ABSTRACT Title of Thesis: EFFECTS OF TYPICAL CONCENTRATIONS OF NITROGEN AND PHOSPHORUS IN AGRICULTURAL TREATMENT WETLANDS ON POLYCULTURES OF TYPHA LATIFOLIA L. AND JUNCUS EFFUSUS L. AND A TEST OF THE N:P RATIO AS A PREDICTIVE TOOL. Timothy W. Culbertson, Master of Science, 2005 Thesis Directed By: Associate Professor Andrew H. Baldwin, Department of Marine-Estuarine-Environmental Science Nitrogen removal in agricultural treatment wetlands is determined by the health of macrophytes. Nitrogen is removed by microbes dependant on an oxic-anoxic boundary layer created by oxygen flow through macrophyte culms and by nitrogen sequestration in plant biomass. Phytotoxic ammonia concentrations can limit plant growth. Orthophosphate fertilization to balance the nitrogen to phosphorus tissue ratio may increase biomass. One test of this hypothesis investigated Juncus effusus and Typha latifolia ammonia phytotoxicity: ammonia concentrations above 150 mgL -1 suppress Typha growth and above 300 mgL -1 do not affect Juncus growth. The second experiment altered mesocosm N:P ratios. Some alleviation of ammonia toxicity was shown; highest biomass production at toxic levels occurred at a 15:1experimental N:P ratio. The two species used N and P differently: the average N:P ratio was 12:1 in Juncus and 9.1:1 in Typha. More effective waste removal may be based upon more precise control of N:P ratios. EFFECTS OF TYPICAL CONCENTRATIONS OF NITROGEN AND PHOSPHORUS IN AGRICULTURAL TREATMENT WETLANDS ON POLYCULTURES OF TYPHA LATIFOLIA L. AND JUNCUS EFFUSUS L. AND A TEST OF THE N:P RATIO AS A PREDICTIVE TOOL. by Timothy W. Culbertson Thesis Submitted to the Faculty of the Graduate School of the University of Maryland at College Park in partial fulfillment of the requirements for the degree of Master of Science 2005 Advisory Committee: Professor Andrew H. Baldwin, Chair Professor Patrick Kangas Professor Joseph Sullivan ACKNOWLEDGEMENTS This thesis was made possible in part due to the continued help and contributions from many people in the University of Maryland Community. I wish to thank Dr. Andrew H. Baldwin for serving as my mentor, advisor, and committee chair. His support and guidance were wonderful, without which this thesis would not have been completed. I would also like to thank Drs. Patrick C. Kangas and Joseph H. Sullivan for serving on my committee and providing valuable criticism of this work. Several students from Dr. Baldwin?s laboratory provided assistance in the greenhouse and otherwise, and for their aid I would like to thank Talia Chalew, Mike Egnotovich, Rachael Herbert, Kelly Neff, Jessica Peterson, Mary Pittek, Krissy Rusello, and Rebecca Stack. I also wish to thank the staff of the UMCP greenhouses for their help in providing bench space and technical assistance. Additionally, I am deeply grateful to my immediate family for their support in this endeavor, and in particular Katherine R.M. Mackey without whose constant assistance this work would have been impossible. ii TABLE OF CONTENTS List of Tables???????????????????????????. iv List of Figures??????????????????????????? vi Chapter I: Introduction and Background??????????????? 1 Research Experiments????????????????????? 14 Chapter II: Effect of Ammonia on Biomass In Mesocosms of Typha latifolia and Juncus Effusus???????????? 17 Objective?????????????????????????? 17 Methods??????????????????????????. 17 Preparation of Mesocosms????????????????. 17 Treatment Solution, Plant Growth, and Statistical Analysis???. 21 Results and Discussion????????????????????.. 24 Conclusions????????????????????????... 51 Chapter III: The Vegetation N:P Ratio in Mesocosms Subjected to High Ammonia Concentrations??????????????. 53 Objective?????????????????????????? 53 Methods??????????????????????????. 53 Preparation of Mesocosms????????????????. 53 Treatment Solution, Plant Growth, and Statistical Analysis???. 54 Results and Discussion????????????????????.. 59 Mesocosm Biomass Production??????????????. 59 Mesocosm Root-to-Shoot Ratios?????????????... 77 Mesocosm Tissue Nitrogen Concentration?????????? 80 Mesocosm Tissue Mass Nitrogen?????????????.. 95 Mesocosm Tissue Phosphorus Concentration????????... 112 The N:P Ratio as a Predictive Tool????????????.... 123 Conclusions????????????????????????? 132 Chapter IV: Synthesis??????????????????????? 134 Appendix I: SAS Code for Chapter II Statistical Analyses????????... 141 Appendix II: SAS Code for Chapter III Statistical Analyses????????.. 160 References????????????????????????????.. 193 iii LIST OF TABLES 2.1. Summary of nutrient additions to prepare treatment solutions in 6 L increments in 2002???????????????............................. 20 2.2. Tests of fixed effects on total, above-ground, and below-ground biomass of mesocosms of Typha latifolia and Juncus effusus in 2002?????? 25 2.3. Tests of fixed effects on total, above-ground, and below-ground biomass of Typha latifolia in 2002???????????????????.. 29 2.4. Tests of fixed effects on total, above-ground, and below-ground biomass of Juncus effusus in 2002???????????????????... 34 2.5. Summary of statistical analyses of differences in biomass production between Typha latifolia and Juncus effusus over an ammonia gradient for 2002.??????????????????????????. 39 2.6. Test of fixed effects on the root-to-shoot ratio in mesocosms and in Typha latifolia and Juncus effusus in 2002????????????? 44 3.1 Summary of nutrient additions to prepare treatment solutions in 6 L increments in 2003????????????????????.?? 55 3.2 Tests of fixed effects on total, above-ground, and below-ground biomass of mesocosms of Typha latifolia and Juncus effusus in 2003?????? 60 3.3 Tests of fixed effects on total, above-ground, and below-ground biomass of mesocosms of Typha latifolia in 2003?????????????... 66 3.4 Tests of fixed effects on total, above-ground, and below-ground biomass of mesocosms of Juncus effusus in 2003?????????????... 72 3.5 Test of fixed effects on the root-to-shoot ratio in mesocosms and in Typha latifolia and Juncus effusus in 2003???????????..?.. 78 3.6 Tests of fixed effects on total, above-ground, and below-ground tissue nitrogen concentration of mesocosms of Juncus effusus and Typha latifolia in 2003????????????????????... 83 3.7 Tests of fixed effects on total, above-ground, and below-ground tissue nitrogen concentration of Typha latifolia in 2003??????????. 90 3.8 Tests of fixed effects on total, above-ground, and below-ground tissue nitrogen concentration of Juncus effusus in 2003??????????.. 94 iv 3.9 Tests of fixed effects on total, above-ground, and below-ground tissue total mass nitrogen in mesocosms of Typha latifolia and Juncus effusus in 2003????????????????????.. 99 3.10 Tests of fixed effects on total, above-ground, and below-ground tissue total mass nitrogen in Typha latifolia in 2003???????????... 106 3.11 Tests of fixed effects on total, above-ground, and below-ground tissue total mass nitrogen in Juncus effusus in 2003???????????.... 110 3.12 Tests of fixed effects on total, above-ground, and below-ground tissue phosphorus concentration of mesocosms of Typha latifolia and Juncus effusus in 2003????????????????????... 115 3.13 Tests of fixed effects on total, above-ground, and below-ground tissue phosphorus concentration of Typha latifolia in 2003????????? 118 3.14 Tests of fixed effects on total, above-ground, and below-ground tissue phosphorus concentration of Juncus effusus in 2003?????????. 121 v LIST OF FIGURES 1.1. Conceptual framework of experiments summarizing direct and indirect effects of ammonia addition to ATWs??????????????... 5 2.1. Aspects pertaining to greenhouse treatment design?????????... 19 2.2. Interstitial water conductivities (?S) for each mesocosm at sampling dates in 2002 at experimental ammonia concentrations???????? 22 2.3. Total mesocosm biomass along a gradient of ammonia concentration treatments in 2002????????????????.......................... 26 2.4. Total above-ground mesocosm biomass along a gradient of ammonia concentration treatments in 2002????????????????... 27 2.5. Total below-ground mesocosm biomass along a gradient of ammonia concentration treatments in 2002????????????????... 28 2.6. Total Typha latifolia biomass along a gradient of ammonia concentration treatments in 2002????????????????... 30 2.7. Above-ground Typha latifolia biomass along a gradient of ammonia concentration treatments in 2002????????????????... 32 2.8. Below-ground Typha latifolia biomass along a gradient of ammonia concentration treatments in 2002????????????????... 33 2.9. Total Juncus effusus biomass along a gradient of ammonia concentration treatments in 2002????????????????... 35 2.10. Above-ground Juncus effusus biomass along a gradient of ammonia concentration treatments in 2002????????????????... 36 2.11. Below-ground Juncus effusus biomass along a gradient of ammonia concentration treatments in 2002????????????????... 37 2.12. Total species biomass along a gradient of ammonia concentration treatments in 2002??????????????????????.. 40 2.13. Above-ground species biomass along a gradient of ammonia concentration treatments in 2002????????????????... 41 2.14. Below-ground species biomass along a gradient of ammonia concentration treatments in 2002????????????????... 42 vi 2.15. Root-to-shoot ratios for mesocosms along a gradient of ammonia concentration treatments in 2002????????????????... 45 2.16. Root-to-shoot ratio for Typha latifolia along a gradient of ammonia concentration treatments in 2002????????????????... 47 2.17. Root-to-shoot ratio for Juncus effusus along a gradient of ammonia concentration treatments in 2002????????????????... 48 2.18. Species root-to-shoot ratios along a gradient of ammonia concentration treatments in 2002??????????????????????.. 49 3.1 Interstitial water conductivities (?S) for each mesocosm at sampling dates in 2003????????????????????????.. 57 3.2. Total mesocosm biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???????.. 61 3.3. Total above-ground mesocosm biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 63 3.4. Total below-ground mesocosm biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 64 3.5. Total Typha latifolia biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???????.. 67 3.6 Above-ground Typha latifolia biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???.. 69 3.7 Below-ground Typha latifolia biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???.. 70 3.8. Total Juncus effusus biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???????.. 73 3.9. Above-ground Juncus effusus biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???.. 75 3.10. Below-ground Juncus effusus biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???.. 76 3.11. Root-to-shoot ratios for mesocosms along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???????.. 79 vii 3.12 Root-to-shoot ratios for Typha latifolia along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???????.. 81 3.13. Root-to-shoot ratios for Juncus effusus along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???????.. 82 3.14. Total mesocosm tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???.. 84 3.15. Total above-ground mesocosm tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??????????????????????????... 86 3.16. Total below-ground mesocosm tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??????????????????????????... 88 3.17. Total Typha latifolia tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??. 91 3.18. Above-ground Typha latifolia tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??????????????????????????... 92 3.19. Below-ground Typha latifolia tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??????????????????????????... 93 3.20. Total Juncus effusus tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 96 3.21. Above-ground Juncus effusus tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??????????????????????????... 97 3.22. Below-ground Juncus effusus tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??????????????????????????... 98 3.23. Total mesocosm tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???????.. 100 3.24. Total above-ground mesocosm tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 102 viii 3.25. Total below-ground mesocosm tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 104 3.26. Total Typha latifolia tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???????.. 107 3.27. Above-ground Typha latifolia tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 108 3.28. Below-ground Typha latifolia tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 109 3.29. Total Juncus effusus tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003???????.. 111 3.30. Above-ground Juncus effusus tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 113 3.31. Below-ground Juncus effusus tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 114 3.32. Total mesocosm tissue phosphorus concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 116 3.33. Total Typha latifolia tissue phosphorus concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 120 3.34. Total Juncus effusus tissue phosphorus concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003??.. 122 3.35. The vegetation N:P ratio of total Juncus effusus tissue and total Typha latifolia tissue presented as tissue nitrogen versus tissue phosphorus at all experimental N:P ratios and all ammonia concentration treatments in 2003??????????????????????.. 124 3.36. The vegetation N:P ratio of each experimental N:P ratio presented as tissue nitrogen versus tissue phosphorus at all ammonia concentration treatments and for mesocosms of Juncus effusus and Typha latifolia in 2003........................................................................................................... 125 3.37. The vegetation N:P ratio of each experimental N:P ratio presented as tissue nitrogen versus tissue phosphorus at all ammonia concentration treatments and for Typha latifolia in 2003?????????????. 126 ix 3.38. The vegetation N:P ratio of each experimental N:P ratio presented as tissue nitrogen versus tissue phosphorus at all ammonia concentration treatments and for Juncus effusus in 2003????????????? 127 x CHAPTER ONE: INTRODUCTION AND BACKGROUND Of paramount concern in agricultural areas, in particular those in close proximity to natural watersheds of ecological and economic importance, is the control of anthropogenically derived nitrogenous wastes entering those watersheds (Bobbink et al.. 1998, Aerts and Bobbink 1999). In addition to nitrogenous compounds, wastewater produced by dairy farms, feedlots, and hog farms also tends to have high carbon content, high total suspended solids, and often high phosphorus content as well (Cronk 1996); the sum effect of addition of excess amounts of these nutrients to natural waters may contribute to eutrophication (Cronk 1996, Knight et al. 2000). Concentrations of 100- 300 mgL -1 of ammonia and 75-200 mgL -1 of total phosphorus are fairly common for agricultural wastewater. Hammer (1992) reports concentrations of up to 500 mgL -1 in raw livestock wastewater. Production of wastewater in agricultural systems typically stems from cleaning practices (Tanner et al. 1995a), which flush excretions, mud, feed, waste milk, and detergents from barn or dairy parlor floors with high-pressure hoses (Cronk 1996, Tanner et al. 1995a). Flush water volumes tend to be quite high: Soil Conservation Service estimates suggest a total of 100 gallons per cow per day (SCS 1991), and independent researchers suggest almost an order of magnitude higher (Surrency 1993). Raw flush water is normally treated, though not always, prior to addition to natural waters, but tends to retain some portion of its high nutrient load (Clarke 1999, Cronk 1996). In such typically nitrogen-limited natural systems, additions of high nitrogen compounds such as ammonium (NH 3 ) and nitrate (NO 3 - ) tend to alter species 1 composition and biomass production (Shaver and Chapin 1995, Lammerts et al. 1999, Van Duren and Pegtel 2000, Camacho et al. 2003). Biomass production in natural systems tends to be enhanced by the addition of a limiting nutrient; the addition of non- limiting nutrients tends to have little or no effect on biomass production (Vitousek and Howarth 1991, Verhoeven et al. 1996, Gusewell et al. 2003). Species composition is affected by the addition of limiting or non-limiting nutrients, due to differential species responses to different nutrients (Dijk and Olff 1994, Mamolos et al. 1995, Shaver and Chapin 1995, Lammerts et al. 1999). Species responsiveness to nutrient availability differs, then, dependant on relative nutrient requirements and on their ability to take up and use the available nutrients (Kielland 1994, Perez-Corona et al. 1996, Ryser et al. 1997, Aerts and Chapin 2000). In watersheds, the most responsive species tend to be algae and phytoplankton (e.g. Rhee 1978, Downing and MacCauley 1992); responses are most frequently seen to increases in available nitrogen rather than increases in available carbon (Camacho et al. 2003) or in available phosphorus (Jansson et al. 2001). Fertilization increases are characterized by rapid biomass production of one or few species, in response to increases in previously limiting nutrients (Grime 1979, Camacho et al. 2003). This rapid biomass production is associated with decreases in species richness (Bedford et al. 1999, Kent et al. 2000, Green and Galatowitsch 2003, see Mittelbach et al. 2001 for an excellent synopsis), and increases in the presence of invasive species (Green and Galatowitsch 2002), which may end in decrease in economic and ecological function in the watershed or other system (Engelhardt et al. 2001, Ostertag and Verville 2002). 2 Methodologies to remove the nitrogen component of agricultural system wastes were historically limited to the creation or restoration of riparian buffer strips of naturalized vegetation (e.g. Brix 1989, Vought et al. 1993, Konya et al. 1995, Tanner et al. 1995a, Craft 1999). Stemming from the theories that such strips remove contaminants by translocating chemicals into the biomass and providing suitable habitat for microbial colonization (and subsequent redoxic removal mechanisms), the use of engineered wetlands as dedicated waste water purifiers has been investigated and implemented in a range of conditions and at a range of sites (e.g. Hammer 1992, Cronk 1996, Karpiscak et al., 1999, Knight et al. 2000). Current estimates indicate that 65% to 85% of nitrogenous contaminants are removed by vegetation and bacteria as water moves through an agricultural treatment wetland (ATW) (Cronk 1996, Knight et al. 2000). Of nitrogenous compounds (which can include organic forms such as urea, amino acids, amines, and nucleic bases, and inorganic forms such as ammonia, nitrite, nitrate, and nitrogen gases), ammonia is the most problematic to remove, and percentage removal rates are often significantly lower than the removal rates of other compounds, nitrogenous and otherwise (i.e. Mitsch and Gosselink 1993, Hammer and Knight 1994, Cronk 1996, Humenik et al. 1999, Clarke 1999, Knight et al. 2000, Clarke and Baldwin 2002). Ammonia, nitrite, and nitrate are recognized pollutants of natural waters, with toxicity limits of ammonia placed as low as 0.2 mgL -1 (Kadlec and Knight 1996). Maximum removal efficiency of ammonia tends to be less than 90%, with some estimates as low as 48% (Knight et al. 2000). On the other hand, under conditions of maximum efficiency, up to 97% of biological oxygen demand (a surrogate measure for total carbon) can be removed, as can 99% of total nitrogen, and 3 93% of total phosphorus (Cronk 1996, Clarke 1999, Schaafsma et al. 2000,Knight et al. 2000, DeRico 2000, Clarke and Baldwin 2002). As previously explained, poor removal of ammonia or other nitrogenous compounds from wastewater can lead to eutrophication in nitrogen-limited natural systems; nitrogen removal becomes, therefore, an important goal of ATWs (Cronk 1996, Clarke 1999, Knight et al. 2000). Of the portion removed from the wastewater, approximately 40% of this nitrogen is translocated into plant biomass and used for the production of proteins, nucleic bases, protonated hydrocarbons, etc (Reddy et al. 1989). The remaining 60% is removed through the redoxic nitrification- denitrification pathway (Howes et al.. 1981, Gumbricht 1993, Bachand and Horne 2000a), although percentage values as to the treatment effects of plants on nitrogenous wastes range from 90% (Rogers et al. 1991) to just 5% (Hammer 1992). In any event, the importance of the nitrogen-coupled bacterial reaction is not to be overlooked; through this pathway, ammonium (NH 4 + ) is oxidized to nitrate, which is then reduced to nitrogen gas (N 2 or N 2 O), which leaves solution. Figure 1.1 illustrates these direct effects as well as associated indirect effects. Clearly, then, this pathway requires an oxygen gradient; translocation in plants of oxygen from atmosphere to rhizosphere through aerenchymous tissues helps to provide an anoxic-oxygenated interface over which this reaction may occur (Brix 1989, Brix et al. 1992, Bendix et al. 1994, Tornbjerg et al. 1994, Brix et al. 1996, Brix and Sorrell 1996, Brix 1997, Jespersen et al. 1998, Smith et al. 2000). Tanner and Kadlec (2003) consider rhizosphere oxygenation to be the rate-limiting step in nitrogen removal in treatment wetlands. Greater plant biomass results in greater oxygen flow into the rhizosphere, thus increasing the degree of development of this boundary layer. This gas 4 High NH 4 + Plant Biomass N-Uptake Rate Denitrification Nitrification N-Removal Rhizosphere Oxidation + + - + + +? + + - - ? Figure 1.1. Conceptual framework of experiments summarizing direct and indirect effects of ammonia addition to ATWs. Signs associated with effects indicate feedback within the system. Question marks may indicate where further research could add to the precision with which ATWs are modelled. 5 flow rate differs based on plant species, but in each case increases as biomass increases (Brix 1989, Brix et al. 1996, Romero et al. 1999). Thus, biomass may be seen as a rough surrogate for degree of development of oxygenation in the root zone, the resolution of which depends upon the gas flux constant of the species used. Root zone oxygenation is likewise proportional to degree of nitrification given that the bacteria necessary for such are in abundance in an oxygenated system. Decreases in biomass in treatment wetlands are therefore problematic in terms of nitrogen removal on two levels: (1) lower biomass production equates to lower translocation rate of nitrogen into plant tissue; (2) lower biomass equates to lower oxygen flow into the rhizosphere, lower degree of development of the redoxic boundary, and lower rate of nitrification. Additionally, (3) lower biomass equates to less surface area in the water column itself upon which nitrifying bacteria may be supported (a potentially overlooked but experimentally valid assumption; see Bastviken et al. 2003) (Figure 1.1). Difficulties arise in designing ATWs for effective removal of nitrogenous wastes due to a lack of standardization in flow hydrology and chemical load into the system (Hammer and Knight 1994). The reactions which remove nitrogen tend to have a low rate of reaction; slower flows will allow greater amounts of denitrification. In the case of some treatment wetlands, wastes are pretreated to varying degrees by screening, settling basins, oxygenation, etc.; these measures tend to emphasize physical removal of solids more so than redoxic or translocational removal of chemical constituents (Gale et al. 1993, Hammer and Knight 1994, Knight et al. 2000). Wetlands receiving well-treated waste water tend to have a higher potential for removal of biochemicals due primarily to lower concentrations of those chemicals in the influent waste water. Conversely, 6 wetlands receiving untreated or poorly treated wastewater tend to have lower potential for biochemical removal due to high concentrations of those chemicals in the influent water (Hammer and Knight 1994, Knight et al. 2000, Knight 2003), despite potentially higher total mass removal by the wetland. High concentrations of waste water contaminants are problematic due to phytotoxic effects. Ammonium (NH 4 + ) in untreated agricultural waste is known to be phytotoxic, although different species exhibit toxic effects at different concentrations of NH 4 + , though few tested concentrations approached observed ATW ammonia concentrations (Walker and Evans 1978, van der Eerden 1982, Wang 1990, Surrency 1993, Dijk and Eck 1995, Magalhaes et al. 1995, Tanner et al. 1995b, Clarke 1999, Clarke and Baldwin 2002). Surrency (1993) reports Typha latifolia L. growth was suppressed by ammonia concentrations between 160 and 170 mgL -1 . Hill et al. (1997) found that the growth of five wetland species, including both Typha latifolia L. and Juncus roemerianus L., was not affected by field ammonia concentrations of up to 82.4 mgL -1 . Humenik et al. (1999) found Juncus effusus tolerant of ammonia concentrations up to 350 mgL -1 over a two year period. Clarke (1999, Clarke and Baldwin 2002) reported growth inhibition of five species, including Typha latifolia and Juncus effusus, at NH 4 + concentrations in excess of 200 mgL -1 , and discussed build-up of ammonia and time-dependant toxicity levels under continuous-loading treatment conditions. The exact mechanisms behind ammonium toxicity are not known; hypothetical links exists between NH 4 + toxicity and alterations in differential ion selective uptake (such are the cases for other cation toxicities, such as Ca +2 , Mg +3 , Li + , etc.