(Comments are in black, questions in red, and suggested modifications in blue.)

• I applaud the multidisciplinary approach used to track potential improvements in water quality resulting from restoration. However, the current sampling design does not include an important component of this ecosystem – stream algae (periphyton) and macrophytes. In addition to the massive algal blooms that regularly produce episodes of poor water quality including high pH and low dissolved oxygen during summer in Agency and Upper Klamath Lakes, impressive proliferations of periphyton also occur in portions of the Wood River Valley. Nutrient concentrations are probably quite high, so algae, particularly in well-lighted stream reaches, ditches, and irrigation canals. The lack of periphyton monitoring represents a significant obstacle to understanding the nutrient dynamics of the system, a limitation that may result in underestimating potential water-quality benefits from restoration, given the current frequency of water sampling that can miss episodic losses and transport of algae from the streambed. From other studies I’ve conducted in other western watersheds, in the early spring, sometimes quite early, I’ve observed large growths of periphyton in nutrient enriched streams receiving adequate light. During this initial growth phase, periphyton actively take up dissolved nutrients (from the stream or ground water seepage), sometimes depleting concentrations in the streams to levels below detection. In this way, periphyton can be a large nutrient sink, especially for nitrogen– at least initially. Later, typically in June or July, algal productivity often slows in response to factors including lower water velocities (slower nutrient delivery) and/or lower nutrient concentrations. Decaying periphyton may then be transported downstream in floating mats that release nutrients through decomposition. In this way, nutrients from decaying periphyton would be available for uptake in downstream reaches, including blue-green algae growing in Agency and Upper Klamath Lakes. The biological oxygen demand (BOD) that also occurs from the decomposition of algae, particularly during early summer, may result in unfavorably low concentrations of dissolved oxygen, a condition that is exacerbated by the increasing water temperatures. In order to truly measure improvements in water quality resulting from restoration, particularly nutrient concentrations or loads, I suggest including seasonal measures of periphyton biomass (chlorophyll a), making sure to sample about every 2-4 weeks, including early spring. I also suggest doing surveys for macrophytes beds in the study area. If funding permits, also including algal/macrophyte species composition could provide insights into the factors affecting other important components of aquatic ecosystems, such as the invertebrate communities (I would also suggest expanding the network of sites currently being sampled for invertebrates and would add algae).

• I also suggest measuring basic field parameters (temperature, pH, and dissolved oxygen) in the early morning and late afternoon (diel measurements) to capture extremes in these parameters.

• Such measurements can be correlated with algal biomass values for a given reach to evaluate the effects that periphyton has on pH and DO. For this purpose, the use of Quanta multiparameter instruments, however, is not recommended. I have found Hydrolab Quanta performance to be unreliable and unstable, when compared with other higher-quality Hydrolab instruments, such as the Hydrolab DS3 or H20 units. The authors acknowledge the "unsatisfactory" performance of the Quanta for turbidity, but I would add that other data on other parameters, especially dissolved oxygen and pH, might also be unreliable or otherwise inaccurate. It is recommend that proper quality assurance and verification of these data, including independent measures of pH (with separate unit) and periodic dissolved oxygen checks using the Winkler method, would improve data quality.

• Laboratory methods and reporting limits employed for nutrient analyses were sufficiently low to evaluate the nutrient status of the system, and can be used to evaluate potential biological processes, such as uptake of nutrients by algae, denitrification, etc. In the methods, however, it wasn’t stated how samples were filtered, although table 1, p. 36 of the 2003 proposed work indicates that SRP samples do require filtration. However, the table on p. B-6 indicates that the samples were not filtered.

• All samples being analyzed for soluble reactive phosphorus, nitrate-plus-nitrite, and ammonia should be filtered in the field at the time of collection to avoid any potential uptake or transformation of nutrients prior to analysis. Filtration is recommended in Standard Methods for ammonia (method 4500-NH3H), nitrate (4500-NO3F), and SRP (4500-PF). The reporting limits for nutrients presented in the table on page B-3 (2003 Monitoring Plan) indicates a detection limit of 0.010 mg/L. However, in Standard Methods, it is reported that for the method 4500-NO3F (the method employed) the range is 0.50 to 10 mg/L. I didn’t see QA results for standard reference samples that would verify the accuracy of this method at levels near the reporting levels (ca, 0.010 mg/L). Also, according to the table on page B-3, nitrate + nitrite samples can be acidified. Standard Methods does not mention the use of acid for the method used. Could the acid preservation apply to the ammonia samples, not the nitrate?

