The partitioning of copper, lead, and zinc between the water, colloidal and bed sediments as a function of pH is very evident from the combined water, colloid, and bed-sediment data in this report. In Cement Creek at Memorial Park, where the pH at low flow was 3.89, copper concentrations were high in the dissolved phase and labile copper in the bed sediments upstream from Memorial Park was in the range of 50-80 ppm (fig. 24D). In the upper Animas River above Silverton, where the pH at low flow was 6.82, copper concentration in the water were low whereas labile copper concentrations in the bed sediments were in the range of 300-500 ppm (fig. 26D). Smith and others (1992) demonstrated that the partitioning of metals in acidic, iron-rich waters followed a predictable path that could be modeled using sorbtion partitioning data and the iron hydroxide content of these waters (fig. 32A and B). As the colloidal content of the water decreases, the pH at which metals are sorbed by the colloids increases. We interpret the data from our study to indicate that copper is partitioned to the bed sediments through the sorption of copper to the colloids as a function of increasing pH. The aggregation and settling of the colloidal phase to the bed sediments results in a partitioning of the metal load from the dissolved and suspended colloids to storage in a static blanket of colloidal material in the bed sediments.

Dissolved and colloidal lead concentrations in the waters of Mineral and Cement Creeks and the upper Animas River are all less than the limit of detection indicating that lead is partitioned into the colloidal phase at low pH and precipitated to the bed sediments as soon as it reaches the streams. The dramatic differences in the lead concentrations in the bed sediments between Cement Creek (a gradual decrease from 600 to 300 ppm fig. 24E), the upper Animas River (generally greater than 1,000 ppm fig. 26E), and Mineral Creek (1,000 ppm at the headwaters decreasing to about 200 ppm at the confluence, fig. 22E) demonstrates how lead in bed sediments responds over a short distance to the changes in the sources of lead as was shown previously in the study of acidic streams at Mount Emmons near Crested Butte, Colo. (Gulson and others, 1992). This concept is particularly dramatic in the changes in the lead-isotopic compositions seen in bed sediments from the profile of Mineral Creek (fig. 23A and B).

Figure 31. Downstream variation of the iron to zinc weight ratio (A) in colloids and bed sediments of the

Animas River and tributaries (low flow, Oct. 1995), (B) downstream variation of the aluminum to zinc

weight ratio in colloids and bed sediments of the Animas River and tributaries (low flow, Oct. 1995),

and (C) downstream variation of the copper to zinc weight ratio in colloids and bed sediments of the

Animas River and tributaries (low flow, Oct. 1995).

Figure 32. Plots of the predicted sorbtion of metals on suspended iron hydroxides as a function of pH for two

different concentrations of iron colloids (HFO) A, model predictions at 1 g/L and B, model predictions at

0.01 g/L (from Smith and others, 1992).

Zinc concentrations in Cement Creek, where pH at low flow was 3.89, were 1,338 ppb, and zinc in the bed sediments is about 200 ppm above Memorial Park (fig. 24G). In the upper Animas River above Silverton, where the pH at low flow was 6.62, the dissolved zinc concentration was 319 ppb and zinc in the bed sediments was in the range of 3,500-1,000 ppm (fig. 26G) indicating that much of the dissolved zinc has been partitioned to the colloidal phase in the bed sediments as described above. This process also continues through the Animas Canyon reach (fig. 13A). An understanding of this dynamic process of partitioning metals between the dissolved, colloidal, and bed sediment compartments is necessary to evaluate the process of metal transport over the hydrograph in the Animas River watershed (fig. 9) and their impact on water quality.

The geochemical profiles showing the dispersion and dilution of metals in the bed sediments
of the Animas River below Silverton are convincing evidence that the source of these ore
metals in the Animas River watershed is above Silverton. A very important question that
should be addressed is:

Can we isolate the contributions of the three major tributaries, Mineral and Cement Creeks

and the upper Animas River, to determine which tributary provides what portion of the metals

in the bed sediments?

There are two basic approaches to this fundamental question: mass balance calculations like
those outlined in equation 1 and lead-isotope calculations as outlined in equation 2. We
will use a combination of these two approaches to arrive at an estimate of the sediment added
by Mineral and Cement Creeks to the total mass of the bed sediment in the Animas River below
Silverton.

The mass balance or mixing calculations are straight forward. The total concentration of any given metal in the bed sediments is defined by:

where: **T(Me) _{AB}** is the total concentration (ppm) of
the metal

**X** is the fraction of the total metal concentration
contributed by bed sediments from Mineral Creek,

**(Me) _{MC}** is the concentration of the
metal

**Y** is the fraction of the total metal concentration
contributed by bed sediments from Cement Creek,

**(Me) _{CC}** is the concentration of the metal

**Z** is the fraction of the
total metal concentration contributed by bed sediments from the upper
Animas River above the confluence with Cement Creek, and

**(Me) _{UAR}** is the concentration of the
metal

The solution to this equation requires either an independent determination of one of the
variables **X, Y,** or **Z,** or a series of equations which can be solved
simultaneously. The critical question in this approach to ask is:

What metal concentrations have sufficient variation and provide mathematical leverage in

this calculation?

Since there is no significant variation in the geology of the subbasins, there is no significant variation in any of the lithophile element concentrations that can be used to derive percentages of materials added to the Animas River by either Mineral or Cement Creeks. Labile copper and zinc are actively precipitated from solution immediately below the confluence of Mineral and Cement Creeks, so the concentrations of these metals in the bed sediments are highly variable and cannot be used in the calculations. Labile lead concentrations, however, are stable as lead is present in both the colloidal phase and bed sediments in all three drainages.

