A Comparison of Nitrate (NO3) and Silica (SiO2) concentrations in Canajoharie Creek, NY and Conodoguinet Creek, PA A Practical Examination Shippensburg University Department of Geography and Earth Science Rachel Hager 3/11/2016 1
Table of Contents Page 1.0 Introduction..3 1.1 Research Questions 3 2.0 Literature Review.4 2.1 Nitrogen..5 2.2 Dissolved Silica (SiO2)...6 3.0 Specific Study Area.7 4.0 Methods...10 5.0 Results and Discussion 13 6.0 Conclusion...19 7.0 References...22 List of Figures Figure 1. The Conodoguinet Creek watershed in relation to its location within Pennsylvania.... 8 Figure 2. Land use within the Conodoguinet Creek watershed, PA.... 10 Figure 3. The seasonal/temporal relationship between discharge, silica (SiO 2), and nitrate (NO 3) concentrations found in the Conodoguinet Creek... 15 Figure 4. The statistical relationship between nitrate (NO 3) and silica (SiO 2) concentrations for Conodoguinet Creek.... 17 Figure 5. The statistical relationship between nitrate (NO 3) and silica (SiO 2) concentrations without the silica (SiO 2) concentration outlier data point.... 17 Figure 6. Discharge compared to the silica (SiO 2) concentration outlier data point.... 18 Figure 7. The statistical relationship between nitrate (NO 3) and silica (SiO 2) concentrations during drought conditions in the Conodoguinet Creek watershed.... 19 2
1.0 Introduction Across the United States the issue of water quality and agriculture s influence on water quality as a non-point source of pollution is a continuing concern ((Wall et al. 1998). In response to water quality concerns, the National Water Quality Monitoring Council (NWQMC) was formed to bring together experts to develop mutual methods of assessing water quality across the United States (US Geological Survey 2005). The NWQMC collects water quality data, for example discharge, temperature, ph, conductivity, and nutrient concentrations in surface waters. Nutrient concentration data often consists of nitrate (NO3), silica (SiO2), and phosphate concentration measurements that can be used in numerous water quality evaluations. The relationships between certain nutrients and even the discharge of streams can provide important information about the health of streams. The seasonal and temporal relationships between silica (SiO2), nitrate (NO3), and discharge specifically have been documented often and demonstrate correlations between the effect of the season and time of low vs. high discharge rates and the resulting nitrate (NO3) and silica (SiO2) concentrations. 1.1 Research Questions The purpose of this practical exam was to compare the relationships between discharge, nitrate (NO3) concentration, and silica (SiO2) concentration in Conodoguinet Creek, PA and Canajoharie Creek, NY. The principal goal was to create a number of hydrologic graphs to compare the Conodoguinet Creek waters to the results found in (Wall et al. 1998). The specific research questions for this study are as follows: 3
What is the pattern of discharge from the Conodoguinet Creek from August 1996 through June 2002? What is the seasonal/temporal relationship between discharge, nitrate (NO3) concentration, and silica (SiO2) concentration in the Conodoguinet Creek, PA and how does it compare to the Canajoharie Creek, NY? What is the relationship between the nitrate (NO3) concentration and silica (SiO2) concentration in the Conodoguinet Creek, PA and how does it compare to the Canajoharie Creek, NY? How might an outlier point affect the relationship between the nitrate (NO3) concentration and the silica (SiO2) concentration? How do periods of drought affect the relationship between the nitrate (NO3) concentration and silica (SiO2) concentration? 2.0 Literature Review A study was conducted from March 1993 to January 1996 that measured the seasonal and spatial patterns of nitrate (NO3) and silica (SiO2) concentrations in the Canajoharie Creek, NY (Wall et al. 1998). This study compared nitrate (NO3) and silica (SiO2) concentrations to the discharge in the Canajoharie Creek to identify seasonal and temporal variations in the concentration levels (Wall et al. 1998). Wall et al. (1998) also compared the nitrate (NO3) concentrations and silica (SiO2) concentrations to evaluate the statistical relationship between the two nutrient concentrations. This study found that nitrate (NO3) and silica (SiO2) concentration variations are dominated by biological processes (Wall et al. 1998). To better understand the nutrients discussed in the Wall et al. (1998) study and also in this practical exam nitrogen and silica will be discussed below. 4