• Also, it was reported that samples were received by the lab 1-5 days after collection, but actual analysis date is not given (Appendix D, p. 1-5). What were the actual holding times? This may be important, especially if the samples were not filtered in the field. Some additional quality assurance data, including equipment blanks, is recommended.

• In addition to the suite of chemical constituents already being tested for, a few additional samples for organic carbon (total and dissolved), alkalinity, and chloride could be used to help track water sources (Cl^-) and understand how a given algal biomass affects pH (alkalinity). This could help answer questions including "how much production can a reach handle without adverse effects on pH?" Also, because the much of the low-lying areas in the Wood River Valley were historically wetland, they may contain peat soils that upon decay release tannins that stain the water. This is quite noticeable in the Wood River Wetland. It is hypothesized by some that water rich in dissolved organic carbon, such as that from many of the wetlands around Upper Klamath and Agency Lakes, may curtail or prevent the growth of Aphamizomenon, through reducing the amount of light into the water, lowering the pH, or from some other mechanism. Sampling both concentrations of dissolved organic carbon and the character of the carbon (through various fractionation techniques) could be an important component of managing both the uplands in the Wood River Valley and the Wood River Wetland, which is now being managed by the BLM for water storage, water quality, and wildlife habitat goals. If the dissolved organic carbon in the wetland, Sevenmile Canal, or Wood River is shown to inhibit or slow the growth Aphamizomenon, then management of this area could include production of t his particular type of carbon. Sampling might also identify which plant assemblages or soil types are associated with such effects, which could then help guide future restoration activities.

• As a last thought, given the recent acquisition of two large former wetlands (the Wood River Wetland (WRW) by the Bureau of Land Management and the Agency Lake Property managed by the Bureau of Reclamation) that are ripe for restoration, new opportunities for creative wetland management may emerge. For example, the BLM occasionally takes agricultural return-flow water into the WRW from the adjacent property to the north, which is flood irrigated during spring, and either drained to the WRW or pumped into Sevenmile Canal shortly thereafter, and grazed by cattle. The WRW currently supports dense growths of algae, and could become an important location for processing nutrients and sediment from agricultural return water. In addition to inputs from the adjacent property, the WRW might also accommodate additional inputs of irrigation return water via Sevenmile Canal. Under this scenario, nutrients assimilated by the wetland by plant uptake (algae and wetland plants), loss of N through denitrification, and retention of P in the sediments could be held in the WRW until fall, after the algae blooms in Agency and Upper Klamath Lake have subsided. One of the key elements in such a "treatment wetland" approach is knowing what water quality is like in Sevenmile Canal, and knowing how much nutrient-processing capacity the wetland has. There may be a threshold in the assimilative capacity or maximum amount of nutrients the wetland can process without becoming overwhelmed or out of balance. This capacity may change as the WRW matures, and striving to maintain proper ecosystem function will be one of the challenges managing the wetland. In the future, high-nutrient return flows from the remaining irrigated parcels in the Wood River Valley, for example, might also be routed into the Wood River Wetland to be processed, while low-nutrient snowmelt is sent into Agency Lake or to the Agency Lake Property managed by the BOR. The water-quality, biological, and habitat data proposed here will be an important source of high-quality data that will benefit management of the WRW and Wood River Basin as a whole, hopefully allowing for decisions on water to be science-based.

p. ix, last bullet.. see p. 26-27 below

p. x, first bullet.. the most likely source of particulate P is from erosion that results from grazing. While this may be true, another potentially important source of particulate phosphorus, especially in this system, is sloughed algae. Does erosion include manure?

p. 23, first paragraph addition to soil and nutrients, bacteria associated with cattle also can contribute to water quality problems such as low dissolved oxygen levels. Pilot sampling for bacteria (e.g., E. coli and fecal coliform) at a few locations could be used to identify whether this is an issue or not. Bacteria data might exist already, I don’t know.