(Hageman 1984, Britto and Kronzucker 2002). Ammonium ion toxicity may be mediated by 7 incorporation of low concentrations of selectively transported cations such as potassium, Rb + , or Cs + , but not by Ca +2 (Cao et al. 1993), but deficiencies in tissue K + at high ammonium levels have been reported as a potential cause for low growth rates and poor biomass production (Abbes et al. 1995). Ammonia fertilization may also cause soil acidification due in increased H + ion concentration from uptake of ammonia; acidification further complicates uptake mechanisms and may result in biomass decline due to inability to take up other nutrients (Hageman 1984, Dijk and Eck 1995). Toxicity appears to function in some species at the enzyme or mRNA levels rather than by reducing pH or cation uptake. In mustard (Sinapsis alba L.), interference with the transformation of fats to carbohydrates via reduction of RuBPCase, NAD-GDP and NADP-GDP enzymes was found to be the cause of decreased biomass production; pH changes and osmotic effects were found to not interfere with biomass production (Mehrer and Mohr 1989). Ammonia toxicity has been implicated in reduced hormonally-mediated transporter functions (Glass et al. 2002), and auxin-deficient mutants in Arabidopsis display some resistance to ammonia toxicity (Cao et al. 1993). Developments in the field of plant microbiology and discovery of the genetic mechanisms controlling ammonia uptake and its relation to the uptake of other cations or maintenance of cellular pH or cation equilibrium, while outside the scope of this paper, may further elucidate toxic properties of excessive amounts of ammonia (Kronzucker et al. 2001, Glass et al. 2002). In any event, high ammonium levels are one of the factors responsible for decreases in plant biomass in ATWs (Surrency 1993, Hill et al. 1997, Clarke 1999), and macrophyte biomass in ATWs is essential to system health and proper functioning (Kadlec and Knight 1996, Brix 1997, Hunter et al. 2000, Clarke and Baldwin 2002). 8 On the other hand, a moderate concentration of ammonia, where ?moderate? is again defined by the selected plant species, is likewise necessary for optimal system health and functioning. Due to adaptations to flooded soil and the reduced condition found therein (Mitsch and Gosselink 1993), most wetland plants preferentially uptake nitrogen in the ammoniacal form (Hammer and Knight 1994). Plant growth in natural wetland systems is often limited by a lack of abundance of reduced nitrogen in the system; ammonia concentrations below 2 mgL -1 are typical for a number of types of natural wetland (Shaver and Chapin 1995, Kadlec and Knight 1996, Lammerts et al. 1999, Van Duren and Pegtel 2000), with variations in NH 4 + concentration inversely proportional to standing biomass. Phosphate concentrations (PO 4 -3 ) in ATWs are not as high as NH 4 + levels; toxicity of phosphates has not been ascertained (Kadlec and Knight 1996), and researchers have failed to find significant growth inhibition at high phosphate levels (Romero et al 1999). The mechanism by which phosphate uptake occurs is a series of proton pumps in the root epidermis with high H 2 PO 4 - affinity, indicating the necessity of available H + ions to oxidize PO 4 -3 (Smith 2002, Smith et al. 2003). Generally, phosphates tend not to be as mobile in the environment as nitrogenous compounds; in the presence of cations they tend to form precipitates, in the presence of clays they tend to adsorb (Mitsch and Gosselink 1993). Having been removed from solution, these anions are not prone to removal by flow-through, and can accumulate at the redoxic boundary, maintaining availability for plants equipped with ATPase-linked membrane proteins with sufficient affinity for phosphates (Smith 2002). 9 The use of nitrogen to phosphorus ratios in plant tissues for predicting fertilization effects is controversial. Koerselman and Meuleman (1996) reviewed 40 field fertilization experiments in wetlands and concluded that biomass production of the vegetation community is typically limited by nitrogen if the aboveground N:P ratio is low (<14) and by phosphorus if the N:P ratio is high (>16). This result did not coincide well with the by-weight N:P ratio of 7.2:1 proposed by Redfield (1958), nearly 40 years earlier (although Redfield?s prediction was based upon N:P ratios found in marine phytoplankton). They also hypothesized that individual plant population response to fertilization could be predicted using the same ratios due to the fact that over half of the sites included in the review were dominated by a single species making up 80% or more of the aboveground biomass for the site. Tilman (1982, 1985, 1987, 1997, et al. 1999), modeling nutrient limitations in wet meadow and prairie communities, described competition effects in response to fertilization in the ?resource ratio model.? This model predicts that if two nutrients (for example N and P) are limiting, fertilization with one nutrient will cause the other to become relatively scarcer, and will create differential selection or competitive pressure for those species that are strong competitors for the scarce nutrient. Fertilization with the opposite nutrient will create the opposite effect, and will select for a different group of species. Furthermore, the model predicts the competitive ability of plant species for certain resources based on the tissue concentration of a nutrient: strong competitors will have a low concentration, weak competitors a high concentration. Thus, if N and P are limiting, the tissue N:P ratio should indicate which species will be promoted by N enrichment (a high N:P ratio) or by P enrichment (a low N:P ratio). Likewise the 10 addition or subtraction of a necessary limiting factor can enhance or diminish the presence of a certain species (Tilman et al. 1999). These results were corroborated by several authors (Mamolos et al. 1995, Mamolos and Veresoglou 2000). Other authors have investigated whether N:P ratios suggest nutrient limitation patterns at all, although in studies of shorter duration. Wassen et al. (1995), Pegtel et al. (1996), and Gusewell et al. (2003) found that N:P ratios could indicate accumulation of the non-limiting nutrient rather than paucity of the limiting one. McJannet et al. (1995) found different N:P ratios in 41 species of wetland plants grown under the same nutrient concentrations. Lechowicz and Shaver (1982) suggested different N:P ratios reflected differences in growth habit. Gusewell et al. (2003) indicated biomass N:P ratios in plant communities are not the result of species-specific or growth habit-specific uptake rates but the result of differences in the relative availability of N and P. A number of authors (Braakhekke and Hooftman 1999, Roem and Berendse 2000, Gusewell et al. 2002) found that N and P ratios tended to differ more within species between sites or treatments than within species within a given site or treatment, and that variations in N:P ratios within species tended to correlate well with variations in N:P ratios between sites. Ammonium toxicity tends to differ between species (see above; see Tanner et al 1995 for an excellent synopsis). Provided that N and P usage is not significantly different between species as suggested above, use of species specially selected for ammonium concentration tolerance may be beneficial to increasing treatment efficiency. However, since different species exhibit different gas flux constants, highest treatment efficiency may be achieved through a mix of species with graded ammonia tolerances and additive culm oxygen transmittal. It has been suggested that highest biomass production and 11 stability of ecosystem processes occur in wetlands with a high degree of species richness (Tilman 1996, Engelhardt et al. 2001). Mixtures of species, then, may yield elevated treatment efficiency over monocultures due to a suite of morphologies and resource acquisition mechanisms more able to fully capture available light and nutrients. At the same time, increases in fertilization tend to favor the competitor with the best acquisition mechanisms for that nutrient (Tilman et al. 1999), but at toxic or semi-toxic nutrient concentrations, inhibitory effects may prove significant (Goldberg and Barton 1992, Mittelbach et al. 2001). Taken together, the above suggest the possibility that N:P ratios can be used to predict biomass production, although the mechanisms by which this may be done are complex and often related to the species used, the community observed, or the time scale over which the experiment is run. In all, it appears that plant tissue concentrations may closely reflect concentrations of N and P in the surrounding environment; changes in N and P concentrations by fertilization may be reflected proportionally in the biomass, regardless of species-specific uptake mechanisms, or time duration. The predictive power of the N:P ratio has not been applied to agricultural wastewater treatment wetlands; the power and resolution of the tool may change when applied to systems of artificially elevated N and P levels. ATWs represent a unique case of circumstances for the application of a concept prior to this point used to determine end effects, given that treatment wetlands or other means by which to reduce nitrogenous output to the natural environment do not always function as planned with respect to nitrogen removal. Alterations in the environmental N:P ratio based on attempts to correct the vegetation N:P ratio may serve to increase treatment efficiency by removing nutrient limitations and 12 allowing higher biomass production and higher tissue nutrient concentrations (Powell 1974, Rubio et al. 1997, Tu and Ma 2003, Vojtiskova et al. 2004). Environmental N:P ratios in ATWs indicate P-limitation (Clarke 1999, DeRico 2000). Likewise, lack of treatment efficiency, in terms of use by plants or bacteria, in removing nitrogen when compared to phosphorus would seem to indicate an overabundance of nitrogen (Cronk 1996, Knight et al. 2002, Tanner and Kadlec 2003). Leibig?s Law of the Minimum states that biomass production may increase upon the removal of a limiting resource; in this case the limiting resource may be phosphorus. By increasing the available P, N usage in plant tissue may increase as well; increased use, in addition to conversion to N 2 or N 2 O gas along the anoxic boundary, may result in lower cation concentration in solution, which may potentially alleviate ammonium toxicity issues. Additionally, use of a species assemblage over a gradient of specific ammonium toxicities and oxygen flux constants may result in more efficient use of nitrogen, likewise lowering the solution cation concentration and potentially alleviating ammonium toxicity. In this thesis, I hypothesized that by altering N:P ratios in ATWs to better reflect naturally occurring N:P ratios, the phytotoxic level of NH 4 + can be increased to exclude ammonium concentrations in high nitrogen ATWs. The following objectives were addressed in two experiments: (1) Are environmental N:P ratios reflected in plant biomass under high concentrations of NH 4 + and PO 4 -3 ? (2) Is high productivity maintained in high [NH 4 + ] given appropriate [PO 4 -3 ]? (3) How do responses of species grown in mixtures change over a gradient of ammonia and phosphate concentrations? 13 RESEARCH EXPERIMENTS Planting guidelines for agricultural treatment wetlands available from the USDA (SCS 1991) advocate the use of polycultures as opposed to monocultures. Species were selected from these guidelines and based on the work of Clarke (1999 and Clarke and Baldwin 2002), Humenik et al. (1999), Hill et al. (1997), Surrency (1993), in nitrogen tolerances of wetland species. Typha latifolia L. and Juncus effusus L. (hereafter Typha and Juncus) were selected based on nitrogen tolerances; both species have high tolerances for NH 3 when grown in monoculture (Surrency 1993, Humenik et al. 1999, Clarke and Baldwin 2002). Typha is typically considered an excellent competitor, and monotypic stands are often associated with nutrient-rich conditions (Grace and Wetzel 1981, Tilman 1987, Surrency 1993, Clarke 1999, Svengsouk and Mitsch 2000). Juncus is not considered as vigorous a competitor, and ranked lower than Typha in competitive performance when measured against a common phytometer (Keddy et al. 2000). However, both species are clonal graminoids, and as such both should respond in similar fashion to additions of N and P in that additional culm growth should be the result of removal of P-limitation (Tilman 1982, DiTommaso and Aarssen 1989, Aerts et al. 1990, DeKroon and Bobbink 1997, Aerts et al. 1999). Mesocosms were constructed at the greenhouses of University of Maryland (College Park, Maryland, USA) in May 2002. These greenhouses controlled precipitation inputs into the mesocosms, allowing for permanent and maintainable concentrations of solutes. Typha and Juncus were purchased from Environmental Concern, Inc. (St. Michaels, Maryland, USA) in quart pots. Two plants of each species were planted in sterile soil-less media (Metromix; Scotts, Marysville, Ohio, USA) into 6L 14 pots. These pots were then placed in 10L pots lined with 4 mil polypropylene and filled with 6L of water, creating a pot-within-pot design. Nitrogen and phosphorus fertilizers were added as reagent-grade NH 4 Cl and Na 3 (PO 4 ) 2 . In two experiments, mesocosms arranged in this manner were subjected to two experiments to determine the effects of nitrogen and phosphorus concentrations typical in ATWs on biomass, species dominance, and the N:P ratio. The first experiment, in 2002, explored the effects of NH 4 + alone of the two species planted in polyculture. Five ammonia levels were selected using observed ammonia tolerances from Clarke (1999) and Humenik et al. (1999) for monoculture ammonia-N tolerances for the two species: 0, 75, 150, 225, and 300 mgL -1 . Fertilizer solution was replaced weekly and monitored to assure appropriate treatment concentrations, and above- and below-ground biomass was harvested at the end of eleven weeks. Ammonia effects on total biomass, above- and below-ground biomass, and stress tolerance were evaluated using analysis of variance (ANOVA) tools in SAS packages (SAS Institute, Cary, North Carolina, USA). The second experiment, in 2003, explored the effects of NH 4 + and PO 4 -3 on polycultures of Juncus and Typha. Four ammonia levels were selected from the previous year?s data: 75, 150, 225, and 300 mgL -1 , and phosphate was added such that solution N:P ratios of 5:1, 15:1, and 25:1 resulted. Fertilizer solutions were replaced weekly and monitored to assure appropriate treatment concentrations, and above- and below-ground biomass was harvested at the end of ten weeks. Effects of varying N:P ratios and ammonia on total biomass, above- and below-ground biomass, and stress tolerance were evaluated using analysis of variance (ANOVA) tools in SAS packages (SAS Institute, Cary, North Carolina, USA). 15 In attempting to answer the objectives listed above, the following assessments were made after biomass harvesting during the two experiments described above. To test whether tissue N:P ratios reflect environmental N:P ratios under high [NH 3 ] (objective 1), total tissue nitrogen and phosphorus concentrations for all tissues at each treatment, and the mean tissue nitrogen and phosphorus concentrations of Juncus and Typha were compared to the environmental N:P ratio as determined by solute addition. Comparisons were made using blocked ANOVA and data was adjusted for normality and homogeneity of variance. To test whether high productivity is maintained in phytotoxic [NH 3 ] given phosphate additions (objective 2), mean biomass of Typha and Juncus was compared between N treatment levels at equal N:P ratios with data adjusted as required to assure normality and homogeneity of residuals. Phytotoxic concentrations were assumed given the work of previous researchers (Clarke 1999, Humenik et al. 1999). Comparisons between N:P ratios within N treatments were also made for mean total, above-ground, and below-ground biomass for Juncus and Typha. To test differences in phytotoxicity in poly- and mono-culture (objective 3), mean total biomass for Juncus and Typha planted in polyculture (2002 data) was compared within species between treatments using blocked ANOVA with data adjusted as necessary. 16 CHAPTER TWO: EFFECT OF AMMONIA ON BIOMASS IN MESOCOSMS OF TYPHA LATIFOLIA AND JUNCUS EFFUSUS OBJECTIVE The objective of this study was to investigate differences in productivity responses of mesocosms of mixtures of Juncus effusus L. and Typha latifolia L. to ammonia under controlled conditions. This was accomplished in a one greenhouse experiment, during which plants were exposed to levels of 0, 75, 150, 225, and 300 mg NH 4 + L -1 over an 11 week period. In order to supply micronutrients without compromising experimental control, plants were cultured in sterilized soil-less media (Metromix, Scotts, Marysville, Ohio, USA), and treatments were applied by batch- loading with appropriate concentrations (Wang 1991). Typha latifolia (hereafter Typha) and Juncus effusus (hereafter Juncus) were selected for this study based on the work of Clarke (1999), Clarke and Baldwin (2002), Humenik et al. (1999), and Surrency (1993), which indicated high ammonia-N tolerances for these species. Using Juncus and Typha allowed us to investigate changes in the proportion of biomass allocated to each species over the given range of treatments, as well as to investigate changes in the mesocosm biomass on the whole. METHODS Preparation of Mesocosms Plant material was purchased from Environmental Concern, Inc. (St. Michaels, Maryland) in May of 2002 as 125 quart pots of each species, grown from seed under 17 greenhouse conditions. Plants were transported from St. Michaels to the Harrison Laboratory Greenhouses at the University of Maryland, College Park (UMCP) in a covered vehicle, where they were immediately transplanted with their original soil and accompanying microorganisms into 50 mesocosms of approximately 6 L in volume and watered thoroughly to prevent potential transplant shock. Each ?mesocosm? was planted with two quart pots of each species, in a design minimizing North-South and East-West effects (Figure 2.1). Following 2 weeks of acclimatization to UMCP conditions, water, temperature, and transplant, the 6 L pots were then placed within 10 L pots lined with 4 mil. polypropylene and filled with 6 L of greenhouse tap water, creating a pot-within-pot design (Figure 2.1.) such that the water flooded each mesocosm (experimental unit) to the soil surface. These mesocosms were randomly placed in five columns of 10 rows, with columns aligned North-South (labeled A through E) and rows aligned East-West (labeled 1 through 10). Treatments of ammonia concentrations were assigned such that each treatment occurred once in each row and twice in each column, a block design to account for potential effects of shading due to rapid and dense growth of either species (Figure 2.1.). Treatments were applied such that ammonia concentrations of 0, 75, 150, 225, and 300 mgL -1 of NH 4 + were established (Table 2.1.). Ammonia treatments for the 2002 study were prepared using reagent-grade ammonium chloride (Fisher Scientific, Fair Lawn, New Jersey, USA) and greenhouse tap water. Tap water was used due to the large volumes of water changes necessary as a batch-loaded system. Batch loading was chosen over continuous flow based on Wang (1991) and Clarke and Baldwin (2002). Treatment solutions were prepared by the addition of appropriate mass of NH 4 Cl-N and greenhouse tap water to 6 L, and were 18 B. A Column B C D 75 75 75 75 75 75 75 75 75 75 0 0 0 0 0 0 0 0 0 0 225 225 225 225 225 225 225 225 225 225 150 150 150 150 150 150 150 150 150 150 300 300 300 300 300 300 300 300 300 300 Row Column A B C D E A. Row 5:15:1 5:1 5:1 5:15:1 5:1 5:1 5:15:1 5:1 5:1 15:115:1 15:1 15:1 15:115:1 15:1 15:1 15:115:1 15:1 15:1 25:125:1 25:1 25:125:1 25:1 25:125:1 25:1 25:1 25:1 25:11 1 2 2 3 3 4 4 5 5 6 6 7 7 8 8 9 9 75150 Treatment (mgL -1 ) 10 Outer Pot L) Line Inner Pot (6 L) Wate 45 cm C. 45 cm Figure 2.1. Aspects pertaining to greenhouse treatment design (North is towards the top of the page): (A.) indicates ammonia concentration (mgL and column squares used in 2002; (B.) indicates ammonia concentration (mgL -1 ) treatm C.) indicates pot-within pot design used for mesocosm setup in both years. ent design with N:P ratio inclusion used in 2003; ( (10 r r Level 300 225 -1 ) treatment design with row 19 Table 2.1. Summary of nutrient additions to prepare treatment solutions in 6 L increments in 2002, and the associated total mass of ammonium chloride and nitrogen for each treatment. [Ammonia] Treatment NH 4 Cl (g) N (g) 00 75 1.41 0.37 150 2.83 0.74 225 4.24 1.11 300 5.65 1.48 20 replaced once per week. Solution replacements consisted of draining previously existing solution and replenishment with freshly mixed solution of appropriate concentration. Treatment Solution, Plant Growth, and Statistical Analyses Interstitial water was sampled weekly in each experimental unit during solution replacement to insure appropriate ammonium concentration. Due to the number and frequency of measurements and the corresponding frequency of recalibration necessary for such, it was deemed impractical to attempt the use of [NH 3 ] electronic meters. The Orion Ammonia Electrode (Model 95-12) warrants a suggested recalibration after each measurement, requires five to ten minutes per measurement, and suggests measurements not be taken in situ (Thermo Corp., Waltham, Massachusetts, USA). Conductivity meters are robust against need for recalibration, and serve as a useful surrogate for measurements of cation concentration. As ammonium was the only cation added to solution, changes in conductivities were affected primarily by this cation (Clarke 1999). Weekly variations in the initial greenhouse tap water cation-derived conductivity were noted. Conductivities were expected to deviate from desired treatment levels due to evapotranspiration, uptake, nitrification, volatilization, and accumulation on cation- exchange sites in the soil solution. Conductivities for each treatment level in 2002 clearly increased with increasing ammonium concentration, may be found in Figure 2.2; overall greenhouse tap water conductivity averaged 0.7?0.3 microsiemens. Average above-and below-ground biomass of each species was determined for non-experimental plants prior to application of treatments. In 2002, mean above- and 21 0 mg/L 75 mg/L 150 mg/L 225 mg/L 300 mg/L 0 500 1000 1500 2000 2500 3000 3500 4000 6/6/02 6/20/02 7/4/02 7/18/02 8/1/02 8/15/02 8/29/02 9/12/02 9/26/02 Sampling Date Cond uct i v it y ( ? S) 0 mg/L 75 mg/L 150 mg/L 225 mg/L 300 mg/L Figure 2.2. Interstitial water conductivities (?S) for each mesocosm at sampling dates in 2002 at experimental ammonia concentrations. 22 below-ground biomass per Typha plant (n=25) were 3.2 ?0.3g and 5.3? 0.2g, respectively (reported as mean ? SE). Mean above- and below-ground biomass per Juncus plant (n=25) were 4.3?0.9g and 6.4?0.2g, respectively. Mesocosms were destructively harvested in mid-September after eleven weeks of treatment application. Above-ground material for each species was removed to soil surface; below-ground material was retrieved by removing the root ball from each mesocosm, removing particles of the planting medium, and carefully separating Juncus and Typha roots, which were clearly discernible based on size and color. Thus, each mesocosm (experimental unit) was separated into four distinct sampling units. Biomass was dried at 27 degrees Centigrade and weighed until stabilization of mass (approximately three days for above-ground and five days for below-ground samples). Analyses of variance (ANOVAs) were conducted for above- and below-ground biomass and total biomass for each mesocosm and within and between each species. ANOVAs were conducted using SAS Version 8.2 (SAS Institute, Cary, North Carolina, USA). In each case, biomass was the dependant variable; effects of species, column placement (blocking effect), and effects of ammonia treatments were assessed as independent variables. Blocking effect interactions were not analyzed due to significance of first-order blocking effects; significance of first order terms tends to reduce resolution of second order terms. Analyses of variance were likewise conducted for root-to-shoot ratios for Typha and Juncus, both between the species, and within the species at each ammonia concentration. Normality and homogeneity of variances were examined in each case, and log transformations of dependant data were made when necessary. Significant differences were determined at ?=0.05. 