2.1 Nitrogen Nitrogen comes in many forms including dinitrogen (atmospheric nitrogen), organic nitrogen, and inorganic nitrogen. Dinitrogen can be found abundantly in the atmosphere but it can also be found in the soil (Killpack and Buchholz 1993). Organic nitrogen is nitrogen that is created as a byproduct of biological processes in animal forms of life (Killpack and Buchholz 1993). Nitrite (NO2) is a form of organic nitrogen (Heathwaite and Johnes 1996). Inorganic forms of nitrogen include ammonia (NH3) and nitrate (NO3)which can be found both in soil and aquatic environments. Ammonia (NH3), in terms of water quality, is not as great as concern as nitrate (NO3) because it is not as soluble in water and is therefore not easily removed from soil (Killpack and Buchholz 1993). If ammonia (NH3) remains in the soil long enough for bacteria to carry out the nitrification process it is converted in the nitrate (NO3) form of nitrogen and is then susceptible to uptake by plant life, denitrification, or more concerning leaching into ground or surface water sources (Killpack and Buchholz 1993). According to Killpack and Buchholz 1993, some of the most common sources of ammonia (NH3) are decaying organic matter, manure, or fertilizers. However if ammonia (NH3) is carried into water sources, according to Clarke and Baldwin (2002), ammonia (NH3) has been found to be toxic to a variety of plants, including some hydrophilic species of plants that could be found along stream corridors. Ammonia (NH3) at elevated levels of concentration during short period of time has little to no detrimental effect of plants and plant growth while exposure to elevated ammonia (NH3) concentrations for extended periods of time can be detrimental to the growth rate of plants (Clarke and Baldwin 2002). In terms of water quality the nitrate (NO3) form of nitrogen is the form that can be an issue (Killpack and Buchholz 1993). Nitrate (NO3) is very soluble in water and therefore moves 5
easily through aquatic systems and can cause significant water quality degradation (Jaynes et al. 1999). Nitrate (NO3) is also easily added to streams through groundwater discharge and overland runoff from agricultural areas (Jaynes et al. 1999). Agriculture is a major contributor to the level of nitrate (NO3) concentrations found within streams due to over-applications of organic and inorganic fertilizers (Jaynes et al. 1999). Often in water quality studies only nitrate (NO3) or total oxidisable nitrogen (nitrate (NO3) + nitrite (NO2)) are studied due to the relative ease of measurement of these concentrations (Heathwaite and Johnes 1996). It is also frequently assumed that only inorganic forms of nitrogen are available to organisms in water (Heathwaite and Johnes 1996). The measurement of nitrogen in Wall et al. (1998) and even for agencies like the NWQMC reflect the general practice of measuring only nitrate (NO3) or total oxidisable nitrogen (nitrate (NO3) + nitrite (NO2)). 2.2 Dissolved Silica (SiO2) Most dissolved silica (SiO2) is found in streams as a result of the dissolution of silica (SiO2)-containing rock (Iler 1979). There are two forms of dissolved silica (SiO2), a reactive form of dissolved silica (SiO2) and a colloidal form of silica (SiO2) (Iler 1979). The reactive forms of dissolved silica (SiO2) are the most abundant forms of silica (SiO2) in both groundwater and surface water but colloidal forms of dissolved silica (SiO2) are found more often in surface water (Iler 1979). In surface waters reactive forms of silica (SiO2) are assimilated by diatoms (a form of algae) to create a shell made of silica (SiO2) crystals and when the diatoms decompose the silica (SiO2) is released back into the aquatic environment as the reactive form of dissolved silica (SiO2) (Iler 1979). Hornberger et al. (2001) found that dissolved silica (SiO2) follows the pattern of stream discharge with a slight time lag. Many studies have found that silica (SiO2) concentrations 6
decrease in summer months and increase throughout the winter ad through the spring months (Wall et al. 1998). Wang and Evans (1969) hypothesized that the decrease of silica (SiO2) concentrations during the summer was due the increased rate of uptake by diatoms. 3.0 Study Area The study area includes the Conodoguinet Creek watershed in the Cumberland Valley of South Central Pennsylvania (Figure 1). The watershed is located within Franklin, Cumberland, and Perry counties and has a drainage area of 470 square miles. The Conodoguinet Creek drains the Cumberland Valley from west to east to the Susquehanna River. The average daily stream discharge using data from the past 83 years as reported by the U.S. Geological Survey (2016) for the Conodoguinet Creek is 21.12 cubic meters per second (cms) at the gaging station near Hogestown, PA (USGS Station 01570000). 7
