Chedeski Fire . Nutrient Loading into Roosevelt Y Summer 04 - - PowerPoint PPT Presentation

Eutrophication of the Salt River Reservoirs due to the Rodeo- Chedeski Fire.

Nutrient Loading into Roosevelt

Summer 02 Fall 02 Winter 02/03 Spring 03 Summer 03 Fall 03 Winter 04 Spring 04 Summer 04 Sampling_Period 5 10 15 20 25 30 35 Y Y Mean(Ammonia_N_mgPerL_asN) Mean(NitrateNitrite_N_ppm) Mean(Total_P_ppm) Mean(Total_Kjeldahl_Nitrogen_mgPerl_as_N

TOC/DOC in the Salt River above Roosevelt

Summer 02 Fall 02 Winter 02/03 Spring 03 Summer 03 Fall 03 Winter 04 Spring 04 Summer 04 Sampling_Period 5 10 15 20 25 30 35 Y

OverlayChart

Y Mean(TOC_ppm) Mean(DOC_ppm)

Chart

Summer Nutrient Levels from Roosevelt (mean for all sites)

Summer 02 Summer 03 Summer 04 Sampling_Period .0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1.0 1.1 Y Y Mean(Ammonia_N_mgPerL_asN) Mean(NitrateNitrite_N_ppm) Mean(Total_P_ppm) Mean(Total_Kjeldahl_Nitrogen_mgPe

0.1 0.4 0.5 0.6 Summer 02 Summer 03 Summer 04 Sampling_Period All Pairs Tukey-Kramer 0.05 Rsquare Adj Rsquare Root Mean Square Error Mean of Response Observations (or Sum Wgts) 0.767716 0.758782 0.056288 0.189455 55

Summary of Fit

Sampling_Period Error

• C. Total

Source 2 52 54 DF 0.54452819 0.16475544 0.70928364 Sum of Squares 0.272264 0.003168 Mean Square 85.9318 F Ratio <.0001 Prob > F

Analysis of Variance

Summer 02 Summer 03 Summer 04 Level 13 27 15 Number 0.356923 0.167407 0.084000 Mean 0.01561 0.01083 0.01453 Std Error 0.32560 0.14567 0.05484 Lower 95% 0.38825 0.18914 0.11316 Upper 95% Std Error uses a pooled estimate of error variance

Means for Oneway Anova Oneway Anova Oneway Analysis of DO_mg_per_L By Sampling_Period

DO_mg_per_L 0.2 0.3

Components: Chl_a_mgPerm3 DOC_ppm TOC_ppm Ammonia_N_mgPerL_asN NitrateNitrite_N_ppm Total_P_ppm Total_Kjeldahl_Nitrogen_mgPerl_ Prin Comp 1 Prin Comp 2 Prin Comp 3 Prin Comp 4 Prin Comp 5 Prin Comp 6 Prin Comp 7 Chl_a_m DOC_ppm TOC_ppm Ammonia Nitrate Total_P Total_K x y z

Spinning Plot

PCA of Primary Production in Roosevelt

Mean Hypolimnetic DO Levels by Reservoir

DO_mg_per_L 1 2 3 4 5 Apache Canyon Roosevelt Saguaro Reservoir Rsquare Adj Rsquare Root Mean Square Error Mean of Response Observations (or Sum Wgts) 0.022082 0.009755 1.07882 0.719421 242

Summary of Fit

Reservoir Error

• C. Total

Source 3 238 241 DF 6.25467 276.99705 283.25172 Sum of Squares 2.08489 1.16385 Mean Square 1.7914 F Ratio 0.1495 Prob > F

Analysis of Variance

Apache Canyon Roosevelt Saguaro Level 105 13 58 66 Number 0.73124 1.10077 0.47690 0.83864 Mean 0.10528 0.29921 0.14166 0.13279 Std Error 0.52383 0.51133 0.19784 0.57704 Lower 95% 0.9386 1.6902 0.7560 1.1002 Upper 95%

Means for Oneway Anova Oneway Anova

Mean Summer Hypolimnetic DO Levels for all Salt River Reservoirs by Year

Summer 02 Summer 03 Summer 04 Sampling_Period .0 .1 .2 .3 .4 .5 .6 .7 .8 .9 1.0 Mean(DO_mg_per_L) Sampling_Period Summer 02 Summer 03 Summer 04

Chart

PCA of Primary Production in Apache, Canyon, and Saguaro

Components: Chl_a_mgPerm3 DOC_ppm TOC_ppm Ammonia_N_mgPerL_asN NitrateNitrite_N_ppm Total_P_ppm Total_Kjeldahl_Nitrogen_mgPerl_ Prin Comp 1 Prin Comp 2 Prin Comp 3 Prin Comp 4 Prin Comp 5 Prin Comp 6 Prin Comp 7 Chl_a_m DOC_ppm TOC_ppm Ammonia Nitrate Total_P Total_K x y z

Spinning Plot

Mean Chlorophyll a values for Apache, Canyon, and Saguaro Reservoirs by Season and Year

Summer 02 Fall 02 Winter 02/03 Spring 03 Summer 03 Fall 03 Winter 04 Spring 04 Summer 04 Sampling_Period 10 Mean(Chl_a_mgPerm3)

• Autochthonous processes within the

reservoirs may mean eutrophication proceeds unabated long after nutrient loading via the Salt River has diminished.