23 Because experiment-wise error rates would be elevated by using paired t-test to investigate treatment means and mean root-to-shoot ratios, pair-wise comparisons using ANOVA were selected to allow for a more robust analysis of the interactive effects of the independent variables. Additionally, multiple sampling units were assigned to each experimental unit, and ANOVA minimizes the effect of any potential oversampling complications. RESULTS AND DISCUSSION Ammonia concentration significantly affected total, above- and below-ground biomass for each mesocosm at the 5% level in 2002 (Table 2.2). In the case of total biomass for each mesocosm, a definite fertilization effect occurred between the 0 mgL -1 and the 75 mgL -1 ammonia treatments. The average mesocosm biomass for these treatments was 229.4?25.8 g and 508.2?30.1 g, respectively. A statistically apparent toxicity effect occurred in the 225 mgL -1 and the 300mgL -1 ammonia treatment levels (Figure 2.3). Toxicity effects may be seen as an overall decrease in the amount of biomass found at these higher treatment levels. This overall trend may be found in mesocosm above-ground biomass (Figure 2.4), although statistically significant toxicity effect begins at the 225 mgL -1 treatment level. This trend may be found as well in mesocosm below-ground biomass (Figure 2.5). Ammonia concentration likewise significantly affected total, above- and below- ground biomass production of Typha at the 5% level in all years (Table 2.3). Total Typha biomass exhibited a statistically evident fertilization trend in the 75 mgL -1 and 150 mgL -1 ammonia treatments (Figure 2.6). Mean biomass production at these levels was 24 Table 2.2. Tests of fixed effects on total, above-ground, and below-ground biomass of mesocosms of Typha latifolia and Juncus effusus in 2002. Please refer to appendix I for SAS code. Measured Biomass Effect d.f. F Value P Level Total Mesocosm [Ammonia] 4 16.2 p<0.0001 Total Mesocosm Species 1 98.9 p<0.0001 Total Mesocosm [Ammonia] x Species 4 6.9 p=0.0002 Total Mesocosm Column 4 5.1 p=0.0023 Above-ground Mesocosm [Ammonia] 4 23.2 p<0.0001 Above-ground Mesocosm Species 1 145.7 p<0.0001 Above-ground Mesocosm [Ammonia] x Species 4 7.59 p<0.0001 Above-ground Mesocosm Column 4 2.1 p=0.0775 Below-ground Mesocosm [Ammonia] 4 5.4 p=0.0006 Below-ground Mesocosm Species 1 5.5 p=0.0001 Below-ground Mesocosm [Ammonia] x Species 4 16.3 p=0.0066 Below-ground Mesocosm Column 4 3.82 p=0.0005 25 0 100 200 300 400 500 600 0 75 150 225 300 Ammonia Concentration (mgL -1 ) T o t a l M e so c o s m B i o m a ss ( g ) a b b c d Figure 2.3. Total mesocosm biomass along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 26 0 100 200 300 400 500 600 0 75 150 225 300 Ammonia Concentration (mgL -1 ) A b ov e- gr o u n d Mesoc o sm Bi o m ass ( g ) a b b b c Figure 2.4. Total above-ground mesocosm biomass along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 27 0 100 200 300 400 500 600 0 75 150 225 300 Ammonia Concentration (mgL -1 ) B e low- gr ou nd M e s o cosm Bio m as s ( g ) a a a b b Figure 2.5. Total below-ground mesocosm biomass along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 28 Table 2.3. Tests of fixed effects on total, above-ground, and below-ground biomass of Typha latifolia in 2002. Please refer to appendix I for SAS code. Measured Biomass Effect d.f. F Value P Level Total Typha latifolia [Ammonia] 4 19.2 p<0.0001 Total Typha latifolia Column 4 2.2 p=0.0859 Above-ground Typha latifolia [Ammonia] 4 14.8 p<0.0001 Above-ground Typha latifolia Column 4 2.3 p=0.0726 Below-ground Typha latifolia [Ammonia] 4 10.71 p<0.0001 Below-ground Typha latifolia Column 4 1.8 p=0.1420 29 0 50 100 150 200 250 300 350 400 0 75 150 225 300 Ammonia Concentration (mgL -1 ) T o ta l Ty ph a Bio m ass ( g ) ad d a b c Figure 2.6. Total Typha latifolia biomass along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 30 172.2?33.9 g and 222.9?13.3 g respectively, which is elevated from the average 80.9?13.7 g at 0 mgL -1 and the toxicity effect displayed at 300mgL -1 ammonia, with an average of 48.6?10.7 g. Fertilization effects at between 75 mgL -1 and 150 mgL -1 may be seen in above-ground biomass production (Figure 2.7) and below-ground biomass production (Figure 2.8) in Typha as well. The observable trend in total Typha biomass production is similar to the effect produced in overall mesocosm biomass. Juncus effusus total biomass production and above-ground biomass production were significantly affected by ammonia treatment at the 5% level (Table 2.4). Fertilization of Juncus occurred at the 75 mgL -1 ammonia treatment, where an average of 336.1?33.8 g of dry tissue was produced over the course of the experiment; at 0 mgL -1 ammonia, insufficient nitrogen was available, demonstrated by significant chlorosis at this treatment level, and by the low average biomass of 148.6?14.5 g. Above 75 mgL -1 , biomass production was reduced potentially due to ammonia toxicity (Figure 2.9), although differences between biomass production between 75 mgL -1 and 225 mgL -1 and 300 mgL -1 ammonia are not statistically significant; this indicates that, over an ammonia gradient of 150-300 mgL -1 , differential effects on growth are non-significant. The trend in above-ground biomass production is similar, with an observable biomass peak existing at the 75 mgL -1 treatment level (Figure 2.10). However, below-ground biomass production in Juncus was apparently not affected by ammonia concentration (Figure 2.11). This statistically non-significant effect could demonstrate a potential immunity to ammonia toxicity in the roots of Juncus effusus. Juncus root tissue tends to be denser than Typha root tissue, which could allow it to be more tolerant of variable ammonia concentrations. The large ammonia molecule may have difficulty diffusing through this 31 0 50 100 150 200 250 300 350 400 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Ab ov e-grou nd Ty ph a B i o m a ss ( g ) a a b c b Figure 2.7. Above-ground Typha latifolia biomass along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 32 0 50 100 150 200 250 300 350 400 0 75 150 225 300 Ammonia Concentration (mgL -1 ) B e lo w-grou nd Ty p h a Bi o m as s ( g ) a b ac c b Figure 2.8. Below-ground Typha latifolia biomass along a gradient of ammonia concentration treatments in 2002. . Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 33 Table 2.4. Tests of fixed effects on total, above-ground, and below-ground biomass of Juncus effusus in 2002. Please refer to appendix I for SAS code. Measured Biomass Effect d.f. F Value P Level Total Juncus effusus [Ammonia] 4 10.08 p<0.0001 Total Juncus effusus Column 4 4.16 p=0.0146 Above-ground Juncus effusus [Ammonia] 4 14.6 p<0.0001 Above-ground Juncus effusus Column 4 0.8 p=0.5651 Below-ground Juncus effusus [Ammonia] 4 0.93 p=0.4578 Below-ground Juncus effusus Column 4 6.5 p=0.0004 34 0 50 100 150 200 250 300 350 400 0 75 150 225 300 Ammonia Concentration (mgL -1 ) To t a l J u nc us Bio m as s ( g ) a b bc bc c Figure 2.9. Total Juncus effusus biomass along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 35 0 50 100 150 200 250 300 350 400 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Ab ov e-grou nd J u nc us Bi o m ass ( g ) a b bc c c Figure 2.10. Above-ground Juncus effusus biomass along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 36 0 50 100 150 200 250 300 350 400 0 75 150 225 300 Ammonia Concentration (mgL -1 ) B e low- gr ou nd Ju nc us B i o m a ss ( g ) a a a a a Figure 2.11. Below-ground Juncus effusus biomass along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 37 dense root tissue. Also of note is a consistently higher observable overall biomass production and especially overall below-ground biomass, when compared to Typha over all treatment conditions. These results indicate that Juncus is relatively robust under high ammonia concentrations, and optimal growth is maintained even at high ammonia concentrations, eclipsing previous growth inhibitory effects noted at 110 mgL -1 (Clarke 1999, Clarke and Baldwin 2002), and agreeing with Humenik et al. (1999) . Between species total biomass differed significantly at all ammonia concentrations except in the below-ground biomass at low concentrations of ammonia (Table 2.5). Figure 2.12 illustrates significant differences between species within treatments for total biomass. Typha latifolia exhibited impaired growth at higher treatment levels, to a low of 48.6?10.5 g at 300 mgL -1 , with a peak of 222.8?14.3 g at 150 mgL -1 . At this peak alone is there no statistically significant difference between Typha biomass and Juncus biomass. Juncus exhibits peak growth at 75 mgL -1 , and diminishes somewhat at higher concentrations. Compared with Typha, Juncus appears to have higher ammonia tolerance. Figure 2.13 illustrates significant differences in Juncus and Typha in above-ground biomass, with a peak of Typha production at 150mgL -1 , and minimums at 0 and 300 mgL -1 , and a peak of Juncus production at 75 mgL -1 , but without excessive decline thereafter. Below-ground biomass for the species follows the same general observable trend, with significant differences occurring in the 225 mgL -1 to 300 mgL -1 ammonia concentration levels (Figure 2.14). Negative response of biomass production to ammonia levels of 300 mgL -1 was statistically evident for all measured biomass, with decreasing biomass trends of varying statistical magnitude for all ammonia concentrations above 75 mgL -1 . These trends, with 38 Table 2.5. Summary of statistical analyses of differences in biomass production between Typha latifolia and Juncus effusus over an ammonia gradient for 2002. At each ammonia treatment, the results of pairwise comparisons of treatment means of Juncus and Typha are presented. Please refer to Table 2.2 for analysis of the effect of species on biomass production on the whole. Please refer to appendix I for SAS code. Measured Biomass [Ammonia] Treatment d.f. T Value P Level Total Mesocosm 0 185 2.2 p=0.0307 Total Mesocosm 75 185 5.7 p<0.0001 Total Mesocosm 150 185 1.1 p=0.2575 Total Mesocosm 225 185 5.3 p<0.0001 Total Mesocosm 300 185 7.7 p<0.0001 Above-ground Mesocosm 0 86 2.9 p=0.0053 Above-ground Mesocosm 75 86 8.2 p<0.0001 Above-ground Mesocosm 150 86 2.5 p=0.0141 Above-ground Mesocosm 225 86 5.8 p<0.0001 Above-ground Mesocosm 300 86 8.0 p<0.0001 Below-ground Mesocosm 0 86 1.0 p=0.3401 Below-ground Mesocosm 75 86 1.1 p=0.2763 Below-ground Mesocosm 150 86 0.5 p=0.6107 Below-ground Mesocosm 225 86 3.0 p=0.0036 Below-ground Mesocosm 300 86 4.5 p<0.0001 39 0 100 200 300 400 500 600 0 75 150 225 300 Ammonia Concentration (mgL -1 ) T o t a l S p eci es Bi om a s s ( g ) Typha Juncus * * * * Figure 2.12. Total species biomass along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Paired means with an asterisk indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 40 0 100 200 300 400 500 600 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Ab ov e- gr o u n d Sp eci es Bi om ass ( g ) Typha Juncus * * * ** Figure 2.13. Above-ground species biomass along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Paired means with an asterisk indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 41 0 100 200 300 400 500 600 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Bel o w- g r ou nd Sp eci es B i o m ass ( g ) Typha Juncus ** Figure 2.14. Below-ground species biomass along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Paired means with an asterisk indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 42 some fertilization effect occurring at moderate ammonia-N levels followed by toxic response at higher levels, are in agreement with studies investigating responses of species to varying ammonia levels with similar species (Surrency 1997, Hill et al. 1997, Humenik et al. 1999, Clarke and Baldwin 2002). Overall, Typha appears to grow optimally in the range of 75-100 mgL -1 , which corresponds well with the work of previous researchers (Hill et al. 1997, Clarke and Baldwin 2002). Other authors have noted stresses for Typha at levels approaching 170 mgL -1 (Surrency 1993), or a fertilization effect up to 83 mgL -1 (Hill et al. 1997) and 100 mgL -1 (Clarke and Baldwin 2002). Authors have noted tolerances of up to 110 mgL -1 for Juncus (Clarke 1999), but others note inhibition at relatively low ammonia levels (20.8 mgL -1 ; Hill et al. 1997). The tolerance of Juncus tissue, especially of below-ground tissue, may confirm that Juncus is a ?stress tolerator? moreso than Typha, and could therefore be considered dominant under toxic regimes (Emery et al. 2001). Analysis of the root-to-shoot (R:S) ratio is important in considering the role of cytokinins in plant growth. Cytokinin depletion in the rhizosphere corresponds to decreased cytokinin transport to the leaves; low leaf cytokinin levels inhibit cell division (Walch-Liu et al. 2000, Taiz and Zeiger 2002). The role of increased levels of ammonia in decreasing shoot growth may be due to disruption of cytokinin transport and cytokinin synthesis in root apical meristem (Taiz and Zeiger 2002, Vojtiskova et al. 2004); this results in a reduction of cell division and of leaf elongation (Walch-Liu et al. 2000). Root-to-shoot ratio was significantly affected by ammonia, species, and the ammonia by species interaction (Table 2.6). Overall decreases in R:S ratio in the mesocosm are evident in Figure 2.15, but may be found between the 0 mgL -1 and the 75 mgL -1 ammonia 43 Table 2.6. Test of fixed effects on the root-to-shoot ratio in mesocosms and in Typha latifolia and Juncus effusus in 2002. Please refer to appendix I for SAS code. Measured Root-to-Shoot Ratio Effect d.f. F Value P Level Total Mesocosm [Ammonia] 4 11.3 p<0.0001 Total Mesocosm Species 1 17.8 p<0.0001 Total Mesocosm [Ammonia] x Species 4 5.3 p=0.0008 Total Mesocosm Column 4 1.8 p=0.1268 Total Typha latifolia [Ammonia] 4 8.7 p<0.0001 Total Typha latifolia Column 4 2.1 p=0.0978 Total Juncus effusus [Ammonia] 4 12.1 p<0.0001 Total Juncus effusus Column 4 3.1 p=0.0271 44 0.0 1.0 2.0 3.0 4.0 5.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Mes o co s m Ro ot - t o- S h o o t Rat i o ( R g/ S g ) a b b b b Figure 2.15. Root-to-shoot ratios for mesocosms along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 45 treatments, where overall biomass fertilization effect is occurring. Both species show the same general effect, with differences in the R:S ratio evident only in the 0 mgL -1 treatment (Figure 2.18). This demonstrates similarity in the effects of ammonium toxicity on the two species. Decreases in the R:S ratio in Juncus and in Typha individually are statistically evident as well (Table 2.6, Figures 2.16 and 2.17), and indicate suppression of root formation, increased shoot production, or a combination of the two. Since Juncus below-ground biomass is not significantly affected by ammonia concentration, increased shoot production must have occurred to generate this statistically significant decrease in R:S ratio. Variation in the R:S ratio is due primarily to ammonia treatment and the species considered, but may also be due to the interaction of the terms, indicating some level of interaction between the species. Increases in ammonia concentration and subsequent reduction of the R:S ratio corresponds well with similar responses in other species (Walch-Liu et al. 2000, Vojtiskova et al. 2004), although direct measurement of R:S ratio inhibition by ammonia has not been noted in Juncus and Typha prior to this study. The significance of the blocking effect in this analysis was a demonstration of the effects of North-South positioning in the greenhouse (the reader is directed to Tables 2.2 through 2.6 for those instances where the blocking effect is statistically significant). Some variation in biomass production was caused as a result of the orientation of the mesocosms along the axis of the greenhouse, and it was therefore advisable to utilize this fixed effect in experimental design. It is of interest to note that the column effect, that is, the effect of the rising and setting of the sun on providing enough light for the plants, is more evident in Juncus, and especially in below-ground Juncus biomass, where it 46 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Ty p h a R o ot - t o- Sh oot R a ti o ( R g/ S g ) a b b b b Figure 2.16. Root-to-shoot ratio for Typha latifolia along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 47 0.0 0.5 1.0 1.5 2.0 2.5 3.0 3.5 4.0 4.5 5.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) J unc us R o ot - t o- S h oot R a ti o ( R g/ S g ) a b b b b Figure 2.17. Root-to-shoot ratio for Juncus effusus along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Means with different letters indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 48 0.0 1.0 2.0 3.0 4.0 5.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Sp ec ie s R o ot-t o-Sh o o t R a tio (R g/S g ) Typha Juncus * Figure 2.18. Species root-to-shoot ratios along a gradient of ammonia concentration treatments in 2002. Error bars indicate means ? S.E. Paired means with an asterisk indicate significant differences at the 5% level. Due to normality of data and homogeneity of residuals from the ANOVA, transformation was not necessary (see appendix I for SAS code). 49 accounted for more variation than did the ammonia treatment itself (Table 2.4). Juncus is a plant of shorter stature, and was shaded by the taller Typha to produce this effect. The effects of Juncus and Typha upon biomass production or either species are important to note as a result of this study. Although Tilman (1982) found Typha to be a better competitor under N-fertilization regimes, intrageneric competition along a phytotoxic nutrient gradient is a little known trend. The interactive effects of toxicity and competition are poorly documented. Most work to date has focused on field-scale studies along much lower nutrient gradients for fertilization effects typical of natural ecosystems (e.g. Tilman 1982, Tilman et al. 1999, Wilson and Tilman 1995, Mamolos and Veresoglou 2000). In this study, Juncus produced more total biomass than Typha for all ammonia concentrations in 2002 (Figure 2.12) In no case did Juncus produce significantly less biomass than Typha in any biomass measurement and at any ammonia treatment, despite similar initial biomass at the inception of this study. The ability of Juncus to tolerate stress better than Typha may occur as a result of plant morphology in a confined system. Pot size could lead to an inability to establish new culms, and although other studies have indicated that Typha can establish quickly in confined situations (Clarke 1999, Svensouk and Mitch 2000), none of these were conducted in a 10 L pot size. Typha latifolia tends to occur in relatively deep water conditions (Grace and Wetzel 1981); the presence of high amounts of aerenchymous tissue in the stems and the presence of numerous fine root hairs are likely to facilitate high rates of gas exchange through anoxic sediments (Brix et al. 1992). Additionally, the presence of a large rhizome in Typha indicates a high storage capacity for nutrients. Juncus, on the other hand, has aerenchymous tissue with significantly smaller gas 50 volume, and also fewer root hairs; it is typically found in clumps in oxygen-poor but not necessarily anoxic sediments. Juncus has a significantly smaller base culm, certainly without the starchy tuber of Typha latifolia, and therefore uses nutrients as they become available in lieu of storing them. Typha stores nutrients; this capacity may allow it to outcompete other species in times of nutrient stress, but will also force accumulation of excess nutrients in times of overabundance of nutrients. In this study, rhizomes of Typha in high ammonia concentrations tended to be smaller than those of lower concentrations, where they existed at all. Often, stunted, smaller rhizomes were found in higher nutrient situations, exhibiting much necrotic tissue and decay. This finding appears to be unique in the literature; for the most part, managers have focused on eliminating Typha in order to increase diversity in treatment systems (Kadlec and Knight 1996, Mitsch and Gosselink 2000) or developing planting schemes to enhance the competitive abilities of other species (Svengsouk and Mitsch 2000). Extremely elevated ammonia concentrations here helped to limit the dominance of Typha. CONCLUSIONS Above-ground, below-ground, and total biomass for the mesocosms were significantly affected by ammonia concentrations at the 5% level. Biomass measurements were maximized in the 75 to 150 mgL -1 ammonia concentration levels, and declined for higher concentrations of ammonia. Above-, below-, and total biomass of Typha latifolia were significantly affected by ammonia concentrations over a one-year period. Above-ground and total biomass of Juncus effusus were affected by ammonia concentrations at the as well, although in many cases Juncus grew at productive levels 51 even at higher ammonia concentrations. Decreases in cytokinin production due to elevated ammonia levels may have limited cell division, leaf elongation, and overall biomass production in these species are trends shown by a decrease in the root-to-shoot ratio. Evident trends in biomass production demonstrate a high stress tolerance ability of Juncus effusus in high ammonia concentrations typical of constructed wetlands for wastewater treatment; the species demonstrated significantly higher ammonia tolerance than previously reported. Juncus has often been overlooked by managers hoping to create a diverse wetland community (Kadlec and Knight 1996, Mitsch and Gosselink 2000); its performance here may call for a re-evaluation of its competitive dynamic. Observation of the effects of additional limiting or phytotoxic nutrient concentrations may further elaborate competitive performance differences between these two species and others, as differences in tissue ratios often indicate differences in uptake mechanisms and potential effects of limitation. Potential increased genetic knowledge of enzyme control of ammonia toxicity in, for example, Arabidopsis, may lead to better control of toxic effects (Kronzucker et al. 2001, Glass et al. 2002), but application to wetland species known to have specific enzyme adaptations to anoxia and its affect on cation uptake pathways (Mitsch and Gosselink 1993) may not occur for some time. Until the problem of ammonia toxicity can be better understood, pre-treatment of wastes prior to introduction into ATWs probably remains the best policy in maximizing biomass production and subsequent nitrogen removal. 52 CHAPTER THREE: THE VEGETATION N:P RATIO IN MESOCOSMS SUBJECTED TO HIGH AMMONIA CONCENTRATIONS OBJECTIVE The objective of this study was to investigate the role of the vegetation N:P ratio in illustrating relationships in Juncus effusus L. and Typha latifolia L. to mixtures of ammonia and phosphorus under controlled conditions. The specific relationships of interest were the overall biomass production by each species, the effect of the N:P ratio on the nitrogen uptake of each species, and the effect of the experimental N:P ratio on actual vegetation N:P ratio. These objectives were accomplished in a one-year greenhouse study during which plants were exposed to levels of 75, 150, 225 and 300 mgL -1 of ammonia amended with phosphate-P concentrations at N:P ratios of 5:1, 15:1 and 25:1 over a ten week period. Plants were cultured as in Chapter Two. Using Juncus and Typha following the rationale explained above allowed investigation of changes in N:P ratios, such that the N:P ratio may be used to determine species fitness over a range of treatment levels. Growth in these species was tested for limitation by N, P or co- limitation, and the utility of indications of growth limitation by the N:P ratio was examined. METHODS Preparation of Mesocosms Plant material was purchased, transported, and prepared as described in Chapter Two in May of 2003. After an initial acclimation period, each planted mesocosm was 53 placed in a lined 10 L pot (Figure 2.1). These mesocosms (experimental units) were placed in four columns aligned North-South (labeled A through D), and nine rows aligned East-West (labeled 1 through 9). Treatments of ammonia concentrations were applied such that each column received a single ammonia concentration for ease of treatment application (the same concentrations as used in the 2002 experiment, except the 0 mgL -1 treatment, Figure 2.1). Treatments of phosphorus were applied such that, in each row, each N:P concentration was applied three times (Figure 2.1). For this experimental design, each N:P ratio occurred in each ammonia concentration treatment three times. Nitrogen to phosphorus ratios were chosen based on Koerselman and Meuleman (1996) with categories of N-limited, co-limited, and P-limited at ratios of 5:1, 15:1 and 25:1, respectively. The ammonia concentration of 0 mgL -1 was removed as appropriate N:P ratios to a 0 mgL -1 ammonia treatment do not exist. Mass values of total nitrogen, ammonia chloride, sodium phosphate, total phosphate, and total phosphorus, and accompanying N:P ratios are found in Table 3.1. Ammonia treatments were prepared using the 2002 protocol. Treatments of N:P ratio were prepared by adding appropriate amounts of reagent grade Na 3 (PO 4 ) 2 (Fisher Scientific, Fair Lawn, New Jersey, USA) to ammonia treatments and greenhouse tap water to 6 L (see Table 3.1). Solutions were replaced once weekly following the above protocol for 2002 with the addition of the N:P ratio treatment described above. Treatment Solution, Vegetation N:P Ratio, and Statistical Analyses Interstitial water sampling was carried out using the methodology described above for experiments occurring in 2002. As in 2002, conductivities clearly increased with 54 Table 3.1. Summary of nutrient additions to prepare treatment solutions in 6 L increments in 2003, and the associated total mass of ammonium chloride, nitrogen, phosphate, and total phosphorus, and accompanying N:P ratios for each treatment. [Ammonia] Treatment NH 4 Cl (g) N (g) N:P Ratio Treatment P (g) Required NaH 2 PO 4 (g) [Phosphorus] 75 1.41 0.37 5 0.07 0.29 12.34 75 1.41 0.37 15 0.02 0.10 4.11 75 1.41 0.37 25 0.01 0.06 2.47 150 2.83 0.74 5 0.15 0.57 24.67 150 2.83 0.74 15 0.05 0.19 8.22 150 2.83 0.74 25 0.03 0.11 4.93 225 4.24 1.11 5 0.22 0.86 37.01 225 4.24 1.11 15 0.07 0.29 12.34 225 4.24 1.11 25 0.04 0.17 7.40 300 5.65 1.48 5 0.30 1.15 49.35 300 5.65 1.48 15 0.10 0.38 16.45 300 5.65 1.48 25 0.06 0.23 9.87 55 increasing ammonium addition (Figure 3.1A); additions of phosphorus did not significantly affect conductivity; slopes of the lines for conductivity of the three N:P ratio treatments are not significantly different (Figure 3.1B). Hydrogen ion concentration (pH) was measured twice monthly in the 2003 experiment to ensure that the addition of a weak diprotic acid (Na 3 (PO 4 ) 2 ) did not result in significant pH differences between treatments; phosphate concentration was measured twice monthly to identify and correct any deviations from the proposed ratios that may have occurred due to evaporation of solvent, adsorption of phosphate ions, or other forms of accumulation using the appropriate colorimetry protocol (Spectrophotometer Model DR/2500, Hach Co., Loveland, Colorado, USA). Average above- and below-ground biomass of each species was determined for non-experimental plants prior to application of treatments. In 2003, mean above- and below-ground biomass per Typha plant (n=28) were 2.3? 0.4g and 3.3? 0.3g, respectively (reported as mean ? SE). Mean above- and below-ground biomass per Juncus plant (n=28) were 3.9?0.2g and 6.4?0.6g, respectively. Mesocosms were destructively harvested in mid-September after ten weeks of treatment application. Above-ground material for each species was removed to soil surface; below-ground material was retrieved by removing the root ball from each mesocosm, removing particles of planting medium, and carefully separating Juncus and Typha roots, which were clearly discernible based on size and color. Thus, each mesocosm was separated into four distinct sampling units. Biomass was dried at 80 degrees Centigrade and weighed until stabilization of mass (approximately 3 days for above-ground and five days for below- ground samples). 56 A. 75 mg/L 150 mg/L 225 mg/L 300 mg/L 0 500 1000 1500 2000 2500 3000 3500 4000 6/6/2003 6/20/2003 7/4/2003 7/18/2003 8/1/2003 8/15/2003 8/29/2003 9/12/2003 9/26/2003 Sampling Date Conduct i vi t y ( ? S) 75 mg/L 150 mg/L 225 mg/L 300 mg/L B. 0 500 1000 1500 2000 2500 3000 3500 4000 6/6/2003 6/20/2003 7/4/2003 7/18/2003 8/1/2003 8/15/2003 8/29/2003 9/12/2003 9/26/2003 Sampling Date C o nd uc tivity ( ? S) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.1. Interstitial water conductivities (?S) for each mesocosm at sampling dates in 2003: (A.) indicates conductivities based on experimental ammonia concentrations; (B.) indicates conductivities based on experimental N:P ratios. 