p. 24 .. 1) High pH (ca 9.0 units) also can damage brook trout gill lamellae (possibly other salmonids too); 2) Observation: During 2002 sampling occurred late during the season (started in July). I’d suggest that sampling begin in April or early May at the latest. Just this year (2003), I observed a long (6 foot) bright green streamer of filamentous algae (probably Cladophora) at the mouth of Wood River (WRDIKE). By June, however, it was brown and nearly gone; 3) Incorporating longitudinal measures of conductance and temperature in synoptic fashion (i.e., Lagrangian sampling) could identify or confirm dramatic changes in water quality due to changes in land use or from inputs of nutrient-rich ground water. Artesian well water in the area is rich in both SRP and ammonia (~6 mg/L).

p. 25, last sentence..it states that the van Dorn sampler was used when high algal biomass was visible. When and at what sites was this observed? I’m glad that your sampling includes this aspect. This fraction is often missed.

p. 26-27 ..It is stated that phosphorus concentrations increased in the Wood River, Sevenmile Creek, and Annie Creeks in response to land use (grazing induced erosion). While this may be true, the biggest increases in phosphorus occurred at the mouth sites at the Wood River Wetland dike, where water chemistry can be influenced by Agency Lake. The authors acknowledge this, and have installed Doppler velocity gages at these sites. Further, the steady downstream increase in SRP, especially in the Wood River Basin, also may reflect inputs of phosphorus-rich groundwater. Also, couldn’t the increase in particulate P also come from manure?

p. 27..end of third full paragraph ..although return flow from Annie Creek probably increases the nutrient load downstream in Sevenmile Canal, sampling Sevenmile upstream from the dike road (at the Wood River Wetland inlet, for example) might avoid influence from Agency Lake, although this site also may be affected.

p. 28 second full paragraph ..significant warming in this reach".. Warming here also could also be due to the effect of Agency Lake, as mentioned above, but also to the relatively high concentration of dissolved organic carbon in this reach. The stained water efficiently absorbs heat, producing unusually high temperature at the surface compared with the bottom.

p. 28 under pH …Was alkalinity measured? It might help explain why pH fluctuations were dampened at WRDIKE. It could also have been the time of year (periphyton can peak in the early spring). High pH (ca 9.0 units) also can damage brook trout gill lamellae (possibly other salmonids too).

p. 35 third paragraph ..with respect to the "reference reach" sampled for invertebrates.. with the majority of benthic invertebrates consisting of amphipods, chironomids, and oligochaetes, this doesn’t sound like a reference stream. The 9% EPT at this site (9% mayflies) -- were they mostly Baetid mayflies? These are also tolerant. The benthic invertebrate work might be improved by including woody debris as an additional substrate type to be sampled. In many systems, submerged woody debris is the richest habitat for both invertebrates and algae. In some cases, overhanging vegetation also may be the richest habitat for invertebrates. In general, depositional habitats are much less diverse. Were the chironomids identified to species? Given the massive biomass (clouds) of midges in this area, getting this level of identification might better distinguish reference site condition from impacted sites, especially if depositional areas continue to be the only habitat sampled for invertebrates.

figs. 6-1, 6-2 ..1) I’d prefer to see the various components of N in separate plots, instead of combining them into TN. Actually, I’m not sure what TN, NH4, NO3 represents. The persulfate TN includes all N in the sample, so it’s confusing that ammonia and nitrate are also listed on the axis label. Might the black bars represent both forms of inorganic N and SRP? Can you label the plots as such? Again, I’d prefer to see the N plots separated, in order to help sort out potential N sources or transformations, such as nitrification; 2) The sampling dates portrayed are confusing. Although sampling occurred on 7/24, it looks like it was done over a week-long period; 3) I like the downstream trending bars, but TESP is a spring, and takes away from longitudinal pattern. Can you change TESP from a bar to a symbol? 4) Where there’s large variations among sites (e.g., TN and TP, fig. 6-1 and fig. 6-2) I’d suggest noting scale differences somewhere on the plot, its quite large in some cases; 5) In some cases, it looks like sites were cut off (TP concentration for ACAW and SMFS in fig.6-1); 6) Having some indication on the plot as to whether irrigation was occurring would make a nice addition.

fig. 6-14 ..is there any confirmation, such as flow data from the Doppler unit at WRDIKE, that might confirm whether data collected at this site was truly Wood River water, not Agency Lake water?

p. 3 fourth paragraph ..Report states that Wood River at Dixon and Weed Roads have been kept in operation. Two paragraphs previous, however, it states that these gages were continued until 1996-97.

p. 4 .. on meeting Objective 2 (quantifying tailwater) ..is a tall objective. How many weirs can you conceivably install?

