September and into October, in some cases almost climbing back to their earlier (August-early September) range of values. In contrast to the nitrates, the total phosphates had the highest encountered standard deviations ((.09 mg/L) shown in Appendix 1). Values ranged from.018 mg/L to 1.16 mg/L with a season average of.24 mg/L. The pattern, illustrated in Figure 4.2, shows less uniformity than observed in the nitrates, with error bars making it much harder to discern patterns. What can be seen though is a strong spike in values (.50 mg/L) from the middle of June into early July and an even more drastic drop (.55 mg/L) later in the month with values continuing on a slow decline into late September. Vertical Profile The nitrates for the vertical sampling had lower standard deviations on average (.04) than any of the other sampling runs on the lake (Appendix 2). The average nitrate concentration for the column is.86 mg/L, with values falling between.35 mg/L and 1.33 mg/L. Unlike the uniform pattern observed for nitrates for the horizontal sampling, the vertical pattern seen in Figure 4.3 shows more of a grouping by depths that share distinctive behaviors. Temporally, the four shallowest collection depths (2 meters, 4 meters, 6 meters, and 9 meters) showed a regular and fairly consistent, but modest decrease (.25 mg/L per month) throughout the sampling with only minimal dissimilarity between the depths. When the column reaches a depth of 12 meters a different temporal pattern emerges. In these cases low initial values are followed by an increase (.24 mg/L), forming a plateau that is eventually interrupted by a less dramatic drop (.22 mg/L). Though the late- season values are lower than early and mid-season nitrate values at these depths, they still remain higher than at the surface. In looking at the spatial pattern observed for each of the collection depths, a progression is seen throughout the sampling run. Initially the highest nitrates values for each run are seen at depths between 6 and 9 meters. As the run continued, this bracket moved lower into the lake until it finally settled between the depths of 15 and 18 meters, the closest measurements to the bottom of the lake at the site. The vertical profile for total phosphates collected yielded no discernable pattern. Raystown Branch/ Trough Creek Samplings The nitrate values for the Raystown Branch show an initial concentration of 1.4 mg/L, shown in Figure 4.4, in late June followed by slightly lower values as the season moves into July (Appendix 3). It then roughly plateaus, staying between 1.0 mg/L and 1.1 mg/L, through the remainder of July and all of August. In mid-September nitrate levels increase to near initial values (1.35 mg/L) and slightly diminish going into October. The season average for the Raystown Branch was 1.18 mg/L with a range between 1.02 mg/L and 1.40 mg/L. The nitrate pattern for Trough Creek exhibited a similar drop from late June into early July, but at an overall concentration of.54 mg/L. In late July levels surge again to a concentration comparable to that in late June. In mid-August the nitrate concentration drops down to a fairly stable plateau, with values hovering between.28 mg/L and.33 mg/L, for rest of August and into late September. The pattern undergoes another modest decline as October begins, bringing levels to nearly half of those collected in the two prior months. An average of.33 mg/L and a range between.13 mg/L and.54 mg/L were determined for Trough Creek. The reactive phosphates for both sites exhibited a series of increases and decreases through the season that don’t render an identifiable pattern. However, late season concentrations are in both cases more than double those from early in the summer (Figure 4.5). The total phosphates exhibited a similar erratic pattern. Late season patterns were lower than early in the summer. What makes these data interesting is not their exact levels but their levels in relation to the reactive phosphates, with samples processed for total phosphate showing lower concentrations than the recorded reactive phosphate levels for a majority of the sampling. Discussion In trying to understand the observed nutrient patterns of Raystown Lake it is important we note that reservoirs have important characteristics that set them apart from natural lakes. The limnology of lakes is understood largely on the basis of studies of small, natural lakes (Wetzel; 2001). We should expect that the behavior of a reservoir will differ from the well- characterized lake pattern. Recognizing these differences is necessary for an accurate interpretation of information gathered. One of the most notable dissimilarities is in their geomorphology. Whereas lakes tend to be elliptical and become shallower toward their outlet, reservoirs are often long, sinuous, and deep. Reservoirs are typically fed from a tributary, usually some distance from their point of discharge. Because of this distance between the input-discharge points and increasing depth, a gradient representing change in chemical and biological characteristics is developed. This gradient has been segmented into three prominent zones (Figure 5.1), the riverine, transition, and lacustrine (Thornton; 1990). The riverine represents that part of the reservoir still retaining aspects of input, including the capability to suspend fine particulate matter and sediments. These limit the biological production in this shallow, uplake, and well-mixed segment. The transition zone can be defined by the point in the lake where velocity has decreased enough for sediment settling to be prominent, mostly easily recognized by increased light penetration and biologic production. The lacustrine segment is the part of the reservoir that bears