57 Tissue nitrogen and phosphorus concentrations for each species were measured by dry matter combustion and color spectroscopy. Tissue samples from above- and below- ground biomass of each species were removed from the biomass following measurement for the above studies. For nitrogen and phosphorus analysis, 0.05 grams were removed from the initial 5 g samples following grinding for homogenization. The use of only 0.05 grams in lieu of the typical 0.1 grams was necessary due to problems with samples sticking in the tube prior to the spectroscopy runs. Samples were purged of atmospheric gases entering during sample loading, combusted in series at 950 and 850 degrees Centigrade, and measured following written protocols for LECO TruSpec CN Elemental Determiner (LECO Inc., St. Joseph, Missouri, USA). For phosphorus samples CT Kjeldahl Digestion Mixture was used following like protocols. Analyses of variance (ANOVA) were conducted for several effects using SAS Version 8.2 (SAS Institute, Cary, North Carolina, USA). Above- and below-ground biomass ANOVAs were conducted using biomass as the dependant variable, and as independent variables, effects of experimental N:P ratio treatment, ammonia concentration treatment, species, and the interactions of N:P ratio treatment, ammonia treatment, and species were tested. Tissue nitrogen concentration ANOVAs were conducted using above- and below-ground tissue nitrogen concentration as the dependant variable, and as independent variables, effects of experimental N:P ratio treatment, ammonia concentration treatment, species, and the interactions of N:P ratio treatment, ammonia treatment, and species were tested. Interaction terms are of importance due to the fact that a significant N:P ratio by ammonia interaction could potentially infer that biomass production is maintained at high ammonia concentrations given phosphorus 58 fertilization. Total tissue mass nitrogen ANOVAs were conducted using above- and below-ground tissue total mass nitrogen as the dependant variable, and as independent variables, effects of experimental N:P ratio treatment, ammonia concentration treatment, species, and the interactions of N:P ratio treatment, ammonia treatment, and species were tested. Tissue phosphorus concentration ANOVAs were conducted using above- and below-ground tissue phosphorus concentration as the dependant variable, and as independent variables, effects of experimental N:P ratio treatment, ammonia concentration treatment, species, and the interactions of N:P ratio treatment, ammonia treatment, and species were tested. Interaction terms were tested in these cases as well to test if phosphorus additions at different ammonia concentrations significantly affected uptake or total mass of nitrogen and uptake of phosphorus. In each ANOVA, normality and homogeneity of residuals were examined for each variable, and data were log- transformed when necessary. Significant differences were determined at ?=0.05. RESULTS AND DISCUSSION Mesocosm Biomass Production Total biomass was significantly affected by ammonia concentration treatment and species tested (Table 3.2). Total mesocosm biomass declined from an average of 373.0?14.0 g at the 75 mgL -1 ammonia treatment level to an average of 277.5?12.4 g at the 300 mgL -1 ammonia treatment (Figure 3.2). This concurs with the 2002 results, in which biomass declined from the 75 mgL -1 ammonia treatment level to the 300 mgL -1 treatment level. Juncus again exhibited significantly higher overall average biomass than Typha, with an average of 223.5?14.0 g for Juncus versus 83.1?11.1 g for Typha. 59 Table 3.2. Tests of fixed effects on total, above-ground, and below-ground biomass of mesocosms of Typha latifolia and Juncus effusus in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Mesocosm Species 1 3.96 0.0538 Total Mesocosm [Ammonia] 3 7.58 0.0004 Total Mesocosm Species x [Ammonia] 3 0.69 0.5622 Total Mesocosm N:P Ratio 2 0.57 0.5691 Total Mesocosm N:P Ratio x Species 2 0.25 0.7824 Total Mesocosm N:P Ratio x [Ammonia] 6 0.42 0.8608 Total Mesocosm Species x N:P Ratio x [Ammonia] 6 1.1 0.3807 Above-ground Mesocosm Species 1 6.22 0.0171 Above-ground Mesocosm [Ammonia] 3 11.22 <0.0001 Above-ground Mesocosm Species x [Ammonia] 3 0.74 0.5328 Above-ground Mesocosm N:P Ratio 2 0.16 0.8557 Above-ground Mesocosm N:P Ratio x Species 2 1.07 0.3527 Above-ground Mesocosm N:P Ratio x [Ammonia] 6 0.5 0.8033 Above-ground Mesocosm Species x N:P Ratio x [Ammonia] 6 1.14 0.3564 Below-ground Mesocosm Species 1 6.84 0.0128 Below-ground Mesocosm [Ammonia] 3 8.53 0.0002 Below-ground Mesocosm Species x [Ammonia] 3 0.42 0.7364 Below-ground Mesocosm N:P Ratio 2 0.32 0.7268 Below-ground Mesocosm N:P Ratio x Species 2 0.85 0.4351 Below-ground Mesocosm N:P Ratio x [Ammonia] 6 0.41 0.8683 Below-ground Mesocosm Species x N:P Ratio x [Ammonia] 6 0.66 0.6815 60 0.0 100.0 200.0 300.0 400.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) T o t a l Meso co sm Bio m ass ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.2. Total mesocosm biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 61 Observable, although not statistically significant differences between the species at each ammonia concentration level corresponds well to the 2002 data (Table 2.5). Total above-ground mesocosm biomass was significantly affected by species and ammonia concentration treatment level (Table 3.2). A similar trend as for overall mesocosm biomass is evident, with steadily declining biomass production for higher levels of ammonia concentration (Figure 3.3); likewise overall Juncus above-ground biomass, 161.7?11.0 g, is significantly higher than for Typha above-ground biomass, 53.7?7.2 g. Of interest, however, is the slight increase in biomass production between the 75 mg/L -1 ammonia concentration and the 150 mg/L -1 ammonia concentration treatments for the 15:1 N:P ratio. It appears that at co-limited N:P levels, biomass production has a peak at higher nitrogen concentrations, although the peak production itself is not statistically different than the biomass production for other N:P ratios. Total below-ground biomass was significantly affected by ammonia concentration treatment and species tested (Table 3.2). Like above-ground biomass, a similar trend to that of total mesocosm biomass is evident, with toxicity effects at ammonia concentrations of above 75 mgL -1 treatment levels. Juncus effusus biomass production was 60.3?4.9 g, while Typha biomass was just 29.4?4.7 g. At all levels of ammonia treatment, Juncus produced significantly more biomass. Production levels of below- ground biomass were observably (but not statistically) greater at N:P levels of 15:1 for ammonia treatment concentrations of 225 mgL -1 and 300 mgL -1 (Figure 3.4). Of particular interest in this study is the effect of additions of phosphorus in varying ratios to ammonia-N in alleviating potentially toxic ammonia concentrations. Although in overall mesocosm biomass analysis, N:P ratio treatments are not a 62 0.0 50.0 100.0 150.0 200.0 250.0 300.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) A b o v e- g r ou nd Meso cosm Bi om ass ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.3. Total above-ground mesocosm biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 63 0.0 50.0 100.0 150.0 200.0 250.0 300.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Bel o w- g r ou nd M e s o cosm Bi oma s s ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.4. Total below-ground mesocosm biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 64 significant source of variation, some interesting trends are evident. Trends in N:P ratios within ammonia treatments indicate highest biomass production at levels other than 15:1 at lower treatments, and highest biomass production at levels of 15:1 at higher, toxic treatments. Although variability induced by N:P ratio is not statistically significant, the indication of higher biomass production when nutrients are co-limited is in agreement with other authors (Tilman 1982, Mamolos et al. 1995), although it has never been tested in phyto-inhibitory ammonia concentration levels. This unclear result may be elucidated by further testing; the initial indication that N:P ratio may be involved with reduction of ammonia toxicity is encouraging, although acceptance of non-significant data is problematic. Total Typha latifolia biomass production was significantly affected by ammonia concentration treatment alone (Table 3.3). Differences in biomass between ammonia treatments is likewise consistent with previous findings, and illustrate a clear toxic effect of ammonia concentrations higher than 75 mgL -1 (Figure 3.5); although 150 mgL -1 ammonia demonstrated a fertilization effect in the 2002 data. It is possible that year-to- year differences in such variables as temperature or photon flux density could affect nitrogen utilization in Typha. The N:P ratio of 5:1 produced observably higher biomass than the 15:1 or 25:1 N:P ratio, with an overall average of 96.2?11.4 g versus 76.2?14.1 g and 77.0?7.9 g, respectively. This may indicate that at N limitation, higher biomass is produced at all ammonia treatment levels, on average. Observably, the N:P ratio of 15:1 appears have less negative slope than the other lines at higher levels of ammonia. This could indicate that at even higher levels of ammonia, an N:P level of 15:1 may result in lower toxicity and more biomass production (Figure 3.5). Coupled with the statistical 65 Table 3.3. Tests of fixed effects on total, above-ground, and below-ground biomass of mesocosms of Typha latifolia in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Typha latifolia [Ammonia] 3 3.24 0.0452 Total Typha latifolia N:P Ratio 2 0.11 0.8961 Total Typha latifolia N:P Ratio x [Ammonia] 6 0.71 0.6430 Above-ground Typha latifolia [Ammonia] 3 5.4 0.0074 Above-ground Typha latifolia N:P Ratio 2 0.34 0.7191 Above-ground Typha latifolia N:P Ratio x [Ammonia] 6 0.75 0.6190 Below-ground Typha latifolia [Ammonia] 3 3.47 0.0378 Below-ground Typha latifolia N:P Ratio 2 0.28 0.7604 Below-ground Typha latifolia N:P Ratio x [Ammonia] 6 0.46 0.8263 66 0.0 50.0 100.0 150.0 200.0 250.0 300.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) T o ta l Ty p h a Bi om ass ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.5. Total Typha latifolia biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 67 non-significance of differences in biomass production at N:P ratios of 15:1 and 25:1, this finding may indicate that at N:P levels of 15:1, there is less effect of ammonia toxicity. Above-ground Typha biomass production is significantly affected by ammonia concentration treatment (Table 3.3). Difference in biomass production between ammonia treatments is consistent with the overall effect in Typha in 2003 and 2002 and with the above-ground effect in 2002 as well, although the 2002 data indicate toxic effects begin at more than 150 mgL -1 , while the 2003 data indicate toxicity at a lower ammonia concentration (Figure 3.6). The N:P ratio of 5:1 produced observably higher above- ground biomass, similar to the total Typha biomass results as well. The slope of the line of the 15:1 N:P ratio appears to be less negative at higher ammonia concentrations, although the interactive effect of N:P and ammonia concentration is not significant. Below-ground Typha biomass is significantly affected by ammonia treatment alone (Table 3.3), with the overall trend in biomass production observably similar to the total and above-ground Typha biomass curves (Figure 3.7). Highest production occurs at the 75 mgL -1 ammonia concentration, with toxic effects apparent thereafter. Although the N:P ratio is not significant in determining biomass production in Typha, the observable trend of the 15:1 ratio having the lowest negative slope is again potentially meaningful, especially at higher ammonia concentrations. Findings for Typha biomass production concur with the results of other researchers, who indicate a fertilization effect up to around 75 mgL -1 ammonia concentration (Hill et al. 1997). The effects of the N:P ratio in Typha are not particularly statistically apparent; higher biomass production occurs overall at an N:P ratio of 5:1, but the interactive effects of ammonia and the N:P ratio are unclear. The observable trend in 68 0.0 50.0 100.0 150.0 200.0 250.0 300.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) A b ov e - gro u n d Ty p h a Bi om ass ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.6. Above-ground Typha latifolia biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 69 0.0 50.0 100.0 150.0 200.0 250.0 300.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Bel o w- gr o und Ty pha Bi o m ass ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.7. Below-ground Typha latifolia biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 70 the effect of the N:P ratio is similar to that of the entire mesocosm, with higher biomass production at co-limited N:P ratios at high ammonia concentrations (Tilman 1982, Mamolos et al. 1995). Again, the lack of statistical significance of the N:P ratio is problematic in making determinations such as these. Total Juncus effusus biomass was significantly affected by ammonia concentration treatments (Table 3.4). Differences in biomass based on ammonia treatments are likewise consistent with results from other portions of this study, and illustrate toxicity in Juncus at approximately the 150 mgL -1 ammonia concentration. The peak biomass production of 238.2?8.9 g occurs at 150 mgL -1 ammonia treatment level, and is followed by a decline to a low of 193.5?9.9 g at 300 mgL -1 ammonia (Figure 3.8). Similar to the 2002 data, little statistically significant differences exist at higher ammonia treatment levels. The N:P ratio of 25:1 produced 230.3?10.0 g biomass, which is higher (although not statistically) than that produced by the 15:1 N:P ratio treatment level, 228.4?19.0 g, and significantly higher than that produced at an N:P ratio of 5:1, 207.3?9.9 g. Observably, the N:P ratio of 15:1 appears once again to have the least negative slope of the three, which could potentially indicate that at higher levels of ammonia, a co-limited N:P ratio would produce higher biomass, which is consistent with findings from Typha and for the mesocosms on the whole. Above-ground Juncus biomass production is significantly affected by ammonia concentration treatment alone, although N:P ratio treatment almost produces an observable effect as well (Table 3.4). Differences in biomass production across ammonia concentrations is consistent with the overall effect in Juncus in 2002 and 2003, although peak biomass production in 2002 occurs at the 75 mgL -1 ammonia concentration and 71 Table 3.4. Tests of fixed effects on total, above-ground, and below-ground biomass of mesocosms of Juncus effusus in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Juncus effusus [Ammonia] 3 9.46 0.0005 Total Juncus effusus N:P Ratio 2 2.03 0.1595 Total Juncus effusus N:P Ratio x [Ammonia] 6 1.11 0.3949 Above-ground Juncus effusus [Ammonia] 3 9.46 0.0005 Above-ground Juncus effusus N:P Ratio 2 2.03 0.1595 Above-ground Juncus effusus N:P Ratio x [Ammonia] 6 1.11 0.3949 Below-ground Juncus effusus [Ammonia] 3 10.35 0.0003 Below-ground Juncus effusus N:P Ratio 2 2.44 0.1143 Below-ground Juncus effusus N:P Ratio x [Ammonia] 6 1.01 0.4483 72 0.0 50.0 100.0 150.0 200.0 250.0 300.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) T o ta l Jun c us Bi o m ass ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.8. Total Juncus effusus biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 73 occurs in 2003 at the 150 mgL -1 ammonia concentration (Figure 3.9). Although the N:P ratio is not a statistically significant effect in determining biomass production, there is an observable trend of higher biomass production at a level of 15:1 as predicted by Tilman (1982), Mamolos et al. (1995) and by the observable trends in this study. Below-ground Juncus biomass production is significantly affected by the ammonia treatment alone (Table 3.4). The highest production occurs at the 75 mgL -1 ammonia concentration, with lowest production at 300 mgL -1 although observable toxic effects do not seem overly evident (Figure 3.10). An observable trend of high biomass production along the 15:1 N:P ratio may be found in Figure 3.10, although N:P ratio is not a statistically significant source of variation. Findings for Juncus biomass production do concur with the results of Clarke (1999), but yield overall higher tolerances for ammonia than some (Hill et al. 1997); results from the 2003 study corroborate well with the 2002 study in indicating relatively high ammonia tolerance in Juncus. Overall, Juncus again produced more biomass than Typha at all ammonia treatment concentrations, and appeared to be less affected by increased ammonia concentrations, based on observable patterns in the slopes of the lines in Figures 3.8 to 3.10. The trend in the effect of the N:P ratio is much more difficult to observe in Juncus, but appears to agree with that of the entire mesocosm and with that of Typha biomass production, with more biomass produced at high ammonia concentrations if the experimental N:P ratio is co-limited. If the application of phosphorus in varying N:P ratios could alleviate potentially toxic ammonia concentrations, interactions between the N:P ratio treatments and the ammonia concentration treatments should be statistically significant sources of variation. 74 0.0 50.0 100.0 150.0 200.0 250.0 300.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Abo v e- g r ou nd Ju ncu s Bi om ass ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.9. Above-ground Juncus effusus biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 75 0.0 50.0 100.0 150.0 200.0 250.0 300.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Bel o w- gr o und Jun c us Bi o m ass ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.10. Below-ground Juncus effusus biomass production along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 76 This is not the case, but some interesting trends are evident that could be further investigated. It appears that for Juncus, as well as for Typha, that under low ammonia concentrations, best growth is achieved when the N:P ratio treatment is 5:1 or 25:1. Under high ammonia concentrations, best growth is achieved when N and P are co- limiting. A potential competitive effect could account for this trend in biomass production; Tilman et al. (1999) found in field plot studies (using significantly lower concentrations of nutrients than used in this study) that the competitor with the best mechanism for acquiring a limiting nutrient will be favored in the absence of that nutrient. It is possible (although Tilman 1982 states otherwise, again in field plot studies) that Typha is a more efficient competitor for both N and P at lower nutrient concentrations, and is therefore inhibited in growth when N and P are co-limited. On the other hand, Juncus produces biomass relatively well at high ammonia concentrations regardless of P fertilization, which may indicate that Juncus is the better competitor for phosphorus in P-limiting situations. Mesocosm Root-to-Shoot Ratios Mesocosm root-to-shoot ratios were significantly affected by ammonia concentration treatment, N:P ratio treatment, species tested, and the interaction of species and ammonia concentration treatment (Table 3.5). Decreases were evident in the R:S ratio in the mesocosm from 0.53?0.1 at the 75 mgL -1 ammonia treatment to 0.4?0.0 at the 300 mgL -1 ammonia treatment; R:S ratio was maximized at 0.4?0.0 at the N:P ratio treatment of 15:1 (Figure 3.11). Typha exhibited a higher mean R:S ratio at 0.5?0.1 than Juncus, at 0.4?0.0. Decreases following the same trend as for the mesocosm are evident 77 Table 3.5. Test of fixed effects on the root-to-shoot ratio in mesocosms and in Typha latifolia and Juncus effusus in 2003. Please refer to appendix II for SAS code. Measured Root-to-Shoot Ratio Effect d.f. F Value P Level Total Mesocosm Species 1 26.31 <0.0001 Total Mesocosm [Ammonia] 3 13.1 <0.0001 Total Mesocosm Species x [Ammonia] 3 7.16 0.0005 Total Mesocosm N:P Ratio 2 5.62 0.0064 Total Mesocosm N:P Ratio x Species 2 0.95 0.3946 Total Mesocosm N:P Ratio x [Ammonia] 6 0.9 0.5030 Total Mesocosm Species x N:P Ratio x [Ammonia] 6 1.82 0.1157 Total Typha latifolia [Ammonia] 3 10.07 0.0002 Total Typha latifolia N:P Ratio 2 2.26 0.1258 Total Typha latifolia N:P Ratio x [Ammonia] 6 1.13 0.3749 Total Juncus effusus [Ammonia] 3 10.5 0.0001 Total Juncus effusus N:P Ratio 2 9.91 0.0007 Total Juncus effusus N:P Ratio x [Ammonia] 6 2.83 0.0314 78 0 0.5 1 1.5 2 0 75 150 225 300 Ammonia Concentration (mgL -1 ) M e s o c o s m R o ot - t o- Sh oot R a ti o ( R g/ S g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.11. Root-to-shoot ratios for mesocosms along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 79 for Typha (Figure 3.12) and Juncus (Figure 3.13), although N:P ratio is not significant in determining the R:S ratio in Typha (Table 3.5). These data concur with the 2002 data and indicate suppression of root formation, increased shoot production, or a combination of the two with increasing ammonia concentration. The fact that the interaction term of species and ammonia concentration is present indicates differences in the level of stress tolerance between the species. Overall decreases in R:S ratio with increasing ammonia concentration concur with similar responses in other species (Walch-Liu et al. 2000, Vojtiskova et al. 2004). It is not to be expected that biomass allocations to roots and shoots will be the same, based on a number of factors which include fertilization (Vojtiskova et al. 2004, Tilman 1997). The fact that the 15:1 N:P ratio alleviated some of the reduction in cell division and leaf elongation typical of phytotoxic concentrations of ammonia is consistent with the observable trend in increased overall biomass production at the 15:1 N:P level in mesocosms, Typha, and Juncus. Mesocosm Tissue Nitrogen Concentration Total mesocosm nitrogen tissue concentration was significantly affected by species tested, ammonia concentration, N:P ratio treatment, and the interaction of species and ammonia concentration. (Table 3.6). Total tissue nitrogen concentration increased from an average of 19.3?2.1 mg g -1 at the 75 mgL -1 ammonia treatment level to an average of 26.7?0.5 mg g -1 at the 300 mgL -1 ammonia treatment level (Figure 3.14). This trend is to be expected and has been referred to as ?luxury consumption?; as the available nitrogen increases, nitrogen concentration in the biomass should increase as well (Tripler et al. 2002). This finding is one of the main reasons why plants are used in treatment 80 0.0 0.5 1.0 1.5 2.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Ty pha R o ot -to-S h o o t R a tio (R g/ Sg) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.12. Root-to-shoot ratios for Typha latifolia along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 81 0.0 0.5 1.0 1.5 2.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Jun c us Ro ot - t o- S h o o t Rat i o ( R g/ S g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.13. Root-to-shoot ratios for Juncus effusus along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 82 Table 3.6. Tests of fixed effects on total, above-ground, and below-ground tissue nitrogen concentration of mesocosms of Juncus effusus and Typha latifolia in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Mesocosm Species 1 7.9 0.0078 Total Mesocosm [Ammonia] 3 10.32 <0.0001 Total Mesocosm Species x [Ammonia] 3 3.69 0.02 Total Mesocosm N:P Ratio 2 11.37 0.0001 Total Mesocosm N:P Ratio x Species 2 1.89 0.1644 Total Mesocosm N:P Ratio x [Ammonia] 6 1.39 0.2426 Total Mesocosm Species x N:P Ratio x [Ammonia] 6 0.71 0.6426 Above-ground Mesocosm Species 1 10.42 0.0026 Above-ground Mesocosm [Ammonia] 3 12.7 <0.0001 Above-ground Mesocosm Species x [Ammonia] 3 4.79 0.0063 Above-ground Mesocosm N:P Ratio 2 12.94 <0.0001 Above-ground Mesocosm N:P Ratio x Species 2 2.09 0.1374 Above-ground Mesocosm N:P Ratio x [Ammonia] 6 1.51 0.2 Above-ground Mesocosm Species x N:P Ratio x [Ammonia] 6 0.6 0.7309 Below-ground Mesocosm Species 1 16.42 0.0003 Below-ground Mesocosm [Ammonia] 3 17.2 <0.0001 Below-ground Mesocosm Species x [Ammonia] 3 7.38 0.0005 Below-ground Mesocosm N:P Ratio 2 15.35 <0.0001 Below-ground Mesocosm N:P Ratio x Species 2 2.71 0.0794 Below-ground Mesocosm N:P Ratio x [Ammonia] 6 1.99 0.0916 Below-ground Mesocosm Species x N:P Ratio x [Ammonia] 6 0.53 0.7811 83 0.0 10.0 20.0 30.0 40.0 50.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) T o t a l Meso co sm T i ssu e N i t r o g e n ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.14. Total mesocosm tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 84 wetlands: their ability to remove designated elements from wastewater (e.g. Kadlec and Knight 1996). Typha averaged significantly higher tissue nitrogen content, 24.3?3.6 mg g -1 , than Juncus, which averaged 21.7?1.2 mg g -1 tissue nitrogen. Average tissue nitrogen content was highest at the 5:1 N:P ratio. The significant interactive effect of species and ammonia treatment reflects significant differences between tissue nitrogen concentrations in each species across ammonia concentrations. From this study, the hope was that a significant N:P ratio effect would develop; this would imply that as N:P ratios changed, more or less nitrogen could be sequestered in plant tissues. The line following the 15:1 N:P ratio in Figure 3.14 shows a trend towards less tissue nitrogen at low ammonia concentrations. This could be due to an inability of the mesocosms to uptake nitrogen, even under ?ideal? N:P ratios, at low ammonia concentrations. Other authors (de Groot et al. 2003) found nitrogen uptake to be limited by phosphorus availability; at high N:P ratios where phosphorus is the limiting factor, N uptake was inhibited. Total above-ground mesocosm tissue nitrogen concentration was significantly affected by the N:P ratio treatment, species, and ammonia concentration treatment level (Table 3.6). A similar trend as for overall mesocosm tissue nitrogen concentration is evident, with increasing tissue nitrogen with increasing ammonia treatment concentration (Figure 3.15); likewise, average Typha above-ground tissue nitrogen concentration, 24.5?2.9 mg g -1 , is higher than that for Juncus, 21.7?1.2 mg g -1 . Again the 5:1 N:P ratio results in highest tissue nitrogen concentration. The line representing the 15:1 N:P ratio demonstrates the same interesting trend as that for overall tissue nitrogen, an apparent inability of the plant tissues to sequester nitrogen at low ammonia concentrations. In fact, at this 15:1 N:P ratio, a significantly lower nitrogen tissue concentration occurs over the 85 0.0 10.0 20.0 30.0 40.0 50.