?* ^^^^^.i 0-°
: iu in rs. IA i
u x o
ooinoxtr-oviootoro Uxr»
a. x o
-'------' — -- i U*"i^O
' o x in
i «J x *
I
i CJ x Kl
_ . _ . .^ _ ^ . . _. ^ (Jxcg
* U x O
OOOOOOOOOOOOO
ooooooooooooo
ooooooooooooo u<9
ooooooooooooo
-------
Is.
*M
*
-»
V4
or
UJ
£
IU
§
^
>.
i
§
IS.
Ul
ui
£
o
o
M
-I
3
-^
m
VI
Ul IU
O K-
Ul M
^
U. M
P §
_ 1
a
•«
K»
s
O.
^~ g& fs ^ i*/ m
W co 43 •g) «o eg
Iu eg CM
• • • i » i
I
or x x x x x x
n c- in o o m «
U. 0- 0- 0- -H
IU
o
2
-j in -< >» in m in m m
M eg^^rgKi cgegr^
K *4»tno^ gi*4co
~f *^O<0 t *>
^> Z A
— ' 5 o
< 0 Z
> o x
M O
en o
in
M X
r
0 *~*
-j i
1 <._i — +
in
iu
o
ID
n
1
co — PI
i o
IU f»
r. <0
S 2
CO
. in o-
<\j o — < m o
^^ P
A A A
i in r ——
' M < f- |O
: tf> ui^^
; o E A A
II- 1O Q
: tt in Q o o
: 5 w I- nr a
• a* o in o- o-
ir^r«-«4*ooiftcgj)
i —* eg K> i^ o ^4 *coin
i o r-
co co •» in KI co
tt -< ^t-.-<
<»•
KI m
eg •?
*4 ^ ^ gs o
eg in -4 -o o OT
iu u
or.
z ££ i
< o iu tn
ui t- n: tr> >
f tr
o
-.
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
-------
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
I
CO
Ift
<0
**
^
LU
I
j^
§
i"
i
^
U)
in
•*
j*
g
S
K
.j
I
m
u
(o
u u
O 1-
UJ M
K (X
I— <
>
U. M
° i
in
i-
3
B.
«»
m
a
o.
>-
Z •» CO «* -• <1
Ul
u
DC ^ Kl Kl Kl Kl Kl
f* Ifi 1^ M m Kl Kl
?< W Kl Kl Kl Kl Kl
w 6 « « co cj <* in r»
i i i
# N! N X X X X
u o m o o in •<
u. tr- »• o> -H
o
(0
J Mf»Kl>*Kl jP^j?
t >4>KICOK1 CM •» CM
II 1
Islsi uj-
X X X X X Z 1 O
omotno < KI o
o iv m w tx 0 c
»4
* *
* *
J
* *
*
* *
4
» *
** **.
H $
b * »
Ij * * * .
Q_ + *
>. ** S
K * *
1 *'
* +
I 1*:
1 *s**
* +
* +
* *
* * :
*
*
oooooooooo
•2S S^SS^^^TfT
•»
**
iu t- tn
3 Z 03
< O 2
> 0 N
z 5
M O
v> u
S N K T * *
= 3 !
a,*** * o o^— + — o o o *
§ i
(O + * +
in
I * O — CO
IU -O K> O> >•
in t-i o tf> — —
—• «-i * in o
in lu in z— —
fj
- O
5S
M 1-
111
in in c-
1 CM CO
111 * (A
C- Kl K
583.M
•
-
U.
Ul
^J
•
UJ
-------
I
K
IU
i
i
i>
in
in
O
N
v>
IU III
^ 1
U. H
o z
a.
1
(ft
lii
to
lu
£
tn
I
t- in N * in -»
to • »e> • **• ~*
UJ P* M) O » f1*
M *N ^4
X CM CM
M m m m in in
UI co co CO tO CO
3 N CM M CM CM
5 r^ ^ CM r* o>
<9 «^ ^ tO CM
1 1 1 1 1
^ i*> co •* N r*
•* 10 • o* m in
r* ^ ^ • co co
^ • »* • .
CM CM M *
b> in o o in **
O 0* O ••*
• o -o
.? r» CM »4 N CM N >O -«
** «o i *•* in •£
US
g«2-5 e°i
E c» j: t? i; IU-H z 5
13 O UJ M O
NXXXX 2io in u
o in o m o -i N
^•i ^L)lOQ.Q. ex
tA*A«Arstr^o^^ivinro
f^*-*»-*co »Ok*^cOfx|^>
l • CM
0 1/1 -O Kl .