the most resemblance to natural lakes, with high light penetration, high enough biologic production to create the potential for nutrient limitation, and vertical division of the lake into epi,, hypo, and metalimon. Another important distinction between reservoirs and lakes involves residence time. The depth of a lake has a direct effect on the residence time and the concentration of most non-conservative substances, specifically nitrates and phosphates for this discussion, as depicted in Figure 5.2 (Kennedy and Walker; 1990). Because of this difference in behavior, an interpretation of seasonal patterns of nutrient concentrations in reservoirs must take into account this longer residence and its implications. NUTRIENT CONCENTRATIONS AND TEMPORAL NITRATES IN Haley, Christopher J. 1, 1 Department of Geology Figure 4.1 Figure 4.2 Figure 4.3 Figure 4.4Figure 4.5

Nitrates In lakes and reservoirs nitrate concentrations represent only a portion of a set of processes known as the nitrogen cycle. Only a few key components are examined in most studies, including nitrogen (N2), ammonia (NH4+), and nitrate (NO 3 ). Nitrates were selected for this study because prior studies had focused on this nutrient, making it possible to compare patterns from recent years and decades past. Also weighing heavily on the decision was that collection and processing of nitrates was relatively easy and provided direct indication of water quality. Moreover, for at least the well- oxygenated epilimnial samples collected for the longitudinal profile of the lake, the redox conditions should preclude ammonia as a significant constituent. The horizontal profile collected for nitrates exhibits behavior typical of phosphate-limited lakes in temperate regions. Previous investigation of Raystown Lake reveals a consistent behavior in nitrates (Williams, 1978). Williams reported elevated concentrations ( mg/L) of nitrate that accumulated during the winter months (November-February). These high winter concentrations began declining as early as March and continued to decline throughout the summer reaching mg/L by mid-September. Beginning in October the concentrations once again began to rise and continued to do so throughout the rest of the winter months. He attributed the decline in nitrates coinciding with spring turnover to phytoplankton, which aggressively begin to assimilate nitrate during this time. This explanation fits well with this study’s findings, though the length of our sampling period doesn’t extend into spring turnover. Spatially, the lake’s horizontal nitrate profile initially displays a pattern consistent with Thornton’s model division of a reservoir into riverine, transitional, and lacustrine zones (1990). None of the sites selected for nutrient analysis could realistically be identified as riverine conditions. However, although no nutrient data were collected, temperature, depth, dissolved oxygen, and visibility were all recorded on a biweekly basis throughout the summer for a site further uplake. These attributes all showed a strong riverine character. Most notable was a significant amount of suspended particles accompanied by minimal light penetration. Site R22 (22 miles uplake from the breast of the dam) exemplifies the gradational character of the transitional zone in Thornton’s model. Seechi data gathered for this site marks an appreciable increase in light penetration compared to sites surveyed further upstream (Appendix 4). Site R12, while further down lake (18 miles uplake from the dam), shares a similar degree of light penetration. Also, nitrate levels for both sites are lower than those measured for the major input to the lake, the Raystown Branch of the Juniata River, near Saxton. This trait is accounted for in Kennedy and Walker’s residence time/concentration models for basic reservoirs, demonstrated in Figure 5.2 (1990). What prohibits including R12 in a transitional zone is the well-defined stratification observed from the vertical sampling profile gathered between these two sites (Figure 3.1). Where Raystown Lake begins to deviate from Thornton’s model is in its multiple tributaries. With a considerable amount of inflow from Hawn’s Run, James Creek, Coffee Run, Shy Beaver Creek, Trough Creek, Tatman Run, and Shoup Run (Figure 3.1) entering at a variety of locations along the length of the lake. Therefore, substantial portions of what might be considered lacustrine in Thornton’s model are actually riverine and transitional. Defining these zones for the seven additional tributaries would demand an impressive amount of additional work further hindered by seasonal stream fluctuation and a 2.4 meter drawn down that can be expected during recreational periods, drastically altering lake morphology (Williams 1978). Fortunately sites R10, R8, R21, and R1 (Figure 3.1) all lie within regions that can be safely identified as lacustrine, each satisfying the basic criteria of high light penetration, stratification (interpreted from temperature/dissolved oxygen data (Simpson 1998)), and low suspended sediment. The nitrate vertical component profile of Raystown Lake closely follows usual patterns for phosphate-limited temperate lakes, as described by Manny and Wetzel (1982). The shallowest depths (2,4, and 6 meters) adhere to expected concentrations (in relation to the column) and behavior throughout the sampling window of mid-June to early September, with a nearly linear decrease in these months (Figure 4.1). At 9 meters the highest value for column (1.33 mg/L) is attained early on in the sampling and then proceeds down similar linear fashion stated before. The three deeper measurements (12, 15 and 18 meters) show noticeably lower initial values than the four shallower samples, each progressively