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Ab ov e- gr o u n d Mes o co s m T i s s u e Ni t r og en ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.15. Total above-ground mesocosm tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 86 entire ammonia treatment regime,. At co-limited N:P levels, less tissue nitrogen uptake occurs. Total below-ground tissue nitrogen concentration was significantly affected by N:P ratio treatment, species, and ammonia concentration treatment as well (Table 3.6). Like above-ground tissue nitrogen, a similar trend to that of overall mesocosm tissue nitrogen concentration is evident, with greater tissue nitrogen concentrations at higher ammonia concentrations and at the 5:1 N:P ratio (Figure 3.16), and a greater concentration of tissue nitrogen found in Typha. Of interest to note is the fact that average Juncus tissue nitrogen concentration is exactly the same in above- and below- ground tissues, 21.7?1.2 mg g -1 . This indicates that the plant sequesters nitrogen equally in above- and below-ground tissues, which may have implications as to how nitrogen is used in plant tissues and as to how ammonia toxicity occurs in wetland plants. If the N:P ratio treatment was a useful means of alleviating ammonia toxicity, N:P ratio treatments should significantly affect the nitrogen stored and used in plant tissue. For ammonia toxicity to decrease, some of the nitrogen must be used by the biomass. In this experiment, N:P ratio did significantly affect tissue nitrogen concentration, with the N:P ratio of 5:1 producing significantly higher tissue N concentration. In fact, at the N:P ratio which is co-limited, 15:1, the highest stored tissue nitrogen was expected. At a co- limited N:P ratio, both nutrients should be used more quickly for the production of biomass (Tilman 1982), and, while the observable trend in biomass production did show this to an extent, the amount of nitrogen stored in the tissue is the lowest of all three N:P ratio treatments. At sub-optimal N:P ratios, plant tissues should not be able to use nutrients at an optimal rate. Perhaps the difference in this case is between ?use? of 87 0.0 10.0 20.0 30.0 40.0 50.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Be lo w - g r o u n d Meso c o sm T i ssu e N i t r o g en ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.16. Total below-ground mesocosm tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 88 nutrients and ?storage? of nutrients, which should be further explored. Aerts et al.(2001) indicate that N and P tissue concentration in Sphagnum is not controlled by the addition of nutrients to the site, although at significantly lower fertilization rates than in this study. Total Typha latifolia tissue nitrogen concentration was significantly affected by ammonia concentration treatment and N:P ratio treatment (Table 3.7). Tissue nitrogen increased from an average of 19.3?3.6 mg g -1 at the 75 mgL -1 ammonia concentration treatment to 29.5?3.7 at the 225 mgL -1 ammonia concentration treatment, and then declined to the 300 mgL -1 ammonia concentration treatment (Figure 3.17). Tissue nitrogen was highest at the 15:1 N:P ratio. The fact that peak tissue nitrogen concentration appears at the 225 mgL -1 ammonia treatment could indicate the toxic effects of ammonia could potentially limit the overall uptake. Although 15:1 N:P ratio treatments were significant in determining high tissue nitrogen concentrations, this treatment again performed poorly at high ammonia concentrations. At the high phosphorus N:P ratio, the slope of the line indicates that tissue nitrogen concentration may continue to increase at even higher ammonia concentrations. The behavior of the above-ground Typha tissue nitrogen concentration (Figure 3.18) and below-ground tissue nitrogen concentration (Figure 3.19) demonstrate a similar trend; for each case, ammonia concentration and the N:P ratio account for differences in tissue nitrogen concentration. The overall N:P ratio trend is similar in each case as well, with the line representing the N:P ratio of 15:1 demonstrating a peak at 225 mgL -1 ammonia, and the 5:1 N:P ratio line increasing at 300 mgL -1 ammonia. Total Juncus effusus tissue nitrogen concentration is affected by ammonia concentration treatment and the N:P ratio treatment (Table 3.8). Mean tissue nitrogen 89 Table 3.7. Tests of fixed effects on total, above-ground, and below-ground tissue nitrogen concentration of Typha latifolia in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Typha latifolia [Ammonia] 3 7.74 0.0014 Total Typha latifolia N:P Ratio 2 6.56 0.0068 Total Typha latifolia N:P Ratio x [Ammonia] 6 0.86 0.5438 Above-ground Typha latifolia [Ammonia] 3 9.92 0.0004 Above-ground Typha latifolia N:P Ratio 2 7.59 0.0038 Above-ground Typha latifolia N:P Ratio x [Ammonia] 6 0.84 0.5572 Below-ground Typha latifolia [Ammonia] 3 13.52 <0.0001 Below-ground Typha latifolia N:P Ratio 2 9.27 0.0017 Below-ground Typha latifolia N:P Ratio x [Ammonia] 6 1.01 0.4517 90 0.0 10.0 20.0 30.0 40.0 50.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) To t a l T y p ha T i ssu e Ni t r og en ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.17. Total Typha latifolia tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 91 0.0 10.0 20.0 30.0 40.0 50.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Ab ov e- gr o u n d T y ph a T i s s ue N i trog e n (mg g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.18. Above-ground Typha latifolia tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 92 0.0 10.0 20.0 30.0 40.0 50.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Bel o w- g r oun d T y p ha T i ssu e Ni t r og en ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.19. Below-ground Typha latifolia tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 93 Table 3.8. Tests of fixed effects on total, above-ground, and below-ground tissue nitrogen concentration of Juncus effusus in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Juncus effusus [Ammonia] 3 3.03 0.0549 Total Juncus effusus N:P Ratio 2 7.11 0.0049 Total Juncus effusus N:P Ratio x [Ammonia] 6 2.17 0.0922 Above-ground Juncus effusus [Ammonia] 3 3.03 0.0549 Above-ground Juncus effusus N:P Ratio 2 7.11 0.0049 Above-ground Juncus effusus N:P Ratio x [Ammonia] 6 2.17 0.0922 Below-ground Juncus effusus [Ammonia] 3 3.03 0.0549 Below-ground Juncus effusus N:P Ratio 2 7.11 0.0049 Below-ground Juncus effusus N:P Ratio x [Ammonia] 6 2.17 0.0922 94 concentration increases from 17.5?1.0 mg g -1 at 75 mgL -1 ammonia to 24.5?0.6 mg g -1 at the 300 mgL -1 ammonia treatment. Mean tissue nitrogen concentration likewise increases from 20.2?1.7 mg g -1 at an N:P ratio of 15:1 to about 22 mg g -1 at N:P ratios of 5:1 and 25:1. Figure 3.20 demonstrates that, especially under low ammonia concentrations, tissue nitrogen is minimized by the 15:1 N:P ratio treatment. At N-limited and at P- limited N:P ratios, more nitrogen is stored in tissues in Juncus, especially in the lower ammonia concentrations, although this effect is significantly smaller than in Typha. In the related Juncus roemerianus, Brewer (2003) noted tissue nitrogen conservation and little increase in N uptake with fertilization. Emery et al. (2001) noted Juncus effusus to be a ?stress tolerator? and could therefore potentially be dominant under unfavorable situations such as ammonia toxicity; they note that ?stress tolerators? are not typically considered dominant in the presence of nutrients. This trend towards more consistent storage is not similar to the trend in Typha, where at higher ammonia concentrations, the N-limited and P-limited N:P ratios tend to cause less nitrogen storage in tissues. Curves are similar in above-ground Juncus tissue nitrogen concentration (Figure 3.21) and in below-ground nitrogen tissue concentration (Figure 3.22). Mesocosm Total Tissue Mass Nitrogen Total mesocosm tissue nitrogen uptake is affected by ammonia concentration, species, the N:P ratio treatment, and the interaction between species and ammonia concentration (Table 3.9). Total tissue nitrogen decreased with increasing ammonia concentration, from an average of 7301.2?834.3 mg at the 150 mgL -1 ammonia treatment level to 5981.1?854.3 mg at the 300 mgL -1 treatment ( Figure 3.23). This indicates that 95 0.0 10.0 20.0 30.0 40.0 50.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) To t a l J unc us T i ssue Ni tr o g en ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.20. Total Juncus effusus tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 96 0.0 10.0 20.0 30.0 40.0 50.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Ab ov e- gr o u n d Jun c us T i s s u e Ni t r og en ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.21. Above-ground Juncus effusus tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 97 0.0 10.0 20.0 30.0 40.0 50.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Bel o w- g r ou nd J unc us T i ssue Nit r o g en ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.22. Below-ground Juncus effusus tissue nitrogen concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 98 Table 3.9. Tests of fixed effects on total, above-ground, and below-ground tissue total mass nitrogen in mesocosms of Typha latifolia and Juncus effusus in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Mesocosm Species 1 10.09 0.003 Total Mesocosm [Ammonia] 3 9.73 <0.0001 Total Mesocosm Species x [Ammonia] 3 2.77 0.0548 Total Mesocosm N:P Ratio 2 5.68 0.0069 Total Mesocosm N:P Ratio x Species 2 1.3 0.2848 Total Mesocosm N:P Ratio x [Ammonia] 6 1.17 0.3406 Total Mesocosm Species x N:P Ratio x [Ammonia] 6 0.62 0.7147 Above-ground Mesocosm Species 1 13.03 0.0009 Above-ground Mesocosm [Ammonia] 3 12.62 <0.0001 Above-ground Mesocosm Species x [Ammonia] 3 3.86 0.0167 Above-ground Mesocosm N:P Ratio 2 6.4 0.004 Above-ground Mesocosm N:P Ratio x Species 2 1.59 0.2179 Above-ground Mesocosm N:P Ratio x [Ammonia] 6 1.31 0.2744 Above-ground Mesocosm Species x N:P Ratio x [Ammonia] 6 0.61 0.721 Below-ground Mesocosm Species 1 15.76 0.0003 Below-ground Mesocosm [Ammonia] 3 12.33 <0.0001 Below-ground Mesocosm Species x [Ammonia] 3 3.85 0.0171 Below-ground Mesocosm N:P Ratio 2 6.45 0.004 Below-ground Mesocosm N:P Ratio x Species 2 1.95 0.1561 Below-ground Mesocosm N:P Ratio x [Ammonia] 6 1.43 0.23 Below-ground Mesocosm Species x N:P Ratio x [Ammonia] 6 0.42 0.863 99 4000 6000 8000 10000 0 75 150 225 300 Ammonia Concentration (mgL -1 ) T o tal Mesocos m Ti ssu e N i tro g en (g) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.23. Total mesocosm tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 100 total nitrogen uptake may be based primarily on total biomass rather than on tissue nitrogen concentration. Although more nitrogen is taken up at high ammonia treatment levels, more total nitrogen is removed from wastewater at peak biomass production. Total N uptake is greatest at the 5:1 N:P ratio treatment; this finding is similar to tissue N concentration results (Figures 3.23 and 3.14, respectively). Since Juncus produced more biomass than Typha at all treatment levels, it sequestered significantly higher mass of N in the tissue than Typha, although Typha has higher tissue N concentrations. Although this may seem to indicate that Typha and Juncus therefore operate in the same manner in terms of overall N removal, since Juncus produces more biomass and is more tolerant of high ammonia concentrations, it is able to remove more N from wastewater. Total above-ground mesocosm tissue mass nitrogen was significantly affected by the same effects as total mesocosm tissue mass nitrogen, and in much the same way; ammonia concentration, species, the N:P ratio treatment, and the interaction between species and ammonia concentration affect total above-ground mesocosm tissue mass nitrogen (Table 3.9). Total above-ground tissue mass nitrogen decreased with increasing ammonia concentration, due in part to the fact that less biomass was produced at high ammonia concentrations, and despite higher tissue nitrogen concentrations at high ammonia concentrations (Figure 3.24). Juncus had higher above-ground tissue mass nitrogen at all treatment levels, and averaged 3526.4?409.6 mg over all treatments. Typha averaged 1242.3?286.7 mg over all treatments. Highest N uptake occurs at the 5:1 N:P ratio treatment; this is similar to overall mesocosm tissue N and to tissue N concentration results. This may indicate a result of a low N:P ratio, and demonstrates 101 2000 4000 6000 8000 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Ab ov e - g r ou nd M e s o c o sm Tis s u e Nitr og e n ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.24. Total above-ground mesocosm tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 102 that when less N is available, plant tissues sequester more overall N and sequester N in higher concentrations. Total below-ground tissue mass nitrogen was significantly affected by N:P ratio treatment, species, and ammonia concentration treatment and by the interaction of species and ammonia concentration as well (Table 3.9). Like above-ground tissue mass nitrogen, a similar trend to overall mesocosm mass nitrogen is evident, with greater mass nitrogen at lower ammonia concentrations and at the 5:1 N:P ratio (Figure 3.25), and a greater mass of N found in Juncus. Again, this trend indicates a clear relationship between total biomass production and total N uptake; since Juncus produced more biomass, total N in Juncus tissue is higher than that of Typha, despite a higher N concentration in Typha. If the amending the N:P ratio could improve ATW functioning, N:P ratio treatments should significantly affect the total mass N stored in plant tissues. In this experiment, the N:P ratio of 5:1 produced significantly higher N mass uptake in plant tissues, as well as producing higher overall tissue N concentration. At a co-limited N:P ratio, nutrients should be used for biomass production, and therefore should be evident in high tissue mass N (Tilman 1982). In fact, at the lower, N-limited N:P ratio (5:1), higher N mass uptake and higher N concentration occurred. If the 5:1 N:P ratio had likewise produced overall higher biomass, then a precedent could be set for managers to improve ATW function by setting an N:P ratio of 5:1. Despite the 5:1 N:P ratio not statistically influencing biomass, the fact that over-fertilization with phosphorus may serve to remove additional nitrogen from wastewater may be a practical means by which to control ammonia concentration in ATW effluent. 103 0 1000 2000 3000 4000 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Be low-g r ou nd M e s o co sm Tis s u e Nit r og en (g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.25. Total below-ground mesocosm tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 104 Total Typha latifolia tissue mass nitrogen was significantly affected by ammonia concentration treatment alone (Table 3.10). Tissue mass N decreased with from an average of 2731.0?796.3 mg at the 75 mgL -1 ammonia concentration treatment to 787.5?89.3 mg at the 300 mgL -1 ammonia concentration treatment (Figure 3.26). This concurs well with the tissue N concentration results, and indicates N uptake problems at the higher ammonia concentrations, as well as suppressed biomass production at those levels. Although the N:P ratio treatment was not significant in determining Typha tissue N mass, the 5:1 N:P ratio produces observably higher N mass uptake, which is a similar overall trend to total mesocosm N uptake, and may as well indicate the management implications of this study. The behavior of above-ground Typha tissue mass N (Figure 3.27), and below-ground tissue mass N (Figure 3.28) demonstrates similar trends. In fact, the 5:1 N:P ratio produces statistically significant high N mass for these measurements (Table 3.10). Again, the decrease in biomass at high ammonia treatments appears to affect total N uptake in Typha above- and below-ground tissues; this corresponds well with total mesocosm effects. Total Juncus effusus tissue mass nitrogen is affected by ammonia concentration treatment, the N:P ratio treatment, and their interaction (Table 3.11). Mean tissue mass N increases from 4068.9?588.2 mg at the 75 mgL -1 ammonia concentration to 5222.41?318.1 mg at the 225 mgL -1 ammonia concentration, then drops to 4798?333.8 mg at the 300 mgL -1 ammonia concentration. Highest tissue mass N is produced by the 25:1 N:P ratio treatment. Figure 3.29 illustrates the interactive effects of ammonia concentration and N:P ratio treatment in affecting tissue N mass, and shows the higher ammonia treatments have relatively similar overall N mass uptake. This corresponds 105 Table 3.10. Tests of fixed effects on total, above-ground, and below-ground tissue total mass nitrogen in Typha latifolia in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Typha latifolia [Ammonia] 3 6.03 0.0046 Total Typha latifolia N:P Ratio 2 3.19 0.0637 Total Typha latifolia N:P Ratio x [Ammonia] 6 0.85 0.5493 Above-ground Typha latifolia [Ammonia] 3 8.11 0.0011 Above-ground Typha latifolia N:P Ratio 2 3.71 0.0435 Above-ground Typha latifolia N:P Ratio x [Ammonia] 6 0.94 0.4915 Below-ground Typha latifolia [Ammonia] 3 7.42 0.0019 Below-ground Typha latifolia N:P Ratio 2 3.75 0.0437 Below-ground Typha latifolia N:P Ratio x [Ammonia] 6 0.82 0.5655 106 0 1000 2000 3000 4000 0 75 150 225 300 Ammonia Concentration (mgL -1 ) To t a l Ty p h a Tissu e Nitr og en ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.26. Total Typha latifolia tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 107 0 1000 2000 3000 4000 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Abov e-g r ou nd Ty p h a T i ss ue Nitr oge n (g) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.27. Above-ground Typha latifolia tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 108 0 1000 2000 3000 4000 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Bel o w-gr oun d Ty ph a Ti s s u e Ni t r og en (g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.28. Below-ground Typha latifolia tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 109 Table 3.11. Tests of fixed effects on total, above-ground, and below-ground tissue total mass nitrogen in Juncus effusus in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Juncus effusus [Ammonia] 3 14.14 <0.0001 Total Juncus effusus N:P Ratio 2 10.71 0.0008 Total Juncus effusus N:P Ratio x [Ammonia] 6 2.83 0.0386 Above-ground Juncus effusus [Ammonia] 3 12.31 0.0001 Above-ground Juncus effusus N:P Ratio 2 8.98 0.0018 Above-ground Juncus effusus N:P Ratio x [Ammonia] 6 2.27 0.0806 Below-ground Juncus effusus [Ammonia] 3 16 <0.0001 Below-ground Juncus effusus N:P Ratio 2 12.56 0.0003 Below-ground Juncus effusus N:P Ratio x [Ammonia] 6 3.47 0.0173 110 2000 4000 6000 8000 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Total J u nc us Ti ssu e Ni t r og en ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.29. Total Juncus effusus tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 111 well with overall biomass production results for Juncus, with little toxicity effect at higher ammonia concentrations. Of interest to note is that the P-limited 25:1 N:P ratio produced higher tissue N mass than co-limited or N-limited ratios, as in the case of overall mesocosm and Typha tissue mass N. This may indicate interactions with Typha growth at 5:1 and 15:1 N:P levels, and may also indicate the ?stress tolerator? syndrome described above (Emery et al. 2001). Curves are similar for above-ground Juncus tissue mass N (Figure 3.30) and below-ground tissue mass N (Figure 3.31), although in below- ground tissue mass N, there is a less evident effect. This corresponds well with lower biomass and N concentration effects for Juncus previously noted. Mesocosm Tissue Phosphorus Concentration Total mesocosm tissue phosphorus concentration is affected by ammonia concentration, species, and the interactions between N:P ratio treatment and species and action between ammonia and species (Table 3.12). Since phosphorus administered by the N:P ratio is the only phosphorus available to the plant (other than phosphorus present in the tap water, which should be consistent between mesocosms), the non-significance of the N:P ratio is unclear and may represent additional phosphorus sources such as the greenhouse tap water or the medium in which the plants were grown. Mean tissue phosphorus concentration in Juncus is 2.1?0.2 mg g -1 , while that of Typha is 2.7?0.5 mg g-1, indicating that, like nitrogen, Typha typically sequesters more P in tissues. It is the significance of the ammonia concentration which is of particular interest here; tissue phosphorus concentrations increase from a low of 1.6?0.1 mg g -1 at the 75 mgL -1 ammonia treatment to 3.1?0.4 mg g -1 at the 300 mgL -1 ammonia treatment (Figure 3.32). 112 2000 4000 6000 8000 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Ab ov e - g r ou nd J unc us T i s s ue Nit r og en (g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.30. Above-ground Juncus effusus tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 113 0 1000 2000 3000 4000 0 75 150 225 300 Ammonia Concentration (mgL -1 ) Be low-g r ou nd Ju ncu s Ti ssu e Ni t r o g en ( g ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.31. Below-ground Juncus effusus tissue mass nitrogen along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 114 Table 3.12. Tests of fixed effects on total, above-ground, and below-ground tissue phosphorus concentration of mesocosms of Typha latifolia and Juncus effusus in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Mesocosm Species 1 503.07 <0.0001 Total Mesocosm [Ammonia] 3 26.7 <0.0001 Total Mesocosm Species x [Ammonia] 3 6.06 0.0014 Total Mesocosm N:P Ratio 2 0.03 0.9671 Total Mesocosm N:P Ratio x Species 2 5.02 0.0104 Total Mesocosm N:P Ratio x [Ammonia] 6 0.4 0.8769 Total Mesocosm Species x N:P Ratio x [Ammonia] 6 0.78 0.5876 Above-ground Mesocosm Species 1 539.76 <0.0001 Above-ground Mesocosm [Ammonia] 3 18.14 <0.0001 Above-ground Mesocosm Species x [Ammonia] 3 3.51 0.0222 Above-ground Mesocosm N:P Ratio 2 0.07 0.9302 Above-ground Mesocosm N:P Ratio x Species 2 3.51 0.0378 Above-ground Mesocosm N:P Ratio x [Ammonia] 6 0.27 0.9486 Above-ground Mesocosm Species x N:P Ratio x [Ammonia] 6 0.55 0.7703 Below-ground Mesocosm Species 1 151.6 <0.0001 Below-ground Mesocosm [Ammonia] 3 31.37 <0.0001 Below-ground Mesocosm Species x [Ammonia] 3 10.15 <0.0001 Below-ground Mesocosm N:P Ratio 2 0.57 0.5708 Below-ground Mesocosm N:P Ratio x Species 2 4.27 0.0197 Below-ground Mesocosm N:P Ratio x [Ammonia] 6 0.64 0.6983 Below-ground Mesocosm Species x N:P Ratio x [Ammonia] 6 0.83 0.5559 115 0 1 2 3 4 5 0 75 150 225 300 Ammonia Concentration (mgL -1 ) T o t a l M e s o c o s m Ti s s u e P h o s por hus ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.32. Total mesocosm tissue phosphorus concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Above- and below- ground total phosphorus concentrations are not shown. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 116 This indicates a link between the ammonia concentration treatment applied and the ability of the biomass to absorb phosphorus. Other researchers have found using non- wetland species (Allium sp.) that at high ammonia concentrations, P uptake is inhibited, and have cited this as a potential cause of low biomass production (Abbes at al. 1995). The findings here suggest otherwise: since tissue nitrogen levels are so high, the amount of phosphorus needed by those tissues to supply a ?normal? ratio is elevated as well. This finding shows some level of control by the plants over the tissue level N:P ratio. The P-enriched 5:1 N:P level treatment experiences an observable decline in total P absorption in the highest ammonia treatment level, indicating that some blockage in P assimilation occurs at toxic ammonia levels. Ammonia toxicity tends to decrease membrane transport efficiencies as well as decrease the functioning of enzymes or of mRNA related to nutrient uptake and storage; the blockage in P uptake may be due to this effect. Phosphorus deficiencies have been noted at phytotoxic nutrient concentrations by Abbes et al. (1995). Finding and using mutants which overexpress genes encoding high- affinity phosphate receptors may be one means by which this deficiency may be overcome, although for the time being mutant strains overexpressing phosphate transporters have been reported in Arabidopsis alone (Smith 2002). Figures for above- and below-ground mesocosm phosphorus tissue concentration are not shown, but show the same overall trends as for total mesocosm phosphorus tissue concentration (Table 3.12). Total Typha tissue phosphorus concentration is affected by ammonia treatment alone (Table 3.13). Mean tissue phosphorus concentrations increased from a low of 1.6 ?0.3 mg g -1 at the 75 mgL -1 ammonia treatment to 3.5?0.8 mg g -1 at the 300 mgL -1 117 Table 3.13. Tests of fixed effects on total, above-ground, and below-ground tissue phosphorus concentration of Typha latifolia in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Typha latifolia [Ammonia] 3 33.94 <0.0001 Total Typha latifolia N:P Ratio 2 2.75 0.0838 Total Typha latifolia N:P Ratio x [Ammonia] 6 1.15 0.3636 Above-ground Typha latifolia [Ammonia] 3 24.72 <0.0001 Above-ground Typha latifolia N:P Ratio 2 3.8 0.0367 Above-ground Typha latifolia N:P Ratio x [Ammonia] 6 1.02 0.4374 Below-ground Typha latifolia [Ammonia] 3 37.51 <0.0001 Below-ground Typha latifolia N:P Ratio 2 0.89 0.4219 Below-ground Typha latifolia N:P Ratio x [Ammonia] 6 1 0.4492 118 ammonia treatment (Figure 3.33), and observably increased from a low of 2.0?0.3 mg g -1 at 25:1 N:P ratio treatment to 3.6?0.7 mg g -1 at the 5:1 N:P ratio treatment. The overall trend in Typha tissue phosphorus concentration is similar to that of mesocosm tissue phosphorus, with the greatest slope evident in the 25:1 N:P ratio treatment. This corresponds to the P-limited N:P ratio treatment, and implies that plants will continue to accumulate phosphorus if P-limited even under toxic ammonia concentrations. Again, figures for above- and below-ground Typha phosphorus tissue concentration are not shown and trends correcpond to total Typha phosphorus tissue concentration (Table 3.13). It is of interest to note that the N:P ratio affects above-ground phosphorus concentration and that the below-ground tissue phosphorus concentration is significantly affected by ammonia alone. Total Juncus effusus tissue phosphorus concentration is significantly affected by ammonia treatment alone (Table 3.14). Mean tissue phosphorus concentrations increased obsrvably from a low of 1.8?0.3 mg g -1 at 25:1 N:P ratio treatment to 2.5?0.1 mg g -1 at the 5:1 N:P ratio treatment (Figure 3.34). The overall trend in Juncus tissue phosphorus is similar to that of the mesocosms on the whole, with the 25:1 N:P ratio exhibiting the greatest slope, and the 5:1 N:P ratio declining significantly between the 225 mgL -1 and the 300 mgL -1 ammonia treatments. Figures for Juncus above- and below-ground tissue phosphorus concentrations are not shown, but correspond well to overall Juncus tissue phosphorus concentrations (Table 3.14). Of interest to note is the significant effect of the N:P ratio in determining phosphorus concentration in below-ground biomass, an opposite trend to that of Typha. 