                   Responses to Earthjustice's Comments on
                      the Department's Reference Sites
                                 Draft 10-2-09
                     FDEP Standards and Assessment Section
                                       
This document provides the Department of Environmental Protection's (DEP) responses to Earthjustice's recent comments on the reference sites that DEP selected for development of numeric nutrient criteria.  Earthjustice's main premise seems to be based on a position that any minor disturbance on the landscape should prevent a site from being categorized as a nutrient reference site.  However, it has never been either EPA's or DEP's intent (or contention) that the selected reference sites be "pristine" and completely devoid of anthropogenic sources of nutrients.  Our goal has been to identify sites that are minimally affected by humans and that have been demonstrated to sustain aquatic life use support.  Further, some of Earthjustice's presumptions about silviculture and potential sources of nutrient loading are not supported and are not consistent with the available information. 
 Designated Use and Reference Condition
The Clean Water Act directs States to establish water quality standards to protect the designated use, in this case, healthy, well balanced aquatic communities.  DEP followed an extensive process to identify candidate reference sites and then demonstrated that the nutrient reference sites had optimal habitat, healthy invertebrates, and no issues with algal or plant biomass.  However, in an abundance of caution, DEP conducted site visits and biological assessments, and omitted sites with greater than minimal landscape disturbance near the stream or sites where biological conditions that did not meet the use.   As a result of this vetting process, DEP can confidently conclude that the nutrient concentrations of the reference site population were characteristic of the levels needed to sustain the designated aquatic life use support.  
Earthjustice's comments seem to imply that the CWA directs states to establish criteria based on conditions that occurred prior to human inhabitation of North America.  For example, Earthjustice contends that Orange Creek, Little Orange Creek, and the Waccasassa River are too disturbed by humans for use as nutrient reference sites based on landscape disturbance.  However, DEP evaluated both the land uses and biological condition of each site (see evaluations of each site later in this document) and these sites were characterized by benign land uses and exceptional biological health with mean SCI scores above 64.  
Landscape Development Intensity Index and Reference Condition

DEP used the Landscape Development Intensity Index (LDI), which is highly correlated with biological health, within a 100 meter corridor as a starting point for candidate reference site selection.  DEP selected 2 as a screening level because there is no trend in the relationship between SCI score and LDI for LDI values below 2 (see figure below and DEP's Technical Support Document (TSD)).  
                                       
                                         
   Stream Condition Index vs. Landscape Development Intensity Index.  Within range of low LDI scores, there is no relationship between LDI and SCI.  SCI shows significant decreases  with LDI in higher LDI ranges (see DEP Nutrient TSD).

The use of a 100 meter natural buffer as a factor that would mitigate nutrient inputs is clearly established in the literature, and a review of this information was included in DEP's TSD.  As noted in the TSD, the farther a particular land use is to a stream, the less influential that land use is on the stream.  The literature (including EPA sources) also clearly indicates that riparian buffer zones are very effective in reducing nutrient inputs to near natural background levels.  However, in an abundance of caution, DEP rejected many sites that met the preliminary criterion of <2 LDI to be certain of the "minimal disturbance" goal.  

Earthjustice incorrectly argues that the LDI coefficient for silviculture (1.6) did not include fertilization as part of the emergy calculation, and that this oversight makes forestry operations a more adverse land use than the LDI would indicate.  University of Florida Professor, Mark Brown, confirmed on 9-30-09 that forest fertilization was accounted for in the LDI calculation and that an application rate of Total Phosphorus of 142 pounds/acre over a 25-year rotation was assumed (Doherty 1995, Pritchett 1983).  This issue will be discussed further below.
To confirm the effectiveness of the LDI corridor approach, EPA's contractor, Tetra Tech, conducted an analysis to determine if landscape disturbance beyond the 100 meter buffer was associated with additional risk of nutrient enrichment.  That analysis supported the use of the 100 meter corridor approach:  "For the 135 WBIDs evaluated, the near field LDIs are a better predictor of TN and TP than the Stream Corridor LDI.  However, the linear regression models developed using the data presented in Figures 7 and 8, while statistically significant, provide little meaningful benefit (r[2] < 10%) to using one of the alternative LDI approaches in comparison to the Stream Corridor LDI".  The analysis also showed that the entire watershed LDI approach was less effective at predicting nutrients than either the corridor or near-field LDIs.  The ultimate conclusion from this detailed analysis was that Tetra Tech recommended removal of WBIDs 1653 and 3401, since these particular sites had elevated TP relative to the measures of LDI.  DEP removed these WBIDs from the reference site distribution, and re-analyzed the data.  Note that the Tetra Tech analysis did not conclude that any other reference sites should be removed.  