Figure 1. The Conodoguinet Creek watershed in relation to its location within Pennsylvania. The geology throughout the area is comprised of Cambrian age metamorphic rocks on South Mountain, Cambrian and Ordovician carbonate rocks, and Ordovician siliciclastic rocks in the southern and northern parts of the valley respectively (Zairchansky 1986). Geologic units like resistant quartzite, phyllite, conglomerate, limestone, and dolomite can be found in the study area as well. There is an abrupt change from non-reactive metamorphic rocks to highly soluble carbonate rocks which is important when it comes to understanding the groundwater flow in the area because the chemicals and other soluble compounds in the rocks will be dissolved in the groundwater and change the water chemistry which can affect how that water can be used (Zairchansky 1986). It should be noted that the geology of the area also indicates that the 8
Conodoguinet Creek watershed includes an area of karstic geology which may affect the way nutrients and water flows through the watershed. In the Cumberland Valley, the slopes of Blue and South Mountains are steep and therefore remain forested while the valley floor consists of gently rolling hills that are fertile, and is excellent for farming (Vogelmann et al. 1998a; Vogelmann et al. 1998b). The types of soils found throughout the study area include Berks shaly silt loam, Edom silty clay loam, Hagerstown silt loam, Hagerstown rock outcrop, and Huntingdon silt loam (Zairchansky 1986). Figure 2 provides a visual presentation of the land use throughout the Conodoguinet Creek watershed. The distribution of land use throughout the Conodoguinet Creek watershed is primarily agricultural land use (51.49%) followed by forested land use (37.68%). The third most common land use is urban land use (8.93%) and the least common land use is other (1.91%). 9
Figure 2. Land use within the Conodoguinet Creek watershed, PA. 4.0 Methods The data that were used to create the map of the Conodoguinet Creek watershed and the land use map within the Conodoguinet Creek were collected from various sources. The datasets that were used to create the maps include the Chesapeake Bay Land Cover Data (CBLCD) series (Irani and Claggett 2010), the Pennsylvania State Boundary (Pennsylvania Department of Transportation 2016), the Pennsylvania County boundaries (Myers and Bishop 1999), NHD Flowline Susquehanna (US Geological Survey 2007-2014), and Hydrologic Unit Maps (Seaber et al. 2007). The data were then imported into the ArcMap system where it could be 10
manipulated to describe the location of the Conodoguinet Creek watershed in relation to the state of Pennsylvania and the land use within the watershed. The CBLCD layer was reclassified into 4 classes: 0 (other), 2 (urban), 4 (forest), and 8 (agriculture). After reclassifying the CBLCD layer the area that each category of land use occupies in the Conodoguinet Creek watershed was calculated by multiplying the count of each raster cell in each category by the size of each cell (30 m x 30m = 900 m 2 ). The percentage of each land use category in the watershed was calculated by dividing the total square meters in each category by the total square meters in the entire watershed and then multiplying the result by 100. To analyze the pattern of discharge from the Conodoguinet Creek from August 1996 through June 2002, daily discharge statistics were downloaded from the US Geological Survey s website (US Geological Survey 2005). The daily discharge statistics were collected from the Hogestown, PA station for the Conodoguinet Creek and has a hydrologic unit code of 01570000. The data were downloaded in cubic feet per second (cfs), then converted to cubic meters per second (cms) by using Equation 1. After the data were converted to cubic meters per second a hydrograph was created to create a visual presentation of the daily discharge of the Conodoguinet Creek. (1) The seasonal/temporal relationship between discharge, nitrate (NO3) concentration, and silica (SiO2) concentration was analyzed by comparing the Conodoguinet Creek daily discharge data to the National Water Quality Assessment (NAWQA) (US Geological Survey 2005) data. Before analysis could be conducted, however the data between the two data sets were converted into a useable format. The NAWQA data consists of nitrate (NO3) and silica (SiO2) concentrations that was collected once or twice a month and some months not at all. To compare the two data sets 11