119 0 1 2 3 4 5 0 75 150 225 300 Ammonia Concentration (mgL -1 ) T o ta l Typ h a T i s s u e Ph os ph or u s ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.33. Total Typha latifolia tissue phosphorus concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Above- and below- ground total phosphorus concentrations are not shown. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 120 Table 3.14. Tests of fixed effects on total, above-ground, and below-ground tissue phosphorus concentration of Juncus effusus in 2003. Please refer to appendix II for SAS code. Measured Biomass Effect d.f. F Value P Level Total Juncus effusus [Ammonia] 3 4.49 0.0123 Total Juncus effusus N:P Ratio 2 2.38 0.1144 Total Juncus effusus N:P Ratio x [Ammonia] 6 0.21 0.9703 Above-ground Juncus effusus [Ammonia] 3 5.1 0.0072 Above-ground Juncus effusus N:P Ratio 2 0.96 0.3961 Above-ground Juncus effusus N:P Ratio x [Ammonia] 6 0.16 0.9858 Below-ground Juncus effusus [Ammonia] 3 5.2 0.0066 Below-ground Juncus effusus N:P Ratio 2 3.83 0.0359 Below-ground Juncus effusus N:P Ratio x [Ammonia] 6 0.49 0.8127 121 0.0 1.0 2.0 3.0 4.0 5.0 0 75 150 225 300 Ammonia Concentration (mgL -1 ) T o ta l Jun c us T i s s u e Ph os ph oru s (mg g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.34. Total Juncus effusus tissue phosphorus concentration along a gradient of ammonia concentration treatments and at three N:P ratios in 2003. Above- and below- ground total phosphorus concentrations. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 122 The N:P Ratio as a Predictive Tool These analyses allow for interpretation of the vegetation N:P ratio in Typha and Juncus. Figure 3.35 illustrates the differences between mean vegetation N:P ratios in Typha and Juncus and allows comparison of each to the tissue N:P ratios predicted by Redfield (1958) and Koerselman and Meuleman (1996); the tissue N:P levels of both species tissues falls within the predicted N:P ratio for all values except very high levels of nitrogen. At high levels of nitrogen, Juncus has lower N:P ratios, and Typha has higher N:P ratios. Due to the capacity of the vegetation to uptake nutrients preferentially, the N:P ratio generally stays between the co-limited ratio of Koerselman and Meuleman (14:1; 1996) and the co-limited ratio of Redfield (7.2:1, 1958). Of interest is that, if one were to predict nutrient limitation by the Korselman and Meuleman ratio, biomass would be slightly N-limited, whereas the Redfield ratio would predict slight P limitation. Since the plants themselves show affinities for different nutrients at different concentrations, the total N:P ratio of the biomass should fall within these ratios if in fact the biomass was co-limited by N and P. Koerselman and Meuleman (1996) demonstrate that vegetation N:P ratios of <14:1 show N limitation and >16:1 show P limitation in wetland plants. Redfield (1958) suggests a vegetation N:P ratio of 7.2:1 by weight. Despite administering experimental N:P ratios of 5:1 and 25:1, Figure 3.36 clearly shows that measured vegetation N:P ratios did not reach such levels. In the case of the high nitrogen 5:1 N:P ratio treatment, the vegetation N:P ratio does appear to approach the actual treatment ratio. Only in high nitrogen Typha tissue were measured N:P ratios less than the Redfield ratio, and in very low nitrogen were measured vegetation N:P ratios more than the Koerselman-Meuleman co-limited ratio (Figure 3.37). The response of Typha to 123 N:P=7.2:1 N:P=14:1 Typha Juncus 0.0 1.0 2.0 3.0 4.0 5.0 6.0 10.0 20.0 30.0 40.0 Total Species Tissue Nitrogen (mg g -1 ) T o t a l S p e c i e s T i s s u e Ph os ph or u s ( m g g -1 ) Juncus Typha Figure 3.35. The vegetation N:P ratio of total Juncus effusus tissue and total Typha latifolia tissue presented as tissue nitrogen versus tissue phosphorus at all experimental N:P ratios and all ammonia concentration treatments in 2003. N:P ratios of 7.2:1 (Redfield 1958) and 14:1 (Koerselmen and Meuleman 1996) are provided for comparison and to show possible limitations. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 124 N:P=7.2:1 N:P=14:1 experimental N:P=5:1 experimental N:P=15:1 experimental N:P=25:1 0.0 1.0 2.0 3.0 4.0 5.0 6.0 10.0 20.0 30.0 40.0 Total Mesocosm Tissue Nitrogen (mg g -1 ) T o t a l Meso c o sm T i ssu e P h o s p h o r u s ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.36. The vegetation N:P ratio of each experimental N:P ratio presented as tissue nitrogen versus tissue phosphorus at all ammonia concentration treatments and for mesocosms of Juncus effusus and Typha latifolia in 2003. N:P ratios of 7.2:1 (Redfield 1958) and 14:1 (Koerselmen and Meuleman 1996) are provided for comparison and to show possible limitations. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 125 N:P=7.2:1 N:P=14:1 experimental N:P=5:1 experimental N:P=15:1 experimental N:P=25:1 0 1 2 3 4 5 6 10 20 30 40 Total Typha Tissue Nitrogen (mg g -1 ) T o ta l T y ph a T i s s u e P h os ph or u s ( m g g -1 ) N:P ratio=5:1 N:P ratio=15:1 N:P ratio=25:1 Figure 3.37. The vegetation N:P ratio of each experimental N:P ratio presented as tissue nitrogen versus tissue phosphorus at all ammonia concentration treatments and for Typha latifolia in 2003. N:P ratios of 7.2:1 (Redfield 1958) and 14:1 (Koerselmen and Meuleman 1996) are provided for comparison and to show possible limitations. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix II for SAS code). 126 N:P=7.2:1 N:P=14:1 experimental N:P=5:1 experimental N:P=15:1 experimental N:P=25:1 0.0 1.0 2.0 3.0 4.0 5.0 6.0 10.0 20.0 30.0 40.0 Total Juncus Tissue Nitrogen (mg g -1 ) To t a l Jun c us T i s s u e Ph os ph oru s (mg g -1 ) N:P=5:1 N:P=15:1 N:P=25:1 Figure 3.38. The vegetation N:P ratio of each experimental N:P ratio presented as tissue nitrogen versus tissue phosphorus at all ammonia concentration treatments and for Juncus effusus in 2003. N:P ratios of 7.2:1 (Redfield 1958) and 14:1 (Koerselmen and Meuleman 1996) are provided for comparison and to show possible limitations. Standard errors are not presented due to graphic difficulty in their interpretation. Due to normality of data and homogeneity of residuals from the ANOVA, log transformation was not necessary (see appendix 2 for SAS code). 127 elevated ammonia concentrations even in high phosphorus concentrations is to uptake more phosphorus than necessary. An N:P ratio of less than 5:1 with high tissue nitrogen may be an artifact of phytotoxicity in Typha and requires further study. In Juncus, vegetation N:P ratios never fell outside predicted ratios (Figure 3.38), although extrapolation of the lines for each N:P treatment could potentially reach such levels. In age-symmetric and age-asymmetric mixtures of Typha and Schenoplectus tabernaemontanai (K.C. Gmel) Palla in treatment wetlands, Svengsouk and Mitsch (2000) found N:P ratios likewise typically bounded by ratios of 7.2:1 and 14:1. When nitrogen and phosphorus were added in experimental concentrations of 3:1, only Schenoplectus tissue exhibited an N:P ratio of less than 7.2:1. Here, mesocosm N:P ratios tend to fall within the bounds of ?normal? (taken here to mean lower nitrogen levels) natural system ratios, perhaps due to the fact that adjustments to such low ratios were not made. This finding is unsupported by other authors (Koerselman and Meuleman 1996, Gusewell et al. 2003): the tissue N:P ratio is more indicative of the nutrient requirements of the species than of relative availability rates of these nutrients (Tilman 1997). The effects of nutrient supply rates on long-term community composition and biomass production may differ from the findings reported here for short-term fertilization (Tilman 1982, Chapin et al. 1997). Long-term changes in community composition may be based on changes in levels of competition following removal of nutrient deficiencies, as predicted by Tilman (1982). This is especially evident considering the extremely high nutrient concentrations used in this experiment. Most experiments regarding short- and long-term nutrient fertilization studies have been field-based in locations of oligotrophic 128 growth (see Koerselman and Meuleman 1996 for a synopsis of fertilization experiments with acceptable treatment design). Systems such as dune slacks, bogs, fens, and northern-latitude freshwater wetlands tend to have higher species diversity, lower nutrient availability, and subsequent lower production when compared to natural systems such as tidal marshes or artificial systems such as treatment wetlands. McJannet et al. (1995) indicate that no differences exist N:P ratios in plant of ?fertile? and ?infertile? ecosystems, ?fertility? was defined by a nitrogen soil concentrations in excess of 10 ppm, which is significantly lower than concentrations addressed in this study. Although many researchers have investigated the effects of nutrient additions, few report actual nutrient concentrations achieved (which is often due to fertilization with slow-release type fertilizers; McJannet et al. 1995, Gusewell et al. 2003) and few have observed the effects of nutrient additions along gradients (but see Clarke and Balwin 2002, Romero et al. 1999) as applied to the N:P ratio. Measurement of the N:P ratios in plant communities in ATWs instead of in artificially manipulated greenhouse experiments may provide increased interpretation of the role of the N:P ratio in determining biomass composition at high nitrogen concentrations. Tissue N:P ratios for each species were compared with those suggested by Koerselman and Meuleman (1996) and Redfield (1958) (Figures 3.35, 3.37, 3.38). The effect of species on determining tissue N and P concentrations, and thus the vegetation N:P ratio, is significant. This contradicts the work of some researchers. In field experiments, N:P ratios tend to differ more within species (given different sites or different nutrient concentrations) than between species at a given site (Koerselman and Meuleman 1996, Gusewell et al. 2002), or between growth forms rather than between 129 species (Aerts et al. 1999). On the other hand, in a study of 41 wetland species, McJannet et al. (1995) found significant differences in N:P ratios between species (n=5 samples of each species). Evidently, additional work is necessary to characterize the N:P ratio between species. A major component of the N:P tool is its potential for characterizing competitive relationships between species. The ?resource ratio model? (Tilman 1982, 1985, 1997) predicts that the concentration of a nutrient in plant biomass might indicate the ability of the species to compete for this nutrient. Thus, in elevated N conditions, the resource ratio model predicts that species with the higher N:P ratio will exhibit a competitive advantage because strong competitors have low concentrations of a limiting nutrient (Tilman 1982). The measured average N:P ratio of Juncus was 12.0:1; the measured average N:P of Typha was 9.1:1. As treatment N levels increased, Juncus sequestered less N in the biomass (Figure 3.20) than Typha (Figure 3.17), and slightly less P (Figure 3.34) than Typha (Figure 3.33). Thus, despite increases in ammonia concentration, the N:P ratio in Juncus remained at a level more suitable for growth as evidenced in Figure 3.32. If N and P biomass sequestering differs between species, as shown here, a real potential for more precise ATW planting schemes should be investigated. Especially in regards to the fact that high levels of ammonia tend to be limiting to plant biomass in these systems due to phytotoxic effect, those species which tend to accumulate lower amounts of N in tissue may play an important role in designing ATWs to handle high N- loading rates. Gusewell et al. (2003) state that species with higher N:P ratios would respond in the long term to additions of nitrogen, while species with lower N:P ratios would respond to additions of phosphorus, and that these predictions are based on the 130 work of Tilman?s (1982) resource ratio model. In point of fact, it appears that in extremely high ammonia, P-limited systems, species exhibiting a higher N:P ratio (Juncus) would have higher biomass than species exhibiting a low N:P ratio. Perhaps since N levels are so high in ATWs, those species which are able to exclude toxic ammonia-N from their tissues have a competitive advantage through stress tolerance. Additionally, tissue N concentrations must be compared with total N removed by each treatment condition. Of interest is that, despite higher N loading in the 300 mgL -1 treatment, less total N was removed due to lower biomass (Figure 3.23). By finding the optimal N level or by amending the N:P level, managers could achieve removal of the highest amount of N over time. Planting to achieve such function would require schemes that focus on stress tolerators such as Juncus in high ammonia concentrations or highly competitive species such as Typha under more ideal ammonia concentrations. The 15:1 N:P level accumulated the least total tissue nitrogen in the experiment; this finding is surprising considering this ratio is ?optimal? and ?co-limiting? based on the work of other researchers (Korselman and Meuleman 1996). Total tissue N over the course of the experiment was highest in the 5:1 level, where, due to an overabundance of P, macrophytes should not have been able to utilize interstitial nitrogen; the fact that high tissue N resulted under these conditions needs additional study. CONCLUSIONS Utility of the N:P ratio as a predictive tool in high-ammonia ATWs may be recommended under the conditions predicted by the model of Koerselman and Meuleman (1996). Optimal predictive power is achieved by use of a larger system, here the 131 mesocosm, in situ the ATW itself; differences in the N:P ratios based on the species themselves can likewise be recommended based on the data presented, with trends in the data possibly indicating differences in N and P usage in Juncus and in Typha. The fact that differences in N and P uptake are based on species is not in agreement with other authors, and represents an important distinction in creating ATWs. Koerselman and Meuleman (1996) found significant differences in the N:P ratio alone, not in the respective differences in the masses of N and P in the biomass tissue. McJannet et al. (1995) likewise found statistically significant differences in N:P ratio alone, as opposed to differences in the masses of N and P within 41 wetland species. Other authors only supply information regarding the results of tests on the N:P ratio (versus the Redfield ratio or t-tests between species or sites) without providing information regarding the individual terms of N uptake and P uptake sequestered in the biomass (Gusewell et al. 2003, Romero et al. 2003, Svengsouk and Mitsch 2000, Tilman 1996). This is primarily due to a lack of gradient-based, controlled bench-scale analyses on the subject. The aforementioned authors used primarily large-scale in situ testing strategies on actual functioning ATWs. While large scale applications of the N:P ratio test may prove worthwhile in the long run in determining means to effect higher denitrification rates in ATWs, issues with species specific N and P loading rates should be addressed prior to planting treatment wetlands in experiments such as this. Certainly, these systems differ from natural systems with lower amounts of nitrogen, and the results of fertilizing them with levels on orders of magnitude higher than natural systems will elicit changes in vegetation dynamics atypical of most field studies. At the same time, however, vegetation uptake of N and P must occur in both 132 types of systems for growth of plant material to occur. Differences in the way ammonia- N forces physiologic change in these species should be further addressed, and may provide additional insight as to how the N:P ratio may be used to as a management tool in ATWs to improve treatment efficiency. 133 CHAPTER FOUR SYNTHESIS High ammonia concentrations are one of a number of potential limits to vegetation colonization of ATWs. Additional limits include substrate stability, herbivory, water depth, and the potential effects of other phytotoxic compounds such as sulfates, nitrates, or soluble metals such as iron, copper, or aluminum (Surrency 1993, Cronk 1996, Clarke 1999). However, the exceptionally high levels of nitrogen, especially those found in ammonia form, have been cited or inferred frequently as the most limiting phytotoxic element in ATWs (Wang 1991, Surrency 1993, Hammer and Knight 1994, Dijk and Eck 1995, Cronk 1996, Kadlec and Knight 1996, Verhoeven et al. 1996, Hill et al. 1997, Bobbink et al. 1998, Humenik et al. 1999, Aerts and Bobbink 1999, Clarke 1999, Clarke and Baldwin 2002). Graded tests of ammonia toxicity have dominated the literature on this subject. It is important to note, however, that nitrogen plays an important role as a plant nutrient as well, and therefore the fate of ammonia in ATWs may be of multiple paths (Mitsch and Gosselink 2003): it may be transformed to nitrates, nitrogen gases or bacterial biomass, it may be transformed to plant biomass, or it may exit the system as ammonia. Experimental ratios of the percentage ammonia removed by the former processes are about 60% removal by bacterial action, and about 40% removal by plant uptake (Howes et al. 1981, Gumbricht 1993, Bachand and Horne 2000b). Therefore, while it is important to achieve optimal spatial distribution of redoxic conditions for removal of nitrogen via the nitrification-denitrification pathway (and plant 134 biomass plays important roles in providing oxygen to the root zone and in providing a substrate upon which the reaction may occur), it is likewise important to maintain optimal growth of macrophytes in constructed wetlands. In natural systems, plant growth tends to be limited by nitrogen, phosphorus, and to a lesser extent, potassium (Mitsch and Gosselink 1993, Taiz and Zeiger 2002). Nitrogen is rarely limiting in ATWs, and ammonia concentrations ranging to 500 mgL -1 have been reported (Hammer 1992). Potassium is likewise seldom limiting due to its prevalence in mineral soils. To increase biomass production, then, it would seem ideal to increase phosphorus concentrations to approach ideal co-limited natural ratios, such as those proposed by Redfield (1958) or Koerselman and Meuleman (1996). This formed the basis of the exploration conducted in this thesis. Although, in this study, differences in biomass production along an ammonia gradient with N:P ratio manipulation failed to produce reliably significant results, certain conclusions regarding the effects of high ammonia treatments and the N:P ratio on biomass may be made from the studies conducted. In the initial study, where no modification of the N:P ratio was conducted, convex curves indicating the response of both mesocosms and individual species to increasing ammonia concentrations indicates that ammonia has a fertilization effect at low to moderate concentrations, and an inhibitory effect at high concentrations (Figures 2.3, 2.6, and 2.9). The maximum of total biomass production for a mesocosm of Juncus and Typha planted together occurred at between 75 mgL -1 and 150 mgL -1 , with the patterns of treatment effect of above- and below-ground biomass being similar. Additions of phosphate to create co-limited systems did not result in a statistically significant change in biomass production due to 135 the low sample size of N:P ratio treatments within ammonia treatments (n=3), but served to illustrate a pattern of highest production at phytotoxic nitrogen concentrations at an N:P level of 15:1 (Figures 3.2, 3.5 and 3.8). By planting these species in mesocosms, a better idea of field-level ATW functioning was achieved by noting effects of the system as a whole; an ammonia regime of less than 150 mgL -1 is suggested due to impaired system functioning above that level. When applied to mesocosms of Typha and Juncus planted together, ammonia concentrations forced a shift in biomass towards dominance by Juncus at all treatment concentrations. This indication seems novel; dry-matter production by Typha tends to be significantly higher than other species. In fact, in most cases, Typha tends to be assumed as a highly competitive species (Grace and Wetzel 1981, Tilman 1987, Svengsouk and Mitsch 2000). Despite the warnings of ATW planting literature not to plant monocultures of cattails (SCS 1991, Kadlec and Knight 1996), here Typha was seriously out-produced by Juncus at all treatment concentrations. The fact that Juncus was less inhibited at higher concentrations is in accordance with other researchers (Hill et al. 1997, Clarke 1999, Emery et al. 2001., Clarke and Baldwin 2002); the interactive effects of phytotoxic gradients of ammonia and N:P ratios on species planted in polyculture have yet to be tested in great detail. Work on polycultures in treatment wetlands has focused on the effects of additions of single levels of N, P and co-limiting N and P, or on the effects of asymmetric-density plantings on the dominance of Typha (Svengsouk and Mitsch 2000). These findings may serve as a bench-level test of the trade-off between competitive performance and stress tolerance noted elsewhere (Grace 1981, Clarke 1999, Emery et al. 2001, Clarke and Baldwin 2002); in cases where stress tolerance is essential 136 in supporting biomass production, species which are highly competitive will perform poorly. With regard to management issues, planting Juncus in ATWs known to have high nitrogen loading may be preferable for sustained biomass production. Statistically significant results for biomass production or tissue nitrogen concentration were shown only in a few instances by altering the N:P ratio for either species. There is some indication based on the pattern of biomass production that Juncus and Typha will both produce more biomass when N:P ratios are adjusted to co-limitation, especially under higher ammonia treatments (Figure 3.2). However, there is also evidence that tissue nitrogen concentrations will decline if a 15:1 N:P ratio is used (Figure 3.14). The fact that non-significant results were shown for alterations in the N:P ratio in these species is of special import due to the significance effect of species on both N uptake and P uptake. Juncus stores less nitrogen in the tissue, and therefore has a higher N:P ratio at 12.0 than Typha at 9.1. This is evident in the morphology of the two species as well; with a large rootstock, Typha stores nutrients in the tissue for times of resource shortage, whereas Juncus does not. Although other authors have noted that differences in tissue N:P ratios do not indicate dominance, these studies have tended to occur in low- nutrient environments. In ATWs that are dominated by high nitrogen concentrations, those species which can sequester less N in the tissue may tolerate concentrations which other species which cannot. The management implication in ATWs receiving high N inputs may be to plant species without large rootstocks (such as Juncus, with its smaller corm), thereby minimizing storage and accumulation of phytotoxic levels of tissue nitrogen. 137 Typically, N:P ratios tend to differ between species, whereas N uptake and P uptake are dependant on differences in major phenotypic groups or ecological boundaries (McJannet et al. 1995, Gusewell et al. 2003). In this study, N:P ratios did differ based on species. Nitrogen and phosphorus uptake was dependant on species as well, despite the ecological and phenotypic type similarity of the species used. Both species occur in shallow water as emergent wetland clonal graminoids. In systems of high stress, perhaps it is the ability to exclude a phytotoxic chemical than to uptake a limiting one that sustains growth. It would be of interest to test the N:P ratios and N and P uptake parameters of field-scale ATWs to determine whether this trend is consistent. The N:P ratio has been used in the past to predict nutrient limitations as according to Koerselman and Meuleman (1996); N:P ratios over 16:1 indicate P limitation, and under 14:1 indicate N limitation. In this case, most N:P ratios fell under 14:1, indicating N limitation. The implications of this are twofold. Because N:P ratios were below 14:1 for most treatments, even in experimental abundance of N (at N:P ratios of 25:1), a problem with nitrogen uptake is to blame. This physiologic response indicates an impairment of the mechanism for nitrogen uptake at phytotoxic N concentrations. The down-regulation of ammonia-N uptake in this case is significant in that this is the first occasion in which it is noted; there is an upper limit to the amount of nitrogen able to be fixed in the biomass regardless of availability of other co-limiting nutrients. Secondly, fertilization effects of phosphorus do not alleviate phytotoxic N regimes, because of limitations in the amount of nitrogen that can be moved through plant biomass. This infers that the phytotoxicity of ammonia-N to plants is due to absolute concentrations and is not able to be manipulated by experimental methods, such as by increasing the 138 availability of phosphorus. Implications of these findings are such that managers should limit the concentration of ammonia-N in the system at all times, and that potential phytotoxic effects are independent of concentrations of co-limiting nutrients. The utility of the N:P ratio in determining nutrient limitations in ATWs is therefore suspect due to high nitrogen concentrations endemic to such systems. Koerselman and Meuleman (1996) and Gusewell et al. (2003) indicate potential limitations to the use of the N:P ratio in determining nutrient availability in systems due to differences in within-species N:P ratios. Tables 3.6 and 3.12 show this species effect on tissue nitrogen concentration and tissue phosphorus concentration, respectively. Measured tissue nitrogen and tissue phosphorus concentrations corresponded primarily with ammonia concentration and species, and to a lesser extent the experimental N:P ratios. Individual species N uptake and P uptake differ significantly, as do differences in the overall N:P ratio between species. This implies that in field-scale research, biomass from a particular ATW may not be pooled to determine environmental limitations. In management implications, however, the absolute concentration of N probably affects biomass to a greater degree than nutrient limitations. Tests of the ecological and environmental aspects of agricultural treatment wetlands are a new trend in biological research. While appropriate engineering of the structural and fluid properties of ATWs is of great importance, a major ?black box? in ATW design is the impact of plants to the system. The effects of plants towards increasing nitrification rates (and subsequently denitrification rates due to coupled reactions) by increasing substrate oxidation are well known, but management decisions necessary to maintain and increase biomass production in species found in ATWs must be based on research as to 139 the limiting conditions imposed by chemical parameters of ATWs. The adverse effects of high ammonia levels are but a single example of the difficulties imposed by extremely eutrophic settings; eutrophication on a fertilization level in natural systems tends to inhibit processes seen as environmentally valuable. Field scale experimentation with the elevated ammonia concentrations typical of ATWs should be the next step in understanding the effects of high levels of nutrients on the biomass and functioning of wetland macrophytes. 