Comment on fig. 3-1 ..Expand site labels for the Sevenmile Creek sites (presently, the Sevenmile Creek at Sevenmile Road is not labeled as such.

p. 7 second paragraph ..Why not include Mares Egg spring? The name implies blue-green algae like this water, which might mean its high in phosphorus. Is it?

p. 21 ..near the bottom ..on the use of bi-weekly.. restate as "every two weeks"

p. 22 .. under turbidity samples ..If you have to reduce the amount of sampling, retain storm sampling if possible.

p. 23 ..under continuous monitoring ..I suggest increasing the frequency of continuous measurements for diel parameters up to a maximum of 30 minutes between samples. Although hourly data is good, 30-min readings, especially for temperature, DO, and pH would be better. Also, consider adding chlorophyll a as an additional parameter to be measured continuously.

p. 24 second paragraph ..in addition to the point QA checks with the NIST thermometer at the continuous monitor, do cross sectional and local longitudinal temperature measurements to ensure the monitor is located in a representative area. With all the ground water inputs and local surface warming (especially in areas with high dissolved organic carbon), there could be considerable local variations in water temperature.

p. 24 nutrient sampling procedures .. I like the use of the van Dorn to capture coarse particulate organic matter, but certain particulars, such as how samples will be filtered, are not included. You’re reporting SRP, so I’m assuming you did filter the samples. Also, I see that the lab does native sample spikes and replicates, and that 10% of samples also will be true field replicates. This level of QA seems appropriate, given the good QA seen from the lab during 2002. In addition to these QA samples, I would suggest adding both equipment blanks and laboratory blanks (bottle blank to check lab, not equipment) for nutrients, and if the standard reference samples are not already being submitted blindly, I would do so. I suggest including reference samples that are representative of the entire data set, especially to verify accuracy of low level NO₃ and NH₄ values. I didn’t see QA results for such samples.

General comment: with the potential to increase the amount of woody debris in these streams through restoration, it would be good to including sampling the wood for benthic invertebrates.

p. 27 ..Do the aerial surveys include the Wood River Wetland? If so, I’m interested

p. 30 ..longitudinal profiles of water depth/chemistry also may be obtained using a small Doppler unit/hydrolab that can be towed with a kayak.

p. 34 ..requiring that pools be 6 feet deep seems to be a stringent requirement. Trout like pools much less deep than 6 feet, especially if there’s some form of cover.

Appendices:

B-3, top table ..shows that nitrate + nitrite samples can be acidified and held for 28 days (as mentioned in the General Comments section, this may be in error). Holding nutrient samples for this long may give inaccurate values. What were the actual holding times? (they are not reported in either document). Also, NH₄ samples are analyzed with EPA method 350.1, which includes preservation with sulfuric acid and isolation with distillation in boric acid. Was this actually done? (DEQ doesn’t do it as part of their typical protocol, and they also use the EPA 350.1 method).

B-7, calibration procedures ..look good, except that it states that periodic recalibrations will be performed. I would suggest that daily calibrations for dissolved oxygen (DO), pH, conductivity, and turbidity be done on all instruments (making sure to bracket the observed values). Also, recheck the DO at each site and recalibrating as necessary, particularly if elevation is much different from where it was last calibrated. For continuous monitors, pay close attention to fouling. Typically, only our most pristine sites can go more than 2 weeks without cleaning. Also, I would re-calibrate, even if the values are within 10-15%. Also, see comments about using Quanta Hydrolabs in the General Comments section.

B-7, cross-sectional measurements ..also evaluate local longitudinal variations in field parameters, including potential variations with depth (especially where high levels of dissolved organic carbon occur).

C-14, disposal of formazine ..this is a known carcinogen and should be disposed of as hazardous waste, not down the drain.

Turbidity monitoring at continuously operated sites has been discontinued due to the poor performance of the Quanta turbidimeter. Calibrations of the turbidimeters that are used to analyze turbidity for the turbidity samples collected with the DH-48 are performed at the GMA lab where all regulations for the disposal of hazardous chemicals are followed.