• This will have consequences, some

severe and others subtle, on water quality entering the Valley for some years to come.

Algal Toxins in the Salt River Reservoirs

• We routinely sample for anatoxin-

a, microcystin, and cylindrospermopsin.

• We first discovered C. raciborskii in

Arizona in 2001.

• Numbers have increased in all

reservoirs surrounding the Valley since that time

Analytical Methods

• Anatoxin-a, Saxitoxin

– HPLC after fluorescent derivatization.

• Microcystin

– Protein phosphatase inhibition assay.

• If greater than 0.5 µg/L, confirmed by

HPLC using a PDA detector.

• Cylindrospermopsin

– HPLC using a photodiode array detector

• Detection limit for all assays is less than 0.1

µg/L

Fish Kills

• First major fish kill occurred in Apache

in March of 2004.

• Subsequent fish kills occurred in

Canyon, Saguaro, and again in Apache throughout the spring and early summer.

• Multiple species involved.
• All water quality variables were

“normal”

• A major fish kill occurred in the

riverine portion of Saguaro on 6/10/04.

• Smaller fish (e.g., threadfin

shad) were noticed dead or moribund in Canyon on 6/9/04.

• Relatively large background levels
• f microcystin in all watersheds.
• Etiology of the fish kills implicate a

fast-acting neurotoxin such as anatoxin-a.

• While levels of C. raciborskii

steadily rose throughout the summer of 2004, only very low levels of cylindrospermopsin have been found.

Persistence/Degradation of Toxins

• Both cylindrospermopsin and

microsystin are environmentally stable compounds.

• Anatoxin-a, however, is rapidly

degraded by sunlight and alkaline conditions with a half life of perhaps

• nly a few hours.

Anatoxin-a

• Potent neurotoxin which causes

rapid death by respiratory arrest.

• Postsynaptic, depolarising,

neuromuscular, blocking agent that binds strongly to the nicotinic acetylcholine receptor.

• Produced by species of Anabaena,

Aphanizomenon, Oscillatoria, and Microcystis.

• No anatoxin-a found in aqueous

samples.

• However, anatoxin-a found at

toxic levels in stomachs of fish.

• Non-detectable amounts of fast-

degrading toxins in aqueous samples can be dangerously misleading.

Potentially Toxic Cyanobacteria Found in Salt River Reservoirs

• Aphanizomenon flos-aquae
• Anabaenopsis circularis
• Anabaena laxa
• Anabaena schremetievi
• Anabaena torulosa
• Anabaena variabilis
• Cylindrospermopsis raciborskii
• Merismopedia elegans
• Microcystis
• Pseudanabaena
• Oscillatoria aghardii
• Oscillatoria limnetica
• and several more

• It is impossible to determine toxicity based

upon presence of an algal species alone.

• The only way to quantify algal toxins is

through direct measurement of either aqueous

• r biological samples.
• Most of the cyanobacteria found within the

reservoirs are ubiquitous and probably do not produce toxins the majority of the time.

• Based upon our large database of

algae identifications, there is NO correlation between numbers of potentially toxic species and toxic events.

Why Were the Toxic Events Worse in the Upper Reaches of Saguaro?

• Unknown but pump-back storage at

Canyon may play a role.

• This area has had other toxic events

and in 2001 we found over 140 µg/L

• f anatoxin-a.
• This was the highest level of

anatoxin-a ever recorded by the reporting lab.

Toxicity based upon Environmental Conditions

• No correlation to toxicity and

number of species suggests that a few of the suspect species produce copious amounts of toxin at a specific time based upon environmental conditions.

Allelopathy

Defense from Grazing by Zooplankton

Why were no Humans Affected?

• Fish and mollusks are especially

susceptible due to rapid uptake across gills.

• Just because toxicity occurs in fish

does not mean toxicity will occur in humans.

• However, fish and zooplankton serve

as important biological indicators of toxicity.

Algal Toxin Summary

• Fish kills probably caused by anatoxin-a.
• Possibly exacerbated by lysing of cells

due to pump-back storage.

• Several potentially toxic species found in

ALL reservoirs surrounding the Valley and no correlation between biomass and toxicity.

• Toxicity probably due to environmental

factors such as removal of nutrient limitation, allelopathy, defense from grazing, etc.

• C. raciborksii probably played no role in

toxic events.

• Dr. Paul Zimba (USDA) growing 2

isolates of C. raciborskii to check for toxicity.

• Unialgal cultures of all potentially

toxic species need to be established and then systematically checked for toxin production under different environmental conditions.

• Without this data, predicting future

toxic events by looking for any individual species is meaningless.

Special Thanks

• Susan Fitch, Linda Taunt, Jenny

Hickman, Sam Rector, and Amanda Fawley from ADEQ.

• Marc Dahlberg, Kevin Bright, and

Larry Riley from AzG&F.

• Dr. Greg Boyer from SUNY

Questions?