140 APPENDIX I SAS CODE FOR CHAPTER II STATISTICAL ANALYSES quit; options ls=80 ps=50 center nodate pageno=1; title1 2002 data; data total; input column$ row$ ammonia$ species$ type$ biomass; datalines; A 1 300 TY a 24.2 A 1 300 TY b 1.2 A 1 300 TY tot 25.4 A 1 300 JU a 157.4 A 1 300 JU b 193.6 A 1 300 JU tot 351 A 1 300 tot tot 376.4 A 1 300 tot b 194.8 A 1 300 tot a 181.6 A 2 150 TY a 96 A 2 150 TY b 90 A 2 150 TY tot 186 A 2 150 JU a 154.2 A 2 150 JU b 92.3 A 2 150 JU tot 246.5 A 2 150 tot tot 432.5 A 2 150 tot b 182.3 A 2 150 tot a 250.2 A 3 225 TY a 105 A 3 225 TY b 68.6 A 3 225 TY tot 173.6 A 3 225 JU a 137.3 A 3 225 JU b 86.8 A 3 225 JU tot 224.1 A 3 225 tot tot 397.7 A 3 225 tot b 155.4 A 3 225 tot a 242.3 A 4 0 TY a 8.3 A 4 0 TY b 33.5 A 4 0 TY tot 41.8 A 4 0 JU a 78.5 A 4 0 JU b 85.6 A 4 0 JU tot 164.1 A 4 0 tot tot 205.9 A 4 0 tot b 119.1 A 4 0 tot a 86.8 A 5 75 TY a 131.2 141 A 5 75 TY b 156.9 A 5 75 TY tot 288.1 A 5 75 JU a 201 A 5 75 JU b 120.7 A 5 75 JU tot 321.7 A 5 75 tot tot 609.8 A 5 75 tot b 277.6 A 5 75 tot a 332.2 A 6 225 TY a 132.6 A 6 225 TY b 96.6 A 6 225 TY tot 229.2 A 6 225 JU a 185.2 A 6 225 JU b 228.3 A 6 225 JU tot 413.5 A 6 225 tot tot 642.7 A 6 225 tot b 324.9 A 6 225 tot a 317.8 A 7 75 TY a 85.2 A 7 75 TY b 85.9 A 7 75 TY tot 171.1 A 7 75 JU a 312.8 A 7 75 JU b 199.9 A 7 75 JU tot 512.7 A 7 75 tot tot 683.8 A 7 75 tot b 285.8 A 7 75 tot a 398 A 8 300 TY a 40.4 A 8 300 TY b 9.4 A 8 300 TY tot 49.8 A 8 300 JU a 208.1 A 8 300 JU b 143.1 A 8 300 JU tot 351.2 A 8 300 tot tot 401 A 8 300 tot b 152.5 A 8 300 tot a 248.5 A 9 0 TY a 52.2 A 9 0 TY b 88 A 9 0 TY tot 140.2 A 9 0 JU a 156 A 9 0 JU b 139.3 A 9 0 JU tot 295.3 A 9 0 tot tot 435.5 A 9 0 tot b 227.3 A 9 0 tot a 208.2 A 10 150 TY a 149.4 A 10 150 TY b 152.8 A 10 150 TY tot 302.2 142 A 10 150 JU a 214.3 A 10 150 JU b 248.6 A 10 150 JU tot 462.9 A 10 150 tot tot 765.1 A 10 150 tot b 401.4 A 10 150 tot a 363.7 B 1 150 TY a 152.8 B 1 150 TY b 78.7 B 1 150 TY tot 231.5 B 1 150 JU a 126.4 B 1 150 JU b 79.9 B 1 150 JU tot 206.3 B 1 150 tot tot 437.8 B 1 150 tot b 158.6 B 1 150 tot a 279.2 B 2 75 TY a 65.1 B 2 75 TY b 118.8 B 2 75 TY tot 183.9 B 2 75 JU a 212.5 B 2 75 JU b 133.8 B 2 75 JU tot 346.3 B 2 75 tot tot 530.2 B 2 75 tot b 252.6 B 2 75 tot a 277.6 B 3 0 TY a 16.4 B 3 0 TY b 73.3 B 3 0 TY tot 89.7 B 3 0 JU a 79.9 B 3 0 JU b 80.5 B 3 0 JU tot 160.4 B 3 0 tot tot 250.1 B 3 0 tot b 153.8 B 3 0 tot a 96.3 B 4 300 TY a 75.5 B 4 300 TY b 34.7 B 4 300 TY tot 110.2 B 4 300 JU a 113 B 4 300 JU b 62.5 B 4 300 JU tot 175.5 B 4 300 tot tot 285.7 B 4 300 tot b 97.2 B 4 300 tot a 188.5 B 5 225 TY a 91.9 B 5 225 TY b 32.7 B 5 225 TY tot 124.6 B 5 225 JU a 179.2 B 5 225 JU b 73.7 143 B 5 225 JU tot 252.9 B 5 225 tot tot 377.5 B 5 225 tot b 106.4 B 5 225 tot a 271.1 B 6 300 TY a 34.8 B 6 300 TY b 16.9 B 6 300 TY tot 51.7 B 6 300 JU a 232.1 B 6 300 JU b 92.2 B 6 300 JU tot 324.3 B 6 300 tot tot 376 B 6 300 tot b 109.1 B 6 300 tot a 266.9 B 7 0 TY a 9.7 B 7 0 TY b 73.5 B 7 0 TY tot 83.2 B 7 0 JU a 65.2 B 7 0 JU b 91.5 B 7 0 JU tot 156.7 B 7 0 tot tot 239.9 B 7 0 tot b 165 B 7 0 tot a 74.9 B 8 150 TY a 148 B 8 150 TY b 126.3 B 8 150 TY tot 274.3 B 8 150 JU a 197.1 B 8 150 JU b 107.3 B 8 150 JU tot 304.4 B 8 150 tot tot 578.7 B 8 150 tot b 233.6 B 8 150 tot a 345.1 B 9 75 TY a 43.3 B 9 75 TY b 76.8 B 9 75 TY tot 120.1 B 9 75 JU a 237.6 B 9 75 JU b 131.7 B 9 75 JU tot 369.3 B 9 75 tot tot 489.4 B 9 75 tot b 208.5 B 9 75 tot a 280.9 B 10 225 TY a 136.7 B 10 225 TY b 55 B 10 225 TY tot 191.7 B 10 225 JU a 148.4 B 10 225 JU b 90.2 B 10 225 JU tot 238.6 B 10 225 tot tot 430.3 144 B 10 225 tot b 145.2 B 10 225 tot a 285.1 C 1 75 TY a 138.1 C 1 75 TY b 164.5 C 1 75 TY tot 302.6 C 1 75 JU a 122.6 C 1 75 JU b 46.9 C 1 75 JU tot 169.5 C 1 75 tot tot 472.1 C 1 75 tot b 211.4 C 1 75 tot a 260.7 C 2 300 TY a 16.8 C 2 300 TY b 9.8 C 2 300 TY tot 26.6 C 2 300 JU a 178.2 C 2 300 JU b 66.4 C 2 300 JU tot 244.6 C 2 300 tot tot 271.2 C 2 300 tot b 76.2 C 2 300 tot a 195 C 3 150 TY a 101.4 C 3 150 TY b 106.1 C 3 150 TY tot 207.5 C 3 150 JU a 185.2 C 3 150 JU b 62 C 3 150 JU tot 247.2 C 3 150 tot tot 454.7 C 3 150 tot b 168.1 C 3 150 tot a 286.6 C 4 225 TY a 107.7 C 4 225 TY b 41 C 4 225 TY tot 148.7 C 4 225 JU a 182.1 C 4 225 JU b 65.6 C 4 225 JU tot 247.7 C 4 225 tot tot 396.4 C 4 225 tot b 106.6 C 4 225 tot a 289.8 C 5 0 TY a 4.1 C 5 0 TY b 21.5 C 5 0 TY tot 25.6 C 5 0 JU a 75.5 C 5 0 JU b 70 C 5 0 JU tot 145.5 C 5 0 tot tot 171.1 C 5 0 tot b 91.5 C 5 0 tot a 79.6 145 C 6 75 TY a 81.6 C 6 75 TY b 89.2 C 6 75 TY tot 170.8 C 6 75 JU a 259 C 6 75 JU b 123.8 C 6 75 JU tot 382.8 C 6 75 tot tot 553.6 C 6 75 tot b 213 C 6 75 tot a 340.6 C 7 150 TY a 98.1 C 7 150 TY b 61 C 7 150 TY tot 159.1 C 7 150 JU a 215.9 C 7 150 JU b 102.4 C 7 150 JU tot 318.3 C 7 150 tot tot 477.4 C 7 150 tot b 163.4 C 7 150 tot a 314 C 8 225 TY a 80.4 C 8 225 TY b 53.8 C 8 225 TY tot 134.2 C 8 225 JU a 190 C 8 225 JU b 81.2 C 8 225 JU tot 271.2 C 8 225 tot tot 405.4 C 8 225 tot b 135 C 8 225 tot a 270.4 C 9 300 TY a 25.8 C 9 300 TY b 69.3 C 9 300 TY tot 95.1 C 9 300 JU a 166.7 C 9 300 JU b 6.9 C 9 300 JU tot 173.6 C 9 300 tot tot 268.7 C 9 300 tot b 76.2 C 9 300 tot a 192.5 C 10 0 TY a 29 C 10 0 TY b 94.7 C 10 0 TY tot 123.7 C 10 0 JU a 62.7 C 10 0 JU b 75.4 C 10 0 JU tot 138.1 C 10 0 tot tot 261.8 C 10 0 tot b 170.1 C 10 0 tot a 91.7 D 1 225 TY a 41.8 D 1 225 TY b 9.3 146 D 1 225 TY tot 51.1 D 1 225 JU a 159.9 D 1 225 JU b 55.4 D 1 225 JU tot 215.3 D 1 225 tot tot 266.4 D 1 225 tot b 64.7 D 1 225 tot a 201.7 D 2 0 TY a 21.1 D 2 0 TY b 21.7 D 2 0 TY tot 42.8 D 2 0 JU a 63.9 D 2 0 JU b 74.8 D 2 0 JU tot 138.7 D 2 0 tot tot 181.5 D 2 0 tot b 96.5 D 2 0 tot a 85 D 3 75 TY a 52.5 D 3 75 TY b 59.5 D 3 75 TY tot 112 D 3 75 JU a 224.5 D 3 75 JU b 64.4 D 3 75 JU tot 288.9 D 3 75 tot tot 400.9 D 3 75 tot b 123.9 D 3 75 tot a 277 D 4 150 TY a 106 D 4 150 TY b 91.7 D 4 150 TY tot 197.7 D 4 150 JU a 137.4 D 4 150 JU b 95.9 D 4 150 JU tot 233.3 D 4 150 tot tot 431 D 4 150 tot b 187.6 D 4 150 tot a 243.4 D 5 300 TY a 6.6 D 5 300 TY b 1.2 D 5 300 TY tot 7.8 D 5 300 JU a 201.2 D 5 300 JU b 82.7 D 5 300 JU tot 283.9 D 5 300 tot tot 291.7 D 5 300 tot b 83.9 D 5 300 tot a 207.8 D 6 150 TY a 113 D 6 150 TY b 112.5 D 6 150 TY tot 225.5 D 6 150 JU a 120.9 147 D 6 150 JU b 100.3 D 6 150 JU tot 221.2 D 6 150 tot tot 446.7 D 6 150 tot b 212.8 D 6 150 tot a 233.9 D 7 300 TY a 16.9 D 7 300 TY b 5.2 D 7 300 TY tot 22.1 D 7 300 JU a 185.9 D 7 300 JU b 83.7 D 7 300 JU tot 269.6 D 7 300 tot tot 291.7 D 7 300 tot b 88.9 D 7 300 tot a 202.8 D 8 0 TY a 32.3 D 8 0 TY b 34.7 D 8 0 TY tot 67 D 8 0 JU a 30.1 D 8 0 JU b 41.7 D 8 0 JU tot 71.8 D 8 0 tot tot 138.8 D 8 0 tot b 76.4 D 8 0 tot a 62.4 D 9 225 TY a 76.7 D 9 225 TY b 49.1 D 9 225 TY tot 125.8 D 9 225 JU a 189.2 D 9 225 JU b 65.3 D 9 225 JU tot 254.5 D 9 225 tot tot 380.3 D 9 225 tot b 114.4 D 9 225 tot a 265.9 D 10 75 TY a 6.8 D 10 75 TY b 5.3 D 10 75 TY tot 12.1 D 10 75 JU a 296.8 D 10 75 JU b 197.9 D 10 75 JU tot 494.7 D 10 75 tot tot 506.8 D 10 75 tot b 203.2 D 10 75 tot a 303.6 E 1 0 TY a 10 E 1 0 TY b 38.9 E 1 0 TY tot 48.9 E 1 0 JU a 63.4 E 1 0 JU b 83.3 E 1 0 JU tot 146.7 148 E 1 0 tot tot 195.6 E 1 0 tot b 122.2 E 1 0 tot a 73.4 E 2 225 TY a 0.3 E 2 225 TY b 2.6 E 2 225 TY tot 2.9 E 2 225 JU a 224.1 E 2 225 JU b 95.2 E 2 225 JU tot 319.3 E 2 225 tot tot 322.2 E 2 225 tot b 97.8 E 2 225 tot a 224.4 E 3 300 TY a 40.5 E 3 300 TY b 28.9 E 3 300 TY tot 69.4 E 3 300 JU a 172.1 E 3 300 JU b 131.4 E 3 300 JU tot 303.5 E 3 300 tot tot 372.9 E 3 300 tot b 160.3 E 3 300 tot a 212.6 E 4 75 TY a 21.9 E 4 75 TY b 15.8 E 4 75 TY tot 37.7 E 4 75 JU a 274 E 4 75 JU b 41.5 E 4 75 JU tot 315.5 E 4 75 tot tot 353.2 E 4 75 tot b 57.3 E 4 75 tot a 295.9 E 5 150 TY a 125.4 E 5 150 TY b 136.7 E 5 150 TY tot 262.1 E 5 150 JU a 104.7 E 5 150 JU b 39.7 E 5 150 JU tot 144.4 E 5 150 tot tot 406.5 E 5 150 tot b 176.4 E 5 150 tot a 230.1 E 6 0 TY a 18.8 E 6 0 TY b 126.9 E 6 0 TY tot 145.7 E 6 0 JU a 32.2 E 6 0 JU b 36.2 E 6 0 JU tot 68.4 E 6 0 tot tot 214.1 E 6 0 tot b 163.1 149 E 6 0 tot a 51 E 7 225 TY a 35.4 E 7 225 TY b 18.1 E 7 225 TY tot 53.5 E 7 225 JU a 240.1 E 7 225 JU b 120.3 E 7 225 JU tot 360.4 E 7 225 tot tot 413.9 E 7 225 tot b 138.4 E 7 225 tot a 275.5 E 8 75 TY a 161.8 E 8 75 TY b 161.4 E 8 75 TY tot 323.2 E 8 75 JU a 89.8 E 8 75 JU b 69.5 E 8 75 JU tot 159.3 E 8 75 tot tot 482.5 E 8 75 tot b 230.9 E 8 75 tot a 251.6 E 9 150 TY a 70.5 E 9 150 TY b 111.9 E 9 150 TY tot 182.4 E 9 150 JU a 147.2 E 9 150 JU b 47.9 E 9 150 JU tot 195.1 E 9 150 tot tot 377.5 E 9 150 tot b 159.8 E 9 150 tot a 217.7 E 10 300 TY a 24.6 E 10 300 TY b 3.3 E 10 300 TY tot 27.9 E 10 300 JU a 101.8 E 10 300 JU b 123.1 E 10 300 JU tot 224.9 E 10 300 tot tot 252.8 E 10 300 tot b 126.4 E 10 300 tot a 126.4 run; title2 biomass ANOVAs in 2002; title3 total biomass for both species in 2002 ANOVA; data total_biomass; set total; if species="tot" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=total_biomass; 150 class ammonia column species; model biomass=ammonia|species column/ddfm=kr outp=resids; *lsmeans ammonia|species / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_biomass normal plot; var biomass; run; *test shows a Shapiro-Wilk's Test (as S-W test from now on) of 0.97, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; *********************************************************** ******************************; title3 aboveground biomass for both species in 2002 ANOVA; data total_above; set total; if species="tot" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=total_above; class ammonia column species; model biomass=ammonia|species column/ddfm=kr outp=resids; *lsmeans ammonia|species / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_above normal plot; var biomass; run; *test shows a S-W test of 0.96, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; *********************************************************** ******************************; title3 belowground biomass for both species in 2002 ANOVA; data total_below; set total; if species="tot" then delete; 151 if type="tot" then delete; if type="a" then delete; run; proc mixed data=total_below; class ammonia column species; model biomass=ammonia|species column/ddfm=kr outp=resids; lsmeans ammonia|species / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_below normal plot; var biomass; run; *test shows a S-W test of 0.96, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; *********************************************************** ******************************; title3 total biomass for TYPHA in 2002 ANOVA; data typha_total; set total; if species="tot" then delete; if species="JU" then delete; if type="b" then delete; if type="a" then delete; run; proc mixed data=typha_total; class ammonia column; model biomass=ammonia column/ddfm=kr outp=resids; *lsmeans ammonia / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_total normal plot; var biomass; run; *test shows a S-W test of 0.94, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; *********************************************************** ******************************; 152 title3 aboveground biomass for TYPHA in 2002 ANOVA; data typha_above; set total; if species="tot" then delete; if species="JU" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=typha_above; class ammonia column; model biomass=ammonia column/ddfm=kr outp=resids; *lsmeans ammonia / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_above normal plot; var biomass; run; *test shows a S-W test of 0.92, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; *********************************************************** ******************************; title3 belowground biomass for TYPHA in 2002 ANOVA; data typha_below; set total; if species="tot" then delete; if species="JU" then delete; if type="tot" then delete; if type="a" then delete; run; proc mixed data=typha_below; class ammonia column; model biomass=ammonia column/ddfm=kr outp=resids; *lsmeans ammonia / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_below normal plot; var biomass; run; *test shows a S-W test of 0.94, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; 153 quit; *shows decent and acceptable homogeneity of resids; *********************************************************** ******************************; title3 total biomass for JUNCUS in 2002 ANOVA; data juncus_total; set total; if species="tot" then delete; if species="TY" then delete; if type="b" then delete; if type="a" then delete; run;proc mixed data=juncus_total; class ammonia column; model biomass=ammonia column/ddfm=kr outp=resids; *lsmeans ammonia / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_total normal plot; var biomass; run; *test shows a S-W test of 0.97, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; *********************************************************** ******************************; title3 aboveground biomass for JUNCUS in 2002 ANOVA; data juncus_above; set total; if species="tot" then delete; if species="TY" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=juncus_above; class ammonia column; model biomass=ammonia column/ddfm=kr outp=resids; *lsmeans ammonia / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_above normal plot; var biomass; run; 154 *test shows a S-W test of 0.98, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; *********************************************************** ******************************; title3 belowground biomass for JUNCUS in 2002 ANOVA; data juncus_below; set total; if species="tot" then delete; if species="TY" then delete; if type="tot" then delete; if type="a" then delete; run;proc mixed data=juncus_below; class ammonia column; model biomass=ammonia column/ddfm=kr outp=resids; *lsmeans ammonia / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_below normal plot; var biomass; run; *test shows a S-W test of 0.96, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; *********************************************************** ********************** Here ends the biomass ANOVA series............................................... *********************************************************** *********************; title2 root-to-shoot ratio ANOVAs in 2002; data rs_ratio; input column$ row$ ammonia$ species$ r_s; log_r_s=log(r_s); datalines; 155 A 4 0 t 4.036144578 A 9 0 t 1.685823755 B 3 0 t 4.469512195 B 7 0 t 7.577319588 C 5 0 t 5.243902439 C 10 0 t 3.265517241 D 2 0 t 1.028436019 D 8 0 t 1.074303406 E 1 0 t 3.89 E 6 0 t 6.75 A 5 75 t 1.195884146 A 7 75 t 1.008215962 B 2 75 t 1.824884793 B 9 75 t 1.773672055 C 1 75 t 1.191165822 C 6 75 t 1.093137255 D 3 75 t 1.133333333 D 10 75 t 0.779411765 E 4 75 t 0.721461187 E 8 75 t 0.997527812 A 2 150 t 0.9375 A 10 150 t 1.022757697 B 1 150 t 0.515052356 B 8 150 t 0.853378378 C 3 150 t 1.046351085 C 7 150 t 0.621814475 D 4 150 t 0.86509434 D 6 150 t 0.995575221 E 5 150 t 1.090111643 E 9 150 t 1.587234043 A 3 225 t 0.653333333 A 6 225 t 0.728506787 B 5 225 t 0.355821545 B 10 225 t 0.402340892 C 4 225 t 0.380687094 C 8 225 t 0.669154229 D 1 225 t 0.222488038 D 9 225 t 0.640156454 E 2 225 t 7.789095267 E 7 225 t 0.511299435 A 1 300 t 0.049586777 A 8 300 t 0.232673267 B 4 300 t 0.459602649 B 6 300 t 0.485632184 C 2 300 t 0.583333333 C 9 300 t 2.686046512 D 5 300 t 0.181818182 156 D 7 300 t 0.307692308 E 3 300 t 0.713580247 E 10 300 t 0.134146341 A 4 0 j 1.09044586 A 9 0 j 0.892948718 B 3 0 j 1.007509387 B 7 0 j 1.403374233 C 5 0 j 0.927152318 C 10 0 j 1.202551834 D 2 0 j 1.17057903 D 8 0 j 1.38538206 E 1 0 j 1.313880126 E 6 0 j 1.124223602 A 5 75 j 0.600497512 A 7 75 j 0.639066496 B 2 75 j 0.629647059 B 9 75 j 0.554292929 C 1 75 j 0.382544861 C 6 75 j 0.477992278 D 3 75 j 0.286859688 D 10 75 j 0.666778976 E 4 75 j 0.151459854 E 8 75 j 0.773942094 A 2 150 j 0.598573281 A 10 150 j 1.160055996 B 1 150 j 0.632120253 B 8 150 j 0.544393709 C 3 150 j 0.334773218 C 7 150 j 0.474293654 D 4 150 j 0.697962154 D 6 150 j 0.829611249 E 5 150 j 0.379178606 E 9 150 j 0.325407609 A 3 225 j 0.63219228 A 6 225 j 1.232721382 B 5 225 j 0.411272321 B 10 225 j 0.607816712 C 4 225 j 0.360241625 C 8 225 j 0.427368421 D 1 225 j 0.346466542 D 9 225 j 0.345137421 E 2 225 j 0.424810353 E 7 225 j 0.501041233 A 1 300 j 1.229987294 A 8 300 j 0.687650168 B 4 300 j 0.553097345 B 6 300 j 0.397242568 157 C 2 300 j 0.372615039 C 9 300 j 0.041391722 D 5 300 j 0.411033797 D 7 300 j 0.450242066 E 3 300 j 0.763509587 E 10 300 j 1.209233792 run; title3 root-to-shoot ratios for mesocosms in 2002 ANOVA; proc mixed data=rs_ratio; class ammonia column species; model log_r_s=ammonia|species column / ddfm=kr outp=resids; *lsmeans ammonia|species / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=rs_ratio normal plot; where species='t'; var log_r_s; run; *test shows a S-W test of 0.97, indicating acceptable normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogenity of resids after log-transform; *********************************************************** ******************************; title3 root-to-shoot ratios for TYPHA in 2002 ANOVA; proc mixed data=rs_ratio; where species='t'; class ammonia column; model log_r_s=ammonia column/ ddfm=kr outp=resids; *lsmeans ammonia / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=rs_ratio normal plot; where species='t'; var log_r_s; run; *test shows a S-W test of 0.97, indicating acceptable normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; 158 *********************************************************** ******************************; title3 root-to-shoot ratio for JUNCUS in 2002 ANOVA; proc mixed data=rs_ratio; where species='j'; class ammonia column; model r_s=ammonia column/ ddfm=kr outp=resids; *lsmeans ammonia / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=rs_ratio normal plot; where species='j'; var r_s; run; *test shows a S-W test of 0.93, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; quit; 159 APPENDIX II SAS CODE FOR 2003 EXPERIMENT quit; options ls=80 ps=50 center nodate pageno=1; title1 2003 data; data total; input column$ row$ ammonia$ n_p$ species$ type$ biomass n_conc p_conc; log_biomass=log(biomass); log_log_biomass=log(log_biomass); log_n_conc=log(n_conc); log_p_conc=log(p_conc); n_mass=n_conc*biomass; Datalines; A 1 300 5 ju a 24.962 1.8905865 130.4 A 1 300 5 ju b 26.082 1.8905865 48.8 A 1 300 5 ju tot 25.522 1.8905865 179.2 A 1 300 5 tot a 25.4485 2.9186226 160.7 A 1 300 5 tot b 26.165 2.9186226 59.5 A 1 300 5 tot tot 25.80675 2.9186226 220.2 A 1 300 5 ty a 25.935 3.946658 30.3 A 1 300 5 ty b 26.248 3.946658 10.7 A 1 300 5 ty tot 26.0915 3.946658 41 A 2 300 15 ju a 26.307 2.0807142 150.8 A 2 300 15 ju b 26.387 2.0807142 49.4 A 2 300 15 ju tot 26.347 2.0807142 200.2 A 2 300 15 tot a 28.3145 3.2900522 167.9 A 2 300 15 tot b 27.8415 3.2900522 55.2 A 2 300 15 tot tot 28.078 3.2900522 223.1 A 2 300 15 ty a 30.322 4.4993902 17.1 A 2 300 15 ty b 29.296 4.4993905 5.8 A 2 300 15 ty tot 29.809 4.4993905 22.9 A 3 300 25 ju a 24.308 1.6025775 162.7 A 3 300 25 ju b 24.786 1.6025475 60.3 A 3 300 25 ju tot 24.547 1.6027475 223 A 3 300 25 tot a 26.2175 2.4452625 176.2 A 3 300 25 tot b 26.156 2.4224525 66 A 3 300 25 tot tot 26.18675 2.4222625 242.2 A 3 300 25 ty a 28.127 3.23305 13.5 A 3 300 25 ty b 27.526 3.423305 5.7 A 3 300 25 ty tot 27.8265 3.2423305 19.2 A 4 300 5 ju a 24.748 1.9672195 154.5 A 4 300 5 ju b 24.82 1.9672195 41.9 A 4 300 5 ju tot 24.784 1.9672195 196.4 A 4 300 5 tot a 24.523 2.6243495 195.2 A 4 300 5 tot b 24.3965 2.6243495 57.6 A 4 300 5 tot tot 24.45975 2.6243495 252.8 A 4 300 5 ty a 24.298 3.2814795 40.7 A 4 300 5 ty b 23.973 3.2814795 15.7 A 4 300 5 ty tot 24.1355 3.2814795 56.4 A 5 300 15 ju a 24.256 1.9063435 147.3 160 A 5 300 15 ju b 25.073 1.9063435 65.3 A 5 300 15 ju tot 24.6645 1.9063435 212.6 A 5 300 15 tot a . . 168.4 A 5 300 15 tot b . . 86 A 5 300 15 tot tot . . 254.4 A 5 300 15 ty a . . 21.1 A 5 300 15 ty b . . 20.7 A 5 300 15 ty tot . . 41.8 A 6 300 25 ju a 22.976 3.0310325 140.8 A 6 300 25 ju b 24.111 3.0310825 58.7 A 6 300 25 ju tot 23.5435 3.0313825 199.5 A 6 300 25 tot a 25.302 3.1684375 157.5 A 6 300 25 tot b 26.969 3.1364375 65 A 6 300 25 tot tot 26.1355 3.1684375 222.5 A 6 300 25 ty a 27.628 3.2423305 16.7 A 6 300 25 ty b 29.827 3.2423305 6.3 A 6 300 25 ty tot 28.7275 3.2423305 23 A 7 300 5 ju a 25.741 2.5417935 111.7 A 7 300 5 ju b 24.698 2.5417935 44.3 A 7 300 5 ju tot 25.2195 2.5417935 156 A 7 300 5 tot a 27.7335 4.3266215 133.3 A 7 300 5 tot b 27.247 4.3266215 46.2 A 7 300 5 tot tot 27.49025 4.3266375 179.5 A 7 300 5 ty a 29.726 6.111449 21.6 A 7 300 5 ty b 29.796 6.1114495 1.9 A 7 300 5 ty tot 29.761 6.1114425 23.5 A 8 300 15 ju a 22.977 3.2256225 128.2 A 8 300 15 ju b 21.442 3.2256722 53.4 A 8 300 15 ju tot 22.2095 3.2256725 181.6 A 8 300 15 tot a 24.748 2.556412 160.9 A 8 300 15 tot b . . 70 A 8 300 15 tot tot . . 230.9 A 8 300 15 ty a 26.519 1.8871575 32.7 A 8 300 15 ty b . . 16.6 A 8 300 15 ty tot 13.2595 0.9435587 49.3 A 9 300 25 ju a . . 135.5 A 9 300 25 ju b . . 57.7 A 9 300 25 ju tot . . 193.2 A 9 300 25 tot a . . 158.8 A 9 300 25 tot b . . 62.7 A 9 300 25 tot tot . . 221.5 A 9 300 25 ty a 27.583 3.0560635 23.3 A 9 300 25 ty b 25.947 3.0560637 5 A 9 300 25 ty tot 26.765 3.0560637 28.3 B 1 225 5 ju a 24.034 3.4510027 176 B 1 225 5 ju b 25.457 3.4510027 49.2 B 1 225 5 ju tot 24.7455 3.4510027 225.2 B 1 225 5 tot a 27.3305 3.6756203 207.1 B 1 225 5 tot b 27.6575 3.6756203 60.1 B 1 225 5 tot tot 27.494 3.6756203 267.2 B 1 225 5 ty a 30.627 3.900238 31.1 B 1 225 5 ty b 29.858 3.900238 10.9 B 1 225 5 ty tot 30.2425 3.900238 42 B 2 225 15 ju a 25.206 1.912815 173.1 161 B 2 225 15 ju b 24.367 1.9128145 56.9 B 2 225 15 ju tot 24.7865 1.9128145 230 B 2 225 15 tot a 25.2395 2.489521 221 B 2 225 15 tot b 24.687 2.489521 76.5 B 2 225 15 tot tot 24.96325 2.489521 297.5 B 2 225 15 ty a 25.273 3.0662225 47.9 B 2 225 15 ty b 25.007 3.0662225 19.6 B 2 225 15 ty tot 25.14 3.0662725 67.5 B 3 225 25 ju a 23.732 1.485091 147.2 B 3 225 25 ju b 23.294 1.485091 40.6 B 3 225 25 ju tot 23.513 1.485091 187.8 B 3 225 25 tot a 25.5495 1.413439 210.5 B 3 225 25 tot b 26.0535 1.413439 56.6 B 3 225 25 tot tot 25.8015 1.413439 267.1 B 3 225 25 ty a 27.367 1.341787 63.3 B 3 225 25 ty b 28.813 1.341787 16 B 3 225 25 ty tot 28.09 1.341787 79.3 B 4 225 5 ju a 22.277 3.066412 148.2 B 4 225 5 ju b 23.907 3.066241 54.2 B 4 225 5 ju tot 23.092 3.066241 202.4 B 4 225 5 tot a 23.569 3.902149 204.4 B 4 225 5 tot b 24.4045 3.902149 69.9 B 4 225 5 tot tot 23.98675 3.902149 274.3 B 4 225 5 ty a 24.861 4.738056 56.2 B 4 225 5 ty b 24.902 4.7380565 15.7 B 4 225 5 ty tot 24.8815 4.7380565 71.9 B 5 225 15 ju a 23.643 2.2152815 154.5 B 5 225 15 ju b 23.84 2.2152815 68.5 B 5 225 15 ju tot 23.7415 2.2152815 223 B 5 225 15 tot a . . 199.3 B 5 225 15 tot b . . 90 B 5 225 15 tot tot . . 289.3 B 5 225 15 ty a . . 44.8 B 5 225 15 ty b . . 21.5 B 5 225 15 ty tot . . 66.3 B 6 225 25 ju a 21.292 1.940391 171.7 B 6 225 25 ju b 20.466 1.9403915 61.9 B 6 225 25 ju tot 20.879 1.9403915 233.6 B 6 225 25 tot a 21.9785 1.8030022 204.6 B 6 225 25 tot b 20.9655 1.8030021 77.4 B 6 225 25 tot tot 21.472 1.8030021 282 B 6 225 25 ty a 22.665 1.665613 32.9 B 6 225 25 ty b 21.465 1.665613 15.5 B 6 225 25 ty tot 22.065 1.665613 48.4 B 7 225 5 ju a . . 150.3 B 7 225 5 ju b . . 58.4 B 7 225 5 ju tot . . 208.7 B 7 225 5 tot a . . 204.5 B 7 225 5 tot b . . 79.6 B 7 225 5 tot tot . . 284.1 B 7 225 5 ty a 39.805 6.0588082 54.2 B 7 225 5 ty b 43.098 6.0588082 21.2 B 7 225 5 ty tot 41.4515 6.0588082 75.4 B 8 225 15 ju a . . 165.2 162 B 8 225 15 ju b . . 79.6 B 8 225 15 ju tot . . 244.8 B 8 225 15 tot a . . 209 B 8 225 15 tot b . . 94.4 B 8 225 15 tot tot . . 303.4 B 8 225 15 ty a 34.325 4.6163167 43.8 B 8 225 15 ty b 36.341 4.6163167 14.8 B 8 225 15 ty tot 35.333 4.6163167 58.6 B 9 225 25 ju a 24.55 1.7548872 154.5 B 9 225 25 ju b 24.146 1.7548872 85.6 B 9 225 25 ju tot 24.348 1.7548872 240.1 B 9 225 25 tot a 25.8425 2.4533627 174.8 B 9 225 25 tot b 26.9565 2.4533627 96.3 B 9 225 25 tot tot 26.3995 2.4533627 271.1 B 9 225 25 ty a 27.135 3.1518382 20.3 B 9 225 25 ty b 29.767 3.1518382 10.7 B 9 225 25 ty tot 28.451 3.1518382 31 C 1 150 5 ju a 19.791 2.6988897 181.3 C 1 150 5 ju b 20.576 2.6988897 48.8 C 1 150 5 ju tot 20.1835 2.6988897 230.1 C 1 150 5 tot a 23.106 3.1970976 267.4 C 1 150 5 tot b 26.5915 3.1970976 81.5 C 1 150 5 tot tot 24.84875 3.1970976 348.9 C 1 150 5 ty a 26.421 3.6953055 86.1 C 1 150 5 ty b 32.607 3.6953055 32.7 C 1 150 5 ty tot 29.514 3.6953055 118.8 C 2 150 15 ju a 10.747 2.127868 177.9 C 2 150 15 ju b 11.452 2.1278682 48.1 C 2 150 15 ju tot 11.0995 2.1278682 226 C 2 150 15 tot a 16.731 2.0677111 251 C 2 150 15 tot b 16.553 2.0677111 84.7 C 2 150 15 tot tot 16.642 2.0677111 335.7 C 2 150 15 ty a 22.715 2.007554 73.1 C 2 150 15 ty b 21.654 2.007554 36.6 C 2 150 15 ty tot 22.1845 2.007554 109.7 C 3 150 25 ju a 27.115 1.568104 190.1 C 3 150 25 ju b 27.065 1.5681042 53.8 C 3 150 25 ju tot 27.09 1.5681042 243.9 C 3 150 25 tot a 24.027 1.5448716 261 C 3 150 25 tot b 24.5085 1.5448716 85.7 C 3 150 25 tot tot 24.26775 1.5448716 346.7 C 3 150 25 ty a 20.939 1.521639 70.9 C 3 150 25 ty b 21.952 1.521639 31.9 C 3 150 25 ty tot 21.4455 1.521639 102.8 C 4 150 5 ju a 22.208 2.303088 154.5 C 4 150 5 ju b 23.251 2.3030887 58.4 C 4 150 5 ju tot 22.7295 2.3030887 212.9 C 4 150 5 tot a . . 232.7 C 4 150 5 tot b . . 113.2 C 4 150 5 tot tot . . 