the NAWQA data was matched by the date that nitrate (NO3) and/or silica (SiO2) concentration was collected to the appropriate daily discharge data point. After the data sets were converted into a useable format a graph was created to create a visual presentation of the seasonal/temporal relationship between discharge and nitrate (NO3) and silica (SiO2) concentrations. This graph was created by following the methods exhibited in Wall et al. (1998). The nitrate (NO3) and silica (SiO2) concentrations were exhibited on the primary vertical (y) axis and the discharge was exhibited on the secondary vertical (y) axis. It should be noted that the discharge was displayed in units of LOG (cubic meters per second). The next step involved studying the relationship between the nitrate (NO3) concentrations and the silica (SiO2) concentrations. This relationship was visually represented by creating a graph. The nitrate (NO3) concentrations were exhibited on the horizontal (x) axis and the silica (SiO2) concentrations were exhibited on the vertical (y) axis. The r 2 value was also calculated by adding a linear trendline to the graph. The methodology to create this graph was also modeled after the Wall et al. (1998) paper. The statistical relationship between the nitrate (NO3) concentration and silica (SiO2) concentration was explored by removing one outlier point (4.08, 33) from the nitrate (NO3) vs. silica (SiO2) concentration graph and observing the change in the r 2 value. The justification for the removal of the outlier data point can be found in Wall et al. (1998). Wall et al. (1998) state that there was a silica (SiO2) concentration data point that was anomalously high which means that the data point deviated from what was normal or expected. However it can also be argued that all data samples that are collected are real and therefore important to the analysis of the data. If the outlier data point cannot be proven to be a typo or a lab error then it must be considered to be real and therefore should remain in the data. 12
In Wall et al. (1998) the relationship between the nitrate (NO3) and silica (SiO2) concentration was found to be stronger during drought conditions. To explore this relationship for the Conodoguinet Creek the data was filtered to display nitrate (NO3) and silica (SiO2) concentrations that were collected during drought conditions. Drought conditions in the Conodoguinet Creek were considered to be instances where the daily discharge was less than 3 cubic meters per second throughout the summer months (June through September). Five instances were found where nitrate (NO3) and silica (SiO2) were collected during drought conditions. These instances were graphed using the same process as the general relationship between nitrate (NO3) and silica (SiO2). After the data analysis was completed the results were compared to the results found in the paper by Wall et al. (1998). The seasonal/temporal relationship between discharge and nitrate (NO3) and silica (SiO2) concentrations for the Conodoguinet Creek was compared to the same relationship demonstrated in Wall et al. (1998) s Figure 3a and the pattern of discharge for the Conodoguinet Creek was compared to the discharge for the Canajoharie Creek using the same figure. The relationship between the nitrate (NO3) and silica (SiO2) concentrations for the Conodoguinet Creek was compared to the same relationship for the Canajoharie Creek. 5.0 Results and Discussion A discharge hydrograph for the Conodoguinet Creek is included as Figure 3. The daily discharge statistics has a great deal of variation. In general, however, there is a pattern of discharge decreasing through the summer months. Then the discharge reaches a low point during the winter months and then begins to increase and reaches a peak during the spring months. In the paper by Wall et al. (1998) the same general pattern can be observed. 13
The seasonal/temporal relationship between the Conodoguinet Creek discharge and the nitrate (NO3) and silica (SiO2) concentrations can be examined in Figure 3. This figure has demonstrated a slightly inverse relationship between daily discharge statistics and nitrate (NO3) concentration while the daily discharge statistics and the silica (SiO2) concentration follow the same pattern of peaks during the spring months and then decreases during the summer months to low points in the winter. These results were then compared to the results found in Wall et al. (1998). Wall et al. (1998) found that as discharge decreased so too did both nitrate (NO3) and silica (SiO2) concentrations and as discharge increased both nitrate (NO3) and silica (SiO2) concentrations increased. 14