Lead-isotopic data provide a powerful tool in evaluating the relative contributions of various sources to the bed-sediment concentrations of various metals. As the primary focus of this study was to evaluate the transport and partitioning of metals between the water column and the bed sediments from the Silverton area downstream, it is essential that we have a measure of the amount of metal added to the bed sediments by tributaries along the course of the Animas River. The measurement of the mass of sediment added by individual tributaries is possible using the lead-isotope ratios. The lead-isotope ratios are a function of mass rather than a function of concentration, allowing us to calculate the relative mass of lead added by individual tributaries that causes changes in the lead-isotopic ratios. These percentages are calculated using the following equation:

where: **PC** is the percent of the metal derived from the contaminant source,

**R _{B}** is the

**R _{T}** is the

**R _{C}** is the

Similar calculations can be made using the^{ 208}Pb/^{204}Pb. Generally,
there is not sufficient analytical resolution to make accurate calculations using the
^{207}Pb/^{204}Pb value unless the source rocks are more than two billion
years old because of the short half-life of the parent isotope for ^{207}Pb,
^{235}U.

The amount of sediment added to the Animas River by
Cement Creek cannot be determined using the lead-isotopic data because there is no
lead-isotopic difference between the lead-isotopic composition in the labile phase of the
bed sediments of Cement Creek and that in the Animas River. There is, however, a difference
between the lead in the labile phase of the bed sediments of Mineral Creek and the Animas
River. Since there is virtually no change in the isotopic composition of labile lead of
the Animas River between site 95ABS113 above the confluence of Cement Creek and site A-72
below the confluence of Mineral Creek, it follows that the amount of labile lead added by
Mineral Creek is small. We used equation 2 to determine the value of **X** in equation 1. The
lead-isotope calculation indicates that about 3 percent of the total mass of the labile lead
in the bed sediments at site A-72 was derived from Mineral Creek. This number is subject to
significant measurement error due to the very small difference in the lead-isotopic values
measured in the sediments of the Animas River. Given these analytical limits, we can use
the lead-isotopic data to assign an upper limit on the contribution of metals from the bed
sediments of Mineral Creek of less than 10 percent of the total labile lead in the Animas
River at site A-72. In reality, the number may be closer to 5 percent.

The chemical data can then be used to estimate the
contribution of Cement Creek to the Animas River from equation 1. As was shown earlier,
both zinc and copper are being actively precipitated from the water below both the confluence
of Cement and Mineral Creeks with the Animas River. Thus, the kinetics of the settling
process and the variability that this introduces into the sampling precludes the use of
either the copper or zinc data from the 2M HCl-1%H_{2}O_{2} digestion in
such a calculation. Only the zinc in the sphalerite component of the sediments is a stable
chemical variable in this mixing zone. Unfortunately, there is little variance in the
zinc data. The zinc in sphalerite from the Animas River is 1,200 ppm, from Cement Creek is
1,000 ppm, and in the Animas River below Silverton at site A-72 is also 1,000 ppm. There
is, however, significant leverage in the lead concentration data which, when combined with
the determination of the limits of the sediment contribution from Mineral Creek using the
lead-isotopic data, can be used to make a mass balance calculation. The lead concentrations
are as follows:

Solving equation 1 and using 5 and 10 percent for the contribution of metals from the bed sediments from Mineral Creek based on the lead data from equation 2, we estimate that Cement Creek contributes about 35 to 40 percent of the total metals in the bed sediments of the Animas River below Silverton at site A-72. The copper data in the insoluble phase give a similar value although the copper data are subject to a much larger fundamental sampling error because the insoluble copper is contained largely in a heavy-mineral phase. If the chemical data had been based on a 50-g sample, the copper data would be more robust.

The lead-isotopic data can also be used to calculate
the contributions of metals to the labile phase of the bed sediments for the Animas River
below Silverton. We will use the lead-isotopic data from site A-72 below Silverton as the
composition of the contaminant, **R _{C}**. Below Silverton, the composition of
the lead
supplied by the tributaries is different for the areas underlain by Precambrian, Paleozoic
and Mesozoic, and Tertiary rocks. The areas are shown by the model for this calculation in
figure 33A values used for

____________________________________________________________________________________________________________________________ | ||||

Locality |
Distance (km) | R |
R |
PC |

____________________________________________________________________________________________________________________________ | ||||

Silverton (A-72) | 26.26 |
100 | ||

KOA Campground |
72.55 |
20.35 |
18.487 |
96 |

Trimble Bridge |
82.35 |
19.25 |
18.510 |
88 |

Durango, 32 |
100.8 |
19.25 |
18.540 |
84 |

Durango, Red Lion Inn |
105.0 |
19.25 |
18.559 |
82 |

Above Lightner Creek |
106.2 |
19.20 |
18.592 |
78 |

Weaselskin Bridge |
122 |
19.20 |
18.586 |
77 |

Above Bondad |
135.5 |
19.20 |
18.637 |
71 |

Cedar Hill |
139.5 |
19.20 |
18.684 |
65 |

Aztec, NM |
171.5 |
19.00 |
18.661 |
57 |

_____________________________________________________________________________________________________________________________ | ||||

model (fig. 33).