As noted previously, Earthjustice seems to presume that any minor disturbance in a WBID, as evidenced by LDIs above natural forest conditions, should result in its rejection as a reference site.  Further, without additional information about their site selection criteria, we are concerned that Earthjustice's comparison of natural forest sites and silviculture sites may be biased.   It appears that Earthjustice concluded that sites with TP concentrations above a pre-determined TP value (which appears to be 70 ug/L of TP) should be removed, while sites with TP concentrations less than 70 ug/L should be retained, even if they have the same (minimal) level of human disturbance.  If this is the case, the statistical analysis provided by Earthjustice is not valid, irrespective of the stated p values. 
To objectively test the claim that forestry increases nutrients in adjacent streams, DEP assembled data from the Peninsula and Panhandle reference sites and performed regression analyses on the percent silviculture in each basin against nutrient concentration (see next four graphs below).  Percent silviculture was calculated based on the most recent (2004) landuse coverage.
                                       
Geometric mean TP vs. % Silviculture in the Peninsula.  The near field drainage includes the drainage area within 10 km upstream of the sample station.  Note that there is no increase in TP with % Silviculture in Peninsular reference sites.
                                       
Geometric mean TN vs. % Silviculture in the Peninsula.  The near field drainage includes the drainage area within 10 km upstream of the sample station.  Note that there appears to be a non-significant decrease in TN with % Silviculture in Peninsular reference sites.
                                       
                                       
Geometric mean TP vs. % Silviculture in the Panhandle.  The near field drainage includes the drainage area within 10 km upstream of the sample station.  Note that there appears to be a non-significant decrease in TP with % Silviculture in Panhandle reference sites.
                                       
Geometric mean TN vs. % Silviculture in the Panhandle.  The near field drainage includes the drainage area within 10 km upstream of the sample station.  Note that there appears to be a non-significant decrease in TN with % Silviculture in Panhandle reference sites.
The results clearly indicate that the percent silviculture in each basin did not increase either the TP or TN, and that a non-significant decreasing trend between the percent silviculture and nutrients were evident in three cases.   Additionally, DEP conducted a Mann-Whitney comparison of TP at high and low silviculture sites in the Peninsula (below).
                                        
Mann-Whitney U Test comparing TP at low silviculture reference sites (<10 % silviculture) with TP at high silviculture sites (>40% silviculture).  Results show there is an 89% probability that there is no difference between the groups.  
The low and high silviculture thresholds correspond well with the level of forestry activities present in the Earthjustice impacted and unimpacted sites, respectively.  Results from the DEP Mann-Whitney test, using objective criteria (% silviculture in each reference site basin), demonstrated that Earthjustice's analysis (grouping selected sites based on non-objective visual interpretation of maps) was fundamentally flawed, and that there is no increase in TP in Peninsula reference sites associated with silviculture.
"Point Sources" and Reference Sites
Earthjustice incorrectly claims that DEP did not consider that channelized systems could discharge nutrients from land uses beyond a riparian zone.  DEP staff actually visited watersheds with nutrients above the midpoint of the reference site nutrient distribution in an effort to detect adverse human influences, and eliminate sites if needed.  If DEP observed channelized conveyances that would likely transport potentially poor water quality to the candidate reference site, DEP eliminated the site.  The sites eventually confirmed as nutrient reference did not have ditches from adverse land uses penetrating the riparian buffer zone.
Forestry and Reference Condition
Earthjustice also incorrectly assumes that forestry is a significant source of nutrients and disturbance.  The following information quantitatively disproves this assumption.  As an initial point, a peer reviewed "Before-After-Control-Impact" Forestry Best management Practices Study conducted by DEP shows that forestry is a benign land use (Vowell, J. L. and R. B. Frydenborg. 2004.  A biological assessment of best management practice effectiveness during intensive silviculture and forest chemical application. Water, Air, & Soil Pollution: Focus 4:297-307). The following is an excerpt from the DEP report of the study, which sampled streams upstream and downstream of five representative forestry sites, both before and after the forests were harvested following established BMPs.  Note that this study, which was based on an earlier version of the SCI, used replicate sampling and Analysis of Variance to determine potential effects of silviculture, including herbicide and fertilizer applications.  The results clearly refute Earthjustice's presumptions concerning forestry.
      "In the Santa Fe River, there were statistically significant improvements in the SCI score during the second sampling event, compared with the initial sampling event (p = 0.01).  Additionally, the test site score was significantly higher than the reference site score during both sampling events (p = 0.005).  Between sampling events, 200 pounds/acre of di-ammonium phosphate was applied on 5 year old pines in the vicinity of the test site.
       The water level was approximately 0.5 m deeper during the August sampling than it was in November, making sampling in this marginally wadeable river more difficult.  This could account for the improvements in the SCI score during the period that the water level was more conducive to collection activities.  The higher SCI values at the test site may be explained by the slight improvements in substrate diversity there, as well as the slight reductions in habitat smothering.  
                                          