Figure 3. The seasonal/temporal relationship between discharge, silica (SiO2), and nitrate (NO3) concentrations found in the Conodoguinet Creek. 15
The statistical relationship between the concentrations of nitrate (NO3) and silica (SiO2) was then examined in Figures 4 and 5. Figure 4 demonstrated the statistical relationship between nitrate (NO3) and silica (SiO2) at Conodoguinet Creek and includes all data points, including an outlier which consisted of a nitrate (NO3) concentration of 4.08 mg/l and a silica (SiO2) concentration of 33 mg/l). The r 2 value associated with this statistical relationship was 0.0028. This r 2 value is considered to be low and to signify a weak relationship between nitrate (NO3) and silica (SiO2) concentrations. The negative relationship between nitrate (NO3) and silica (SiO2) concentration means that as nitrate (NO3) concentrations increase the silica (SiO2) concentrations are more likely to decrease. Figure 5 has shown this same relationship but without the silica (SiO2) concentration outlier point of 33 mg/l. Removing the outlier increased the r 2 value from 0.0028 to 0.1869, strengthening the inverse relationship between nitrate (NO3) and silica (SiO2) concentrations. These results differ greatly from the results found in the Wall et al. (1998) paper. Wall et al. (1998) found a strong positive correlation between nitrate (NO3) and silica (SiO2) concentrations with an r 2 value of 0.93 (without their outlier data point). 16
Figure 4. The statistical relationship between nitrate (NO3) and silica (SiO2) concentrations for Conodoguinet Creek. Figure 5. The statistical relationship between nitrate (NO3) and silica (SiO2) concentrations without the silica (SiO2) concentration outlier data point. 17
Figure 6 depicts the relationship between Conodoguinet Creek discharge and the silica (SiO2) concentration outlier data point. Samples collected before and after the anomalous silica (SiO2) concentration data point was included to give a visual representation how different the outlier data point was from the remaining silica (SiO2) concentration samples with respect to discharge. In this figure all three silica (SiO2) concentration samples occur during discharge flow periods of 9-10 cms and both samples occur during a period of increasing discharge. The rate of discharge was similar during each sample period when the silica (SiO2) concentrations were collected which points to the probability of the outlier data point being a typo or lab error. Due to the probability of the outlier occurring because of lab error or typo error, it will not be included in further analysis of results. Figure 6. Discharge compared to the silica (SiO2) concentration outlier data point. 18
Figure 7 exhibits the relationship between nitrate (NO3) and silica (SiO2) concentrations during drought conditions. According to Figure 7, visually nitrate (NO3) and silica (SiO2) have almost no correlation and even the r 2 value is almost 0 (r 2 = 0.007) which means that there is no correlation (0 = no correlation; 1 = strong correlation). Figure 7. The statistical relationship between nitrate (NO3) and silica (SiO2) concentrations during drought conditions in the Conodoguinet Creek watershed. 6.0 Conclusions Many of the plots created for the Conodoguinet Creek data were similar to the results found in Wall et al. (1998) but there were fundamental differences as well. The discharge graphs demonstrated the same general pattern for both the Conodoguinet and Canajoharie Creeks. The pattern can be described as an increase of discharge through the winter to spring months and a decrease in discharge through the summer through fall months. When nitrate (NO3) and silica (SiO2) concentrations are added to the graph however some differences can be observed. The nitrate (NO3) concentrations in the Conodoguinet Creek 19