• • o
CO CM
tn o ^ o -J
u uj z 3 u in
ox < a r ts
Z U I" 0
«COUJ(/> 1.Z7-Z
t*J H ii O > •» ID " 5 ••
y y trt tn *n 1 1 ^ in -.* g^
+ +
* *
J
* *
J *
* *
^ He
* *
+ *
H * «
a 4- *
s * *
0. * *
+ *
f J|
M * *
_J *
M » »
m + »
< * «
S »
*
Ex
* *
ik +
_i » *
•c * *
? * •*
£ * *
i 1*1
* +
^ +
» * *
* +
* *
*
*
*
in in m
« CO <»
CM i
i- * * *
o * *
Zj i i
n.Kt o — — +i — — o
gi l
i i
to * * •*
_ CM-4O^N^CM^KI
-, -. CM -t
O
0
f-t
•^
CM
CM
CM
CM
fO
O- •* -«
o> * —
co
in tt in K>
CM CM «0 IO
i> CM •*• rx in
< o -> CMlO«
^
in
-------
I
I
I
I
I
Example 4 SAS Regression Analysis
o •
in ••»
ft M
Ift •*
I
I
i
i
ooooooooooooeoooooo
ooooooooooooooooooo
OOOOOODOOOOOOOOOOOO
ooooooooooooooooooo
OOOQOOOOOOOOOOQOOOO
O CM O O
_l -H CM K>
(VI M M CJ
o o o o
o o o o
o o o o
o o o o
o o o o
o CM o o o -tf
* -f m -o f» r»
eg CM O O O
SIM
O-
I DC
a. iu
8?
IS
o o
r» i^
v* CJ
II M
2 Pi
* S*
a a
_ w ui
w iu in in
Sin tu
8° 00
8m (M CM
•* <
(OX II II
s
§§ £
55S w
ft ce in
I
M W CC
P 3 M P
Ik
in '
ec u »- i
-5252
Z O M
M •— O **
jut*~8
in o w in <
c «-• n: >
•-« < S -
uix--
in
u
^
CQ
a:
o o (nininininin«nv)ininm
C£OCZU-^-J->—I—I—I—'-I-I-J-1
in in
=
• cvi a
r M
in
Ji e t-
5SSg
tu ui a
t-
CO
O
. (n u
h- iu <
U -» Q.
ui a
ct < a
Sf-l UI
a >-
u < z
z > t-i
i-t a
•i a
z -o
O Q
tl1
O ^
W «T
CM Z O
CO O O
h- M -4
X >-
-II
gui
in in
fegl
U tft O
£ « tu
u v>
£3«
Sxe^
in o
u, o
o w o
»- V5
SJ
!;!
< HI
§in
ut
a
iu -*
a o- ••
< E o
> < o
M X W
Bui
I
-------
§
ot
ui
§
Ui
HI
IA
3
ID
§
o o o o o o
O •< ac o «
o <
ui o e o
& o t> iin
u. i»i zz
°f|3
^i,??,
••I-©*'
^ «< f*H Q I
§ ..£Srr|
•ui°-S
. i- a
CO O
u
IU
K
g
^
u>
s
£
fc <
o ^
o
• CM
X
g I-
u. o
u, E!
D UJ IU
-j a
i- o 5
< to o
3E ffi W
g s 8
"• o £
N
4) UJ
W *
S u
s g
<0
X>
o o o o
o o o o
o o o o
O o o o
c
o
iu
N
<
ui
(ft
IU
o
10
52 S
I £
< a.
z§
O M
ss
IU UI
m a
w 10
a ui
i rtf »M
M IU tO
Z " K O
M * < IH
<> o tn
< ui n iu
»- -• o a
< u, o n
°5S5
19 X UJ O
UJ UI o
*uj-.5
o -J iu a
O t- o t
« w o b
E K- E o
O -H W M
•»-»*-*
ifl
o
w •
"• IA
- t/>
,.»
S32
IU CD
cc < a
gt-l IU
or H
ssi
Mt>e
1WB
tl<
O *
in «o
CM Z O
CO O o
»- M -.
X I-
-g§
a. <
o: ui
o in !•*
o
t\l
tu n
o ->
O
5 1
0 Z
Z it
§
in a:
OJ UJ Q
IU > •
*?§
O UJ Z
ui t_ < .
h- < -> '
(J O "< •
iu o a i
glo . •
in o-1
a
o
•
S X
z u
&» -.
..
-i ii
a. iu o
?^
CO O
•3- ^
g
u
iu
to
u,
o
i
u.
IU
H
^
§
e
(M
•O
I
o
O
UJ .•
sg
si
IU CO
gg
a •<
iu in
x <
»- to
o
I
»*
PN
in
r-
cvi
M -> O O
I- O CO •
to a Z
z w x
MOO*
Q V
to cn a:
S< O <