      R = Reference site above silviculture operations, T= test site potentially affected by silviculture.
      
      There were no significant time by treatment effects, indicating that the fertilizer addition was not linked to the observed changes.  Invertebrates are affected indirectly by nutrient enrichment, responding to changes in habitat conditions (e.g., caused by algal smothering) or food sources (e.g., proliferation of less nutritious or noxious algae).
                                          
      Periphyton communities respond more quickly to nutrient enrichment effects, and adverse changes in the periphyton community can indicate excessive nutrient enrichment.  The periphyton community in the Santa Fe River was dominated by diatoms, both before and after the nutrient application.  Many of the diatom taxa found at both sites were indicative of low nutrient conditions, such as Eunotia pectinalis, Frustulia rhomboids, Pinnularia major, Achnanthes lanceolata, which is consistent with the water chemistry results.  
      
      Conclusions associated with the Study follow. Habitat quality was in the "optimal" category at all sites, both before and after silviculture activities.  The BMPs appeared successful in controlling erosion, since no additional habitat smothering or stream bank instability was noted after silviculture operations took place.  The decreases in riparian buffer zone width that were noted at three of the four sites were allowed by the BMPs, and did not shift the overall habitat category at any site.There appeared to be no major changes in physical-chemical measurements due to silviculture activities.  Turbidity was low at all sites.  During an intense rain in February, 1997, direct observation at Jack Branch indicated that no turbid water from the clear-cut site was reaching the stream.  Nutrients were relatively low at all sites, both before and after the forestry operations."

Forestry BMP Compliance Survey
The Florida Division of Forestry conducts random assessments of forestry operations throughout the state to determine the relative compliance with Forestry Best Management Practices:  http://www.fldof.com/forest_management/hydrology_index.html
The most recent 2007 compliance survey report states, "For the 2007 Survey, no sites scored below 83% in overall BMP implementation.  Eighty-two percent of the sites scored 100% implementation for all BMPs. The range of compliance scores was 83% to 100%, and the statewide average for overall BMP compliance was 98.6%, a decrease from 99.1% in 2005.  The statewide average compliance for the period of record is 94% and a total of 6,195 individual forestry operations have been surveyed since 1981.  Information from Jeff Vowell (Florida Division of Forestry), co-author of the compliance survey, shows that on average, only 2% of silviculture sites actually receive fertilizer applications (see graphs below).  Since this is a probabilistic survey the results may be extrapolated statewide.
                                       
A Probabilistic Forestry Compliance Survey conducted by the Florida Division of Forestry depicting the percent of silviculture sites actually receiving fertilization treatments since 1995.
                                       
A Probabilistic Forestry Compliance Survey conducted by the Florida Division of Forestry depicting the number of silviculture sites with and without fertilization treatments since 1995.