demonstrate an inverse relationship with discharge while the Canajoharie Creek demonstrates a direct relationship between the two variables. While the Conodoguinet Creek and the Canajoharie Creek demonstrate a different relationship between nitrate (NO3) concentrations and discharge the relationship between silica (SiO2) concentration and discharge are similar. The silica (SiO2) concentration follows the pattern of discharge in both the Conodoguinet Creek and the Canajoharie Creek. When the nitrate (NO3) and silica (SiO2) concentration relationship is examined another fundamental difference between the two streams can be seen. In the Conodoguinet Creek the nitrate (NO3) and silica (SiO2) concentration relationship demonstrates a negative and relatively weak correlation (r 2 = 0.18) when compared to the Canajoharie Creek which demonstrates a positive and relatively strong correlation (r 2 = 0.93). The differences between the two streams could possibly be attributed to the differences in land use (Conodoguinet Creek = primarily forested and Canajoharie Creek = primarily agriculture) and differences in the relationship between nitrate (NO3) and discharge relationships between the two streams. Even during drought conditions the Conodoguinet Creek does not show a relationship between nitrate (NO3) and silica (SiO2) concentrations while Wall et al. (1998) demonstrated a positive relationship between nitrate (NO3) and silica (SiO2) concentrations. King et al. (2006) found that nitrate (NO3) concentrations are generally higher in winter months and which has been show to coincide with the low flow periods of Conodoguinet Creek. The general pattern followed by discharge and silica (SiO2) concentrations that both increase beginning in winter through spring months and decreases through summer and fall months is found in numerous studies (Wang and Evans 1969; Edwards 1974; Casey et al. 1981). 20
These results can be used with other water quality data to help shape water quality standards and protection practices. It should be noted that both this study and the study conducted by Wall et al. (1998) took place in smaller scale watersheds and both watersheds contain high agricultural land use. Further research in larger scale watersheds with more varied land use would be beneficial to discover what types of nitrate (NO3) and silica (SiO2) relationships occur in a more diverse aquatic environment. 21
7.0 References Casey H, Clarke R, Marker A. 1981. The seasonal variation in silicon concentration in chalk relation to diatom growth. Freshw. Biol. 11:335 344. streams in Clarke E, Baldwin AH. 2002. Responses of wetland plants to ammonia and water level. Ecol. Eng. 18:257 264. Edwards A. 1974. Silicon depletions in some Norfolk rivers. Freshw. Biol. 4:267 274. Heathwaite AL, Johnes P. 1996. Contribution of nitrogen species and phosphorus fractions to stream water quality in agricultural catchments. Hydrol. Process. 10:971 983. Hornberger GM, Scanlon TM, Raffensperger JP. 2001. Modelling transport of dissolved silica in a forested headwater catchment: the effect of hydrological and chemical time scales on hysteresis in the concentration discharge relationship. Hydrol. Process. 15:2029 2038. Iler RK. 1979. The chemistry of silica. Irani FM, Claggett P. 2010. Chesapeake Bay Watershed Land Cover Data Series. Jaynes D, Hatfield J, Meek D. 1999. Water quality in Walnut Creek watershed: Herbicides and nitrate in surface waters. J. Environ. Qual. 28:45 59. Killpack S, Buchholz D. 1993. Nitrogen in the Envrionment: Nitrogen s Most Common Forms. King K, Hughes K, Balogh J, Fausey N, Harmel R. 2006. Nitrate-nitrogen and dissolved reactive phosphorus in subsurface drainage from managed turfgrass. J. Soil Water Conserv. 61:31 40. Myers W, Bishop J. 1999. County boundaries of Pennsylvania. Pennsylvania Department of Transportation. 2016. PennDOT - Pennsylvania State Boundaries. Seaber P, Kapinos F, Knapp G. 2007. Hydrologic Unit Maps. US Geological Survey. 2005. National Water Quality Monitoring Council - Water Quality Data. [accessed 2016a Feb 15] US Geological Survey. 2007-2014. NHDFLowline- Susquehanna. US Geological Survey. 2005. USGS Surface-Water Daily Data for Pennsylvania. Vogelmann J, Sohl T, Campbell P, Shaw D. 1998b. Regional land cover characterization using Landsat Thematic Mapper data and ancillary data sources. Environmental Monitoring and Assessment 51:415 428. Vogelmann J, Sohl T, Howard S. 1998a. Regional characterization of land cover using multiple sources of data. Photogram-metric engineering and remote sensing 64:45 57. 22
Wall G, Phillips P, Riva-Murray K. 1998. Seasonal and Spatial Patterns of Nitrate and Silica Concentrations in Canajoharie Creek, New York. Journal of Environmental Quality 27:382 389. Wang W, Evans RL. 1969. Variation of silica and diatoms in a stream. Limnol. Oceanogr. 14:941 944. Zairchansky J. 1986. Soil Survey of Cumberland and Perry Counties, Pennsylvania. US Geological Survey. 23