lower with depth, though all showed a dramatic increase, between June 15 and June 26, bringing their values to comparable if not higher levels than the shallower measurements. In late August these levels retreat, though they remain appreciably higher than the shallow samples. This behavior can be explained by taking into account first a known seasonal behavior in the lake, phytoplankton fixation of nitrates. Increased biological activity that accounts for nitrate depletion horizontally through the lake can also be seen at the shallower depths of this vertical column. The initial increase seen at lower depths can be accounted for by recognizing the column’s position in the transitional zone of Thornton’s model (This determination was made on the basis of a relative comparison of information gathered from this area during this study, extensive work done with dissolved oxygen in previous projects (Simpson 1998), and known lake morphology.) The relevance of assigning the column to a transitional area of the lake is that it implies, most importantly, increased sediment and particulate settling due to the low velocity of the water. This settling carries nitrate down into lower levels of the lake in sediment to which it has sorbed and become immobilized (Keeney, 1973). As the summer progresses into fall, lake nitrate levels become so low above (prior to fall turnover), that the system’s input of nitrate is exceeded by the sedimentation rate. Fall turnover reverses this decline as nitrate begins to accumulate, due in part to decreased biological activity and continued input from precipitation and particulate fallout from atmospheric sources (Wetzel, 2001). Sampling of the two most substantial inputs into Raytown Lake, the Raystown Branch and Trough Creek, embodied the relationship between watershed characteristic and nutrient concentration. While it has been demonstrated that forest areas have the potential to be sources of many components of nitrates (Lepisto 2000), it is far more common for a system to receive the bulk of its nitrates from agricultural land. This determination was substantiated through the sampling of these two near-end members of the agricultural/forest dominated watershed spectrum (Figure 4.4). In comparison, the nitrates from the Raystown Branch are on the order of 3 times the concentration of those for Trough Creek. At the end of the sampling seasons for both inputs we see expected behaviors for both as well. The agriculturally dominated watershed of the Raystown Branch shows the elevated nitrate commonly associated with farming. The forested-dominated watershed of Trough Creek exhibits a gradual decrease as biological activity from the fauna in the lake turn towards conserving and storing nitrates for the winter (Gustafson, 2001, personal communication). Phosphates Phosphorus has been studied extensively in fresh waters. Its role as a limiting nutrient in many streams, lakes, and reservoirs as well as its clear tie to processes like algal bloom and eutophication are the reasons for this. Like many other nutrients observed in aqueous systems, phosphorus can be found in a variety of forms. Convention has been to differentiate it into the four following categories: soluble reactive phosphorus, soluble unreactive phosphorus, particulate reactive phosphorus, and particulate unreactive phosphorus (Strickland and Parsons, 1972). Considerable attention has also been paid to differentiating organic and inorganic forms. For most freshwater data sets though, only total phosphorus and orthophosphate are reported (Wetzel, 2001). Alternate week sampling of total phosphates in the horizontal profile and vertical column of Raystown Lake and both total and reactive phosphates for the Raystown Branch of the Juniata River and Trough Creek was conducted during the course of this study. This sampling scheme was chosen to bring resolution to the behavior of phosphates within the reservoir and offer a means of comparison to prior measurements taken of Raystown by Williams (1978) and Smith (2001). An understanding of the movement of phosphorus in the epilimnion of the reservoir is necessary to explain the patterns seen in the horizontal profile of Raystown Lake. Figure 5.4 gives clarification to processes within this zone of freshwater lakes (Lean, 1973). As designated by the figure, the biological activity (noted as “plankton and other seston”) of freshwater lakes accounts for an overwhelming majority of where phosphorus is tied up. It stands to reason then that fluctuation in phosphorus levels in the lake will have a direct effect on the biological productivity. Direct, almost linear relationships have been established, with increasing chlorophyll content (an indicator of biological productivity) in waters being accompanied by increased total phosphorus (Forsberg and Ryding, 1980). Knowing this is useful because it allows for a loose correlation of phosphorus variation with that of biological productivity. The relationship between phosphorus and biological productivity can be utilized in the examination of temporal behavior in the horizontal profile for total phosphates. The average total phosphate concentration for the longitudinal profile of Raystown Lake was.77 mg/L, with a range of.28 mg/L to 1.16 mg/L. Higher values within this range were measured from mid-June to early July, a peak (averaging.63 mg/L) that temporally associates well with patterns observed by both Williams (1978) and Smith (2001). Where there is a substantial contrast is in the absolute amounts of phosphorus measured by Williams, which are nearly an order of magnitude lower than values collected in this study as VARIATIONS: A STUDY OF PHOSPHATES AND RAYSTOWN LAKE, PA Mutti, Laurence J. 1 at Juniata College