Dissolved Oxygen
Earthjustice notes that some of the reference sites have DO levels below the criteria and claims that low DO is caused by nutrient inputs.  For Classes I and III fresh waters, Florida's dissolved oxygen criterion currently requires that the dissolved oxygen "Shall not be less than 5.0. Normal daily and seasonal fluctuations above this level shall be maintained." (Rule 62-302.530[31]).  Florida's fresh waters are exposed to temperate to subtropical climates and many originate in low oxygen environments, such as swamps and groundwater aquifers. These sources are naturally low in dissolved oxygen, and have natural daily and seasonal fluctuations in dissolved oxygen that fall below 5.0 mg/L.  Since these levels are the result of natural conditions and the native flora and fauna have adapted to this natural variation, they generally do not impact the designated uses.  
Rule 62-302.300(15), F.A.C., states that "the Department shall not strive to abate natural conditions."  Therefore, DEP has conducted a Dissolved Oxygen Study to characterize the causes of naturally low dissolved oxygen levels to more clearly define "natural conditions."  The DO Study consisted of:
         o  Quarterly sampling at 342 sites (both reference and
    non-reference) across Florida
               # 150 lake sites 
               # 160 stream sites
               # 32 canal sites
         o Monitoring consisted of:
               #  4-day YSI deployments
               #  Vertical profile measurements (deploy & retrieval)

Counter to the premise that sites with higher human disturbance (and nutrients) are associated with lower DO, the results demonstrated that reference streams and lakes (LDI < 2) actually exhibited higher frequencies of < 5.0 mg/L DO conditions than sites with higher human disturbance and higher nutrient levels (see Table below). 

         
   