345.9 C 4 150 5 ty a . . 78.2 C 4 150 5 ty b . . 54.8 C 4 150 5 ty tot . . 133 C 5 150 15 ju a . . 172.2 163 C 5 150 15 ju b . . 49.0 C 5 150 15 ju tot . . 221.2 C 5 150 15 tot a . . 216.4 C 5 150 15 tot b . . 72.9 C 5 150 15 tot tot . . 289.3 C 5 150 15 ty a 35.744 2.1147337 44.2 C 5 150 15 ty b 39.911 2.1147337 23.9 C 5 150 15 ty tot 37.8275 2.1147337 68.1 C 6 150 25 ju a 21.557 1.8419985 188.7 C 6 150 25 ju b 21.346 1.8419985 60.3 C 6 150 25 ju tot 21.4515 1.8419985 249 C 6 150 25 tot a . . 253.9 C 6 150 25 tot b . . 103 C 6 150 25 tot tot . . 356.9 C 6 150 25 ty a . . 65.2 C 6 150 25 ty b . . 42.7 C 6 150 25 ty tot . . 107.9 C 7 150 5 ju a 23.714 2.372604 193.2 C 7 150 5 ju b 23.485 2.3726042 51.2 C 7 150 5 ju tot 23.5995 2.3726045 244.4 C 7 150 5 tot a 20.354 2.1007837 245.4 C 7 150 5 tot b 20.3445 2.1007843 74.8 C 7 150 5 tot tot 20.34925 2.1007843 320.2 C 7 150 5 ty a 16.994 1.8289645 52.2 C 7 150 5 ty b 17.204 1.8289645 23.6 C 7 150 5 ty tot 17.099 1.8289645 75.8 C 8 150 15 ju a 23.191 1.842798 213.9 C 8 150 15 ju b 22.878 1.842798 57.5 C 8 150 15 ju tot 23.0345 1.842798 271.4 C 8 150 15 tot a 21.0545 1.9610547 277.5 C 8 150 15 tot b 22.226 1.9610543 90.3 C 8 150 15 tot tot 21.64025 1.9610543 367.8 C 8 150 15 ty a 18.918 2.0793107 63.6 C 8 150 15 ty b 21.574 2.0793107 32.8 C 8 150 15 ty tot 20.246 2.0793107 96.4 C 9 150 25 ju a 22.862 1.4387587 178.7 C 9 150 25 ju b 22.737 1.4387587 66.3 C 9 150 25 ju tot 22.7995 1.4387587 245 C 9 150 25 tot a 19.5745 1.2497783 235 C 9 150 25 tot b 20.487 1.2497783 97.6 C 9 150 25 tot tot 20.03075 1.2497783 332.6 C 9 150 25 ty a 16.287 1.060798 56.3 C 9 150 25 ty b 18.237 1.060798 31.3 C 9 150 25 ty tot 17.262 1.060798 87.6 D 1 75 5 ju a 19.251 2.0141755 143.8 D 1 75 5 ju b 19.787 2.0141755 44.7 D 1 75 5 ju tot 19.519 2.0141755 188.5 D 1 75 5 tot a 21.474 1.8410576 262.1 D 1 75 5 tot b 21.797 1.8410576 115.4 D 1 75 5 tot tot 21.6355 1.8410576 377.5 D 1 75 5 ty a 23.697 1.6679397 118.3 D 1 75 5 ty b 23.807 1.6679397 70.7 D 1 75 5 ty tot 23.752 1.6679397 189 D 2 75 15 ju a 15.123 1.9987757 174.7 164 D 2 75 15 ju b 14.433 1.9987757 67.1 D 2 75 15 ju tot 14.778 1.9987757 241.8 D 2 75 15 tot a 18.519 1.8620626 258.4 D 2 75 15 tot b 18.669 1.8620626 130.1 D 2 75 15 tot tot 18.594 1.8620626 388.5 D 2 75 15 ty a 21.915 1.7253495 83.7 D 2 75 15 ty b 22.905 1.7253495 63 D 2 75 15 ty tot 22.41 1.7253495 146.7 D 3 75 25 ju a 19.931 1.3594965 199.7 D 3 75 25 ju b 19.613 1.3594965 75.6 D 3 75 25 ju tot 19.772 1.3594965 275.3 D 3 75 25 tot a 20.26 1.5885226 273.2 D 3 75 25 tot b 19.581 1.5885226 128.8 D 3 75 25 tot tot 19.9205 1.5885226 402 D 3 75 25 ty a 20.589 1.817548 73.5 D 3 75 25 ty b 19.549 1.817548 53.2 D 3 75 25 ty tot 20.069 1.817548 126.7 D 4 75 5 ju a . . 141.2 D 4 75 5 ju b . . 73.2 D 4 75 5 ju tot . . 214.4 D 4 75 5 tot a . . 235.3 D 4 75 5 tot b . . 144.3 D 4 75 5 tot tot . . 379.6 D 4 75 5 ty a 12.451 2.508661 94.1 D 4 75 5 ty b 12.896 2.508661 71.1 D 4 75 5 ty tot 12.6735 2.508661 165.2 D 5 75 15 ju a 15.769 2.0187715 100.4 D 5 75 15 ju b 15.376 2.0187715 66.4 D 5 75 15 ju tot 15.5725 2.0187715 166.8 D 5 75 15 tot a 14.22 1.6215755 181.5 D 5 75 15 tot b 14.123 1.6215755 128.2 D 5 75 15 tot tot 14.1715 1.6215755 309.7 D 5 75 15 ty a 12.671 1.2243795 81.1 D 5 75 15 ty b 12.87 1.2243795 61.8 D 5 75 15 ty tot 12.7705 1.2243795 142.9 D 6 75 25 ju a 12.864 1.8440585 159.1 D 6 75 25 ju b 11.986 1.8440585 75.4 D 6 75 25 ju tot 12.425 1.8440585 234.5 D 6 75 25 tot a 14.8415 1.4130683 236 D 6 75 25 tot b 13.914 1.4130683 118.6 D 6 75 25 tot tot 14.37775 1.4130683 354.6 D 6 75 25 ty a 16.819 0.9820785 76.9 D 6 75 25 ty b 15.842 0.9820785 43.2 D 6 75 25 ty tot 16.3305 0.9820785 120.1 D 7 75 5 ju a 20.45 2.0483652 163.6 D 7 75 5 ju b 19.042 2.0483652 65.8 D 7 75 5 ju tot 19.746 2.0483652 229.4 D 7 75 5 tot a . . 262 D 7 75 5 tot b . . 129.5 D 7 75 5 tot tot . . 391.5 D 7 75 5 ty a . . 98.4 D 7 75 5 ty b . . 63.7 D 7 75 5 ty tot . . 162.1 D 8 75 15 ju a 15.147 1.4438872 221.6 165 D 8 75 15 ju b 14.856 1.4438872 99.4 D 8 75 15 ju tot 15.0015 1.4438872 321 D 8 75 15 tot a 17.13 1.2500068 246.8 D 8 75 15 tot b 16.1915 1.2500068 118.8 D 8 75 15 tot tot 16.66075 1.2500068 365.6 D 8 75 15 ty a 19.113 1.0561265 25.2 D 8 75 15 ty b 17.527 1.0561265 19.4 D 8 75 15 ty tot 18.32 1.0561265 44.6 D 9 75 25 ju a 21.08 1.316284 164.4 D 9 75 25 ju b 20.452 1.3162847 74.6 D 9 75 25 ju tot 20.766 1.3162847 239 D 9 75 25 tot a 24.529 1.2143286 244.3 D 9 75 25 tot b 24.844 1.2143286 144 D 9 75 25 tot tot 24.6865 1.2143286 388.3 D 9 75 25 ty a 27.978 1.1123725 79.9 D 9 75 25 ty b 29.236 1.1123725 69.4 D 9 75 25 ty tot 28.607 1.1123725 149.3 run; title2 biomass ANOVAs in 2003; title3 total biomass for both species in 2003 ANOVA; data total_biomass; set total; if species="tot" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=total_biomass; class ammonia n_p species; model biomass=species|ammonia|n_p/ddfm=kr outp=resids; *lsmeans species|ammonia|n_p / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_biomass normal plot; var biomass; run; *test shows a Shapiro-Wilk's Test (as S-W test from now on) of 0.97, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 aboveground biomass for both species in 2003 ANOVA; data total_above; set total; if species="tot" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=total_above; 166 class ammonia n_p species; model biomass=species|ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_above normal plot; var biomass; run; *test shows a S-W test of 0.97, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 belowground biomass for both species in 2003 ANOVA; data total_below; set total; if species="tot" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=total_below; class ammonia n_p species; model biomass=species|ammonia|n_p /ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_below normal plot; var biomass; run; *test shows a S-W test of 0.96, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 total biomass for TYPHA in 2003 ANOVA; data typha_total; set total; if species="tot" then delete; if species="ju" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=typha_total; class ammonia n_p; model biomass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; 167 run; title4 checks for normality; proc univariate data=typha_total normal plot; var biomass; run; *test shows a S-W test of 0.97, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent homogeneity of resids; ************************************************************ *****************************; title3 aboveground biomass for TYPHA in 2003 ANOVA; data typha_above; set total; if species="tot" then delete; if species="ju" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=typha_above; class ammonia n_p; model biomass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_above normal plot; var biomass; run; *test shows a S-W test of 0.98, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 belowground biomass for TYPHA in 2003 ANOVA; data typha_below; set total; if species="tot" then delete; if species="ju" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=typha_below; class ammonia n_p; model biomass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_below normal plot; 168 var biomass; run; *test shows a S-W test of 0.97, indicating decent normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 total biomass for JUNCUS in 2003 ANOVA; data juncus_total; set total; if species="tot" then delete; if species="ty" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=juncus_total; class ammonia n_p; model biomass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_total normal plot; var biomass; run; *test shows a S-W test of 0.87, indicating normality, supported graphically; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 aboveground biomass for JUNCUS in 2003 ANOVA; data juncus_above; set total; if species="tot" then delete; if species="ty" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=juncus_above; class ammonia n_p; model biomass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_above normal plot; var biomass; run; 169 *test shows a S-W test of 0.88, indicating normality, supported graphically; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 belowground biomass for JUNCUS in 2003 ANOVA; data juncus_below; set total; if species="tot" then delete; if species="ty" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=juncus_below; class ammonia n_p; model biomass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_below normal plot; var biomass; run; *test shows a S-W test of 0.88, indicating normality, supported graphically; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ **************************** Here ends the biomass ANOVA series...................................................... ************************************************************ ***************************; title2 Tissue Nitrogen Concentration ANOVAs in 2003; title3 total n_conc for both species in 2003 ANOVA; data total_n_conc; set total; if species="tot" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=total_n_conc; class ammonia n_p species; model n_conc=species|ammonia|n_p/ddfm=kr outp=resids; 170 *lsmeans species|ammonia|n_p / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_n_conc normal plot; var n_conc; run; *test shows a Shapiro-Wilk's Test (as S-W test from now on) of 0.88, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 aboveground n_conc for both species in 2003 ANOVA; data total_above; set total; if species="tot" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=total_above; class ammonia n_p species; model n_conc=species|ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_above normal plot; var n_conc; run; *test shows a S-W test of 0.86, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 belowground n_conc for both species in 2003 ANOVA; data total_below; set total; if species="tot" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=total_below; class ammonia n_p species; model n_conc=species|ammonia|n_p /ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; 171 proc univariate data=total_below normal plot; var n_conc; run; *test shows a S-W test of 0.86, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 total n_conc for TYPHA in 2003 ANOVA; data typha_total; set total; if species="tot" then delete; if species="ju" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=typha_total; class ammonia n_p; model n_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_total normal plot; var n_conc; run; *test shows a S-W test of 0.91, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent homogeneity of resids; ************************************************************ *****************************; title3 aboveground n_conc for TYPHA in 2003 ANOVA; data typha_above; set total; if species="tot" then delete; if species="ju" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=typha_above; class ammonia n_p; model n_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_above normal plot; var n_conc; run; 172 *test shows a S-W test of 0.91, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 belowground n_conc for TYPHA in 2003 ANOVA; data typha_below; set total; if species="tot" then delete; if species="ju" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=typha_below; class ammonia n_p; model n_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_below normal plot; var n_conc; run; *test shows a S-W test of 0.88, indicating decent normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 total n_conc for JUNCUS in 2003 ANOVA; data juncus_total; set total; if species="tot" then delete; if species="ty" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=juncus_total; class ammonia n_p; model n_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_total normal plot; var n_conc; run; *test shows a S-W test of 0.91, indicating normality; 173 proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 aboveground n_conc for JUNCUS in 2003 ANOVA; data juncus_above; set total; if species="tot" then delete; if species="ty" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=juncus_above; class ammonia n_p; model n_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_above normal plot; var n_conc; run; *test shows a S-W test of 0.91, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 belowground n_conc for JUNCUS in 2003 ANOVA; data juncus_below; set total; if species="tot" then delete; if species="ty" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=juncus_below; class ammonia n_p; model n_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_below normal plot; var n_conc; run; *test shows a S-W test of 0.91, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; 174 quit; *shows decent and acceptable homogeneity of resids; ************************************************************ **************************** Here ends the tissue nitrogen concentration ANOVA series................................ ************************************************************ ***************************; title2 Tissue Mass Nitrogen ANOVAs in 2003; title3 total n_mass for both species in 2003 ANOVA; data total_n_mass; set total; if species="tot" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=total_n_mass; class ammonia n_p species; model n_mass=species|ammonia|n_p/ddfm=kr outp=resids; *lsmeans species|ammonia|n_p / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_n_mass normal plot; var n_mass; run; *test shows a Shapiro-Wilk's Test (as S-W test from now on) of 0.78, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 aboveground n_mass for both species in 2003 ANOVA; data total_above; set total; if species="tot" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=total_above; class ammonia n_p species; model n_mass=species|ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_above normal plot; var n_mass; run; 175 *test shows a S-W test of 0.78, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 belowground n_mass for both species in 2003 ANOVA; data total_below; set total; if species="tot" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=total_below; class ammonia n_p species; model n_mass=species|ammonia|n_p /ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_below normal plot; var n_mass; run; *test shows a S-W test of 0.78, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 total n_mass for TYPHA in 2003 ANOVA; data typha_total; set total; if species="tot" then delete; if species="ju" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=typha_total; class ammonia n_p; model n_mass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_total normal plot; var n_mass; run; *test shows a S-W test of 0.86, indicating normality; proc plot data=resids vpercent=50; 176 plot resid*pred / vref=0; quit; *shows decent homogeneity of resids; ************************************************************ *****************************; title3 aboveground n_mass for TYPHA in 2003 ANOVA; data typha_above; set total; if species="tot" then delete; if species="ju" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=typha_above; class ammonia n_p; model n_mass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_above normal plot; var n_mass; run; *test shows a S-W test of 0.86, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 belowground n_mass for TYPHA in 2003 ANOVA; data typha_below; set total; if species="tot" then delete; if species="ju" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=typha_below; class ammonia n_p; model n_mass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_below normal plot; var n_mass; run; *test shows a S-W test of 0.86, indicating decent normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; 177 *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 total n_mass for JUNCUS in 2003 ANOVA; data juncus_total; set total; if species="tot" then delete; if species="ty" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=juncus_total; class ammonia n_p; model n_mass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_total normal plot; var n_mass; run; *test shows a S-W test of 0.95, indicating normality, supported graphically; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 aboveground n_mass for JUNCUS in 2003 ANOVA; data juncus_above; set total; if species="tot" then delete; if species="ty" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=juncus_above; class ammonia n_p; model n_mass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_above normal plot; var n_mass; run; *test shows a S-W test of 0.96, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; 178 ************************************************************ *****************************; title3 belowground n_mass for JUNCUS in 2003 ANOVA; data juncus_below; set total; if species="tot" then delete; if species="ty" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=juncus_below; class ammonia n_p; model n_mass=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_below normal plot; var n_mass; run; *test shows a S-W test of 0.95, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *************************** Here ends tissue mass nitrogen ANOVA series................................. ************************************************************ **************************; title2 Tissue Phosphorus Concentration ANOVAs in 2003; title3 total p_conc for both species in 2003 ANOVA; data total_p_conc; set total; if species="tot" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=total_p_conc; class ammonia n_p species; model p_conc=species|ammonia|n_p/ddfm=kr outp=resids; *lsmeans species|ammonia|n_p / pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_p_conc normal plot; var p_conc; run; *test shows a Shapiro-Wilk's Test (as S-W test from now on) of 0.93, indicating normality; 179 proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 aboveground p_conc for both species in 2003 ANOVA; data total_above; set total; if species="tot" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=total_above; class ammonia n_p species; model p_conc=species|ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_above normal plot; var p_conc; run; *test shows a S-W test of 0.93, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 belowground p_conc for both species in 2003 ANOVA; data total_below; set total; if species="tot" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=total_below; class ammonia n_p species; model p_conc=species|ammonia|n_p /ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=total_below normal plot; var p_conc; run; *test shows a S-W test of 0.96, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; 180 *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 total p_conc for TYPHA in 2003 ANOVA; data typha_total; set total; if species="tot" then delete; if species="ju" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=typha_total; class ammonia n_p; model p_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_total normal plot; var p_conc; run; *test shows a S-W test of 0.94, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent homogeneity of resids; ************************************************************ *****************************; title3 aboveground p_conc for TYPHA in 2003 ANOVA; data typha_above; set total; if species="tot" then delete; if species="ju" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=typha_above; class ammonia n_p; model p_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_above normal plot; var p_conc; run; *test shows a S-W test of 0.96, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; 181 ************************************************************ *****************************; title3 belowground p_conc for TYPHA in 2003 ANOVA; data typha_below; set total; if species="tot" then delete; if species="ju" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=typha_below; class ammonia n_p; model p_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=typha_below normal plot; var p_conc; run; *test shows a S-W test of 0.89, indicating decent normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 total p_conc for JUNCUS in 2003 ANOVA; data juncus_total; set total; if species="tot" then delete; if species="ty" then delete; if type="a" then delete; if type="b" then delete; run; proc mixed data=juncus_total; class ammonia n_p; model p_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_total normal plot; var p_conc; run; *test shows a S-W test of 0.96, indicating normality, supported graphically; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; 182 title3 aboveground p_conc for JUNCUS in 2003 ANOVA; data juncus_above; set total; if species="tot" then delete; if species="ty" then delete; if type="tot" then delete; if type="b" then delete; run; proc mixed data=juncus_above; class ammonia n_p; model p_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_above normal plot; var p_conc; run; *test shows a S-W test of 0.99, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 belowground p_conc for JUNCUS in 2003 ANOVA; data juncus_below; set total; if species="tot" then delete; if species="ty" then delete; if type="a" then delete; if type="tot" then delete; run; proc mixed data=juncus_below; class ammonia n_p; model p_conc=ammonia|n_p/ddfm=kr outp=resids; *lsmeans ammonia n_p*ammonia/ pdiff adjust=tukey; run; title4 checks for normality; proc univariate data=juncus_below normal plot; var p_conc; run; *test shows a S-W test of 0.95, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *************************** 183 Here ends tissue phosphorus concentration ANOVA series................................. ************************************************************ **************************; title2 root-to-shoot ratio ANOVAs in 2003; data r_sratio; input column$ row$ ammonia$ n_p$ species$ r_s; log_r_s=log(r_s); datalines; d 1 75 5 j 0.310848401 d 2 75 15 j 0.518413598 d 3 75 25 j 0.402200489 d 4 75 5 j 0.384087006 d 5 75 15 j 0.661354582 d 6 75 25 j 0.448555957 d 7 75 5 j 0.378567852 d 8 75 15 j 0.473915776 d 9 75 25 j 0.45377129 c 1 150 5 j 0.269167126 c 2 150 15 j 0.377993528 c 3 150 25 j 0.265010352 c 4 150 5 j 0.270376616 c 5 150 15 j 0.284552846 c 6 150 25 j 0.268817204 c 7 150 5 j 0.283008943 c 8 150 15 j 0.319554849 c 9 150 25 j 0.371012871 b 1 225 5 j 0.279545455 b 2 225 15 j 0.365721997 b 3 225 25 j 0.388556221 b 4 225 5 j 0.328711727 b 5 225 15 j 0.443365696 b 6 225 25 j 0.481840194 b 7 225 5 j 0.275815217 b 8 225 15 j 0.360512522 b 9 225 25 j 0.554045307 a 1 300 5 j 0.374233129 a 2 300 15 j 0.271197411 a 3 300 25 j 0.39659803 a 4 300 5 j 0.327586207 a 5 300 15 j 0.443312967 a 6 300 25 j 0.416536661 a 7 300 5 j 0.370620774 a 8 300 15 j 0.416903409 a 9 300 25 j 0.425830258 d 1 75 5 t 0.597633136 d 2 75 15 t 0.755579171 d 3 75 25 t 0.647357724 d 4 75 5 t 0.752688172 d 5 75 15 t 0.762022195 d 6 75 25 t 0.76984127 d 7 75 5 t 0.723809524 d 8 75 15 t 0.561768531 184 d 9 75 25 t 0.868585732 c 1 150 5 t 0.379790941 c 2 150 15 t 0.700767263 c 3 150 25 t 0.45210728 c 4 150 5 t 0.500683995 c 5 150 15 t 0.540723982 c 6 150 25 t 0.51572327 c 7 150 5 t 0.449929478 c 8 150 15 t 0.654907975 c 9 150 25 t 0.555950266 b 1 225 5 t 0.350482315 b 2 225 15 t 0.279359431 b 3 225 25 t 0.391143911 b 4 225 5 t 0.409185804 b 5 225 15 t 0.479910714 b 6 225 25 t 0.337899543 b 7 225 5 t 0.252764613 b 8 225 15 t 0.47112462 b 9 225 25 t 0.527093596 a 1 300 5 t 0.353135314 a 2 300 15 t 0.385749386 a 3 300 25 t 0.087962963 a 4 300 5 t 0.339181287 a 5 300 15 t 0.981042654 a 6 300 25 t 0.50764526 a 7 300 5 t 0.422222222 a 8 300 15 t 0.377245509 a 9 300 25 t 0.214592275 run; title3 root-to-shoot ratios for mesocosms in 2003 ANOVA; proc mixed data=r_sratio; class ammonia n_p species; model r_s=species|ammonia|n_p/ ddfm=kr outp=resids; *lsmeans ammonia / pdiff; run; title4 checks for normality; proc univariate data=r_sratio normal plot; var r_s; run; *test shows a S-W test of 0.93, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 root-to-shoot ratios for TYPHA in 2003 ANOVA; proc mixed data=r_sratio; where species='t'; class ammonia n_p; model r_s=ammonia|n_p/ ddfm=kr outp=resids; run; 185 title4 checks for normality; proc univariate data=r_sratio normal plot; where species='t'; var r_s; run; *test shows a S-W test of 0.98, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; ************************************************************ *****************************; title3 root-to-shoot ratio for JUNCUS in 2003 ANOVA; proc mixed data=r_sratio; where species='j'; class ammonia n_p; model r_s=ammonia|n_p/ ddfm=kr outp=resids; run; title4 checks for normality; proc univariate data=r_sratio normal plot; where species='j'; var r_s; run; *test shows a S-W test of 0.93, indicating normality; proc plot data=resids vpercent=50; plot resid*pred / vref=0; quit; *shows decent and acceptable homogeneity of resids; quit; 186 REFERENCES Aerts, R; Wallen, B; Malmer, N; H de Caluwe. 2001. Nutritional constraints on Sphagnum-growth and decay in northern peatlands. Journal of Ecology 89:292- 299. Aerts, R; FS Chapin III. 2000. The mineral nutrition of wild plants revisited: a re- evaluation of processes and patterns. Advances in Ecological Research 30:1-67. Aerts, R; R Bobbink. 1999. The impact of atmospheric nitrogen deposition on vegetation processes in terrestrial, non-forest ecosystems. In: SJ Langan, ed. The impact of nitrogen deposition on natural and semi-natural ecosystems. Kluwer, Dordecht, The Netherlands. Aerts, R; Verhoeven, JTA; Whigham, DF. 1999. Plant-mediated controls on nutrient cycling in temperate fens and bogs. Ecology 80:2170-2181. Aerts, R; Berendse, F; Decaluwe, H; M Schmitz. 1990. Competition in heathland along and experimental gradient of nutrient availability. Oikos 57:310-318. Abbes, C; Parent, LE; Karam, A.; D Isfan. 1995. Onion response to ammoniated peat and ammonium sulfate in relation to ammonium toxicity. Canadian Journal of Soil Science 75:261-272. Bachand, PAM; AJ Horne. 2000a. Denitrification in constructed free-water surface wetlands I: very high notrate removal rates in a macrocosm study. Ecological Engineering 14:9-15. Bachand, PAM; AJ Horne. 2000b. Denitrification in constructed free-water surface wetlands II: effects of vegetation and temperature. Ecological Engineering 14:17- 32. Bastviken, SK; Erikkson, PG; Martins, I; Neto, JM; Leonardson, L; K Tondurski. 2003 Potential nitrification and denitrification on different surfaces in a constructed wetland. Journal of Environmental Quality 32:2414-2420. Bedford, BL; Walbridge, M; A Aldus. 1999. Patterns in nutrient availability and plant diversity of temperate North American wetlands. Ecology 80:2151-2169. Bendix, M; Tornbjerg, T;H Brix. 1994. Internal gas transport in