   It is the experience of DEP scientists that naturally low DO occurs when:
      * Temperatures are elevated;
      * Water velocity is low and stagnant conditions prevail;
      * Waterbodies receive substantial leaf litter from forests and swamps, and natural decomposition processes reduces DO; and
      * Morphological conditions, such as depth, promote depositional processes which in turn lead to low DO.
EPA has acknowledged that Florida has naturally low DO, and has approved many Site Specific Alternative Criteria for DO in Florida waters.  The data show that Earthjustice's contention that low DO results from adverse nutrient enrichment is not supported.
Rebuttals to Specific Reference Site Criticisms
Stevens Branch-  FDEP's technical support document shows that this site is minimally disturbed.  The site had an exceptional SCI score with no algal or plant issues and a very high habitat assessment score.  Direct site observations by field biologists indicated no direct inputs from silviculture to the stream.  Silviculture in Florida, with the high compliance with Forestry BMPs (98%), is a benign land use as demonstrated above.  Outside of the 100 meter buffer, silviculture exists in the watershed, but does not appear to be impacting the stream health or the nutrient levels.  The nitrate levels are 0.01 mg/L in Stevens Branch.  If the nutrient levels were affected by fertilization of forests, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  Nitrate levels below the concentrations in Seven's Branch indicate little to no fertilizer influence on Steven's Branch.  According to Earthjustice, this site was included on the 1994 303(d) list; graded as fair or poor, but it was not assessed due to lack of data.  DEP determined that during the verified period (1999-2006), the area was not impaired for chlorophyll-a.  All other parameters were placed on the 3b list due to insufficient data.  DO had one exceedance in eleven samples (most likely a natural condition fluctuation.  Iron levels were sometimes elevated, also probably associated with the swampy conditions (an iron SSAC is likely).
Cow Creek  -  Direct field observations indicated that human activities (silviculture) were beyond an extensive buffer zone and do not appear to be impacting the stream health or the nutrient levels.  Habitat was optimal, and no algal or plant issues were evident. The nitrate levels are 0.04 mg/L in Cow Creek.  If the nutrient levels were affected by fertilization of forests, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in Cow Creek are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.  Earthjustice claims that DO was an issue here, but FDEP's 2002 TMDL update listed DO as 3c (no causative pollutant).   The site meets standards for numerous other parameters.  During the 2000-2007 verified period, six DO excursions occurred out of 33 samples (therefore, it was not impaired for DO). The site was also not impaired for chlorophyll a.  
Orange reek-   This is one of the few sites in the Peninsula where stoneflies are found.  The SCI (74) was in the exceptional category, clearly supporting the designated use.  Orange Lake is a DEP reference lake, with healthy LVI and chlorophyll near 20 ug/L (paleolimnology at nearby Lake Wauberg indicate that up to 40 ug/K chlorophyll may be a natural background condition in this area).  The majority of the watershed is forested, and direct observation indicates extensive riparian buffer zones and that water from potentially adverse activities does not reach the stream through surface water pathways. Habitat was optimal, and no algal or plant issues were evident. The nitrate levels are 0.09 mg/L in Orange Creek.  If the nutrient levels were affected by human land uses, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in Orange Creek are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.  Currently, this waterbody is on the G1C2 Verified List for DO with TN listed as the causative pollutant, and DEP believes this to be an error. The median BOD is 2.15 mg/L and the stream is characterized by exceptional biological health.  Orange Creek is not impaired for nutrients based on chlorophyll a. The mean of 44 chlorophyll a (corrected) values was 1.24 ug/L.  The absence of algal or plant biomass (not observed in the stream), coupled with low chlorophyll and low BOD, indicates no causative link between the TN (natural in these blackwater streams) and DO.
Withlacoochee R @ 471-  The very undisturbed Green Swamp  forms the headwaters of this system, and streams in the area are naturally low in DO due to sluggish flows and leaf litter decomposition.  There were no ditches and no "ranchettes" observed during field visits.  Habitat was optimal, and no algal or plant issues were evident. The nitrate levels are 0.02 mg/L in Withlacoochee R @ 471.  If the nutrient levels were affected by human land uses, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in Withlacoochee R @ 471are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.  According to report cited by Earthjustice, Mattress Drain, upstream of DEP's reference site, is in category 3c. There is very little data for Mattress Drain (WBID 1435, a very swampy area) and there are no impairments.  Also according to Earthjustice, Grass Creek, upstream of the reference site, is in category 3b.  However, DEP has only 1 data point for several parameters.  According to the report cited by Earthjustice, Pony Creek, upstream of the reference site, is in category 4c (impairment not caused by a pollutant).  Pony Creek was not impaired for nutrients based on C1 assessment.
Econlockhatchee River- This Site had healthy SCI scores and a very low LDI (1.1).  Minimal periphyton growth was observed and fish were abundant.  Earthjustice asserts that the TP levels at the Snowhill Road site were influenced by inputs from the Little Econlockhatchee River.  To test this, DEP evaluated TP data from the Big Econlockhatchee at Pickett Lake Road, which is located approximately five miles upstream of the Little Econlockhatchee, and compared it with TP data from the Snowhill Road location, located approximately six miles downstream from the Little Econlockhatchee.  A two-sided Students t-test was conducted, and the results indicated no statistically significant difference in the TP concentrations between the two sites (although there were only three observations at the Pickett lake Rd. site).  The nitrate levels observed at the Snowhill Rd. site (0.15 mg/L) are well below than the proposed protective 0.35 mg/L nitrate criterion.  Although EPA has listed this site as impaired for nutrients, DO, BOD, and coliforms in 2002, there is no info provided on potential sources by EPA.  DEP has delisted the waterbody, placing DO in category 4c and Category 2 for various other parameters (22, including most metals).  The Econlockhatchee is only impaired for mercury, based on fish consumption advisories.  DEP's mercury TMDL study has shown that oligotrophic, high DOC waters are most susceptible to biological magnification of atmospherically deposited mercury.
Bee Branch-  Direct observation showed that the majority of the area consisted of wet or dry prairie.  Following the DEP standard practices for site recons, an area of citrus groves upstream of the site was scrutinized, and staff walked along the riparian buffer and confirmed that no direct runoff from the citrus was reaching the site.   Further, the site was biologically healthy as evidenced by the SCI, and direct observation indicated there were no issues with algae or adverse aquatic plants.  The nitrate levels are 0.05 mg/L in Bee Branch.  If the nutrient levels were affected by human land uses (in this case, citrus), then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in Bee Branch are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.
Cypress Branch @78-  Direct observations at the site indicated that extensive riparian buffers were present, commonly > 1000 m.  Most of the observed land use was native prairie and pine forest.  The SCI indicated healthy biology, habitat was optimal, and no algal or aquatic plant issues were observed.  The nitrate levels are 0.03 mg/L in Cypress Br @78.  If the nutrient levels were affected by human land uses, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in Cypress Br @78 are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.
St. Johns River @ Deland-  DEP staff confirmed in the field that the point source discharge was downstream of the reference site, and only data upstream from this site was used.  DEP staff concluded from site observations that this location was influenced only by very low levels of anthropogenic stressors and no adverse algal or aquatic plant issues were present.  The nitrate levels are 0.07 mg/L in St. Johns @ Deland.  If the nutrient levels were affected by human land uses, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in St. Johns @ Deland are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.
Moses Creek-  Field observations indicated no direct runoff from forestry operations.  This site had healthy SCIs, optimal habitat, and no algal or aquatic plant issues.  Earthjustice admits that the narrowest riparian buffer was a protective 88 m.  The new subdivision cited by Earthjustice had a state of the art stormwater retention facility, which would result in mitigation of nutrient inputs.  The nitrate levels are 0.01 mg/L in Moses Ck.  If the nutrient levels were affected by human land uses, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in Moses are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.
Steinhatchee River-  Earthjustice supports this site as a reference, even though forestry operations were present in this watershed at a similar magnitude to others that Earthjustice rejected.  DEP site visits indicated extensive riparian buffer zones, healthy biology, and no adverse algal or aquatic plant effects.  The nitrate levels ranged from 0.01 to 0.03 mg/L in the Steinhatchee R.  If the nutrient levels were affected by human land uses, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in Steinhatchee R. are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.
Little Orange Creek- This site had an SCI score (89), in the exceptional category, and no algal or aquatic plant issues.  This is one of the few Peninsular sites where sensitive stoneflies are present.  DEP sampled Redwater Lake and found the Lake Vegetation Index (72) there to be clearly healthy, meaning the lake also supports biological integrity.  Field observations indicated that there were no direct inputs from potentially adverse land uses.  The nitrate levels are 0.02 mg/L in Little Orange Ck.  If the nutrient levels were affected by human land uses, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in Little Orange Ck. are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.
Waccasassa River- This site had an exceptional SCI (69), showing full support of designated use.  EPA scientists visited the entire system with DEP staff, from upstream sites to the estuary, and all confirmed that it is a minimally disturbed system.  Field observations indicated that there were no direct inputs from potentially adverse land uses. See discussion of forestry BMP practices above. The nitrate levels are 0.02 mg/L in Waccasassa R.  If the nutrient levels were affected by human land uses, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in Waccasassa R. are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.
Withlacoochee River @ Stokes Ferry and Trails End-  Earthjustice supports these sites even though they share many land use characteristics  with sites Earthjustice wishes to reject.  Field observations indicated that there were no direct inputs from potentially adverse land uses. Withlacoochee @ Stokes Ferry had an SCI score (69), in the exceptional category, and both sites had no algal or aquatic plant issues.  The nitrate levels range from 0.02 mg/L to 0.1 in these Withlacoochee R sites.  If the nutrient levels were affected by human land uses, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in Withlacoochee R. are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.
Blackwater Creek and Tosohatchee Creek-  Earthjustice supports these sites.  Blackwater Creek has an exceptional (77) SCI score but due to swamp-like conditions, an SCI could not be performed at the Tosohatchee. Both sites had no algal or aquatic plant issues. The nitrate levels range from 0.01 mg/L in Tosohatchee to 0.13 in the Blackwater Ck.  If the nutrient levels were affected by human land uses, then the most notable nutrient, because of its solubility in water and ability to transport through groundwater, is nitrate.  The nitrate levels observed in Tosohatchee and Blackwater Creek are significantly lower than the proposed protective 0.35 mg/L nitrate criterion.

Conclusions
In summary, DEP has followed a well defined, objective approach for establishing minimally disturbed reference sites.  EPA's peer reviewers supported DEP's approach as reasonable and scientifically defensible.  The information presented in the DEP Nutrient TSD and in this document provides credible evidence that the nutrient concentrations of the thoroughly vetted DEP reference site population are characteristic of the levels needed to sustain aquatic life use support.

References 
 
Doherty, S.J. 1995. Emergy evaluations of and limits to forest production.  PhD Dissertation, Department of Environmental Engineering Sciences, University of Florida, Gainesville.
 Pritchett, WL. 1983.  Nutrition and fertilization of slash pine.  In E.L. Stone (ed) The managed slash pine ecosystem; Proc. of Symposium. School of Forest Resources and Conservation, University of Florida.  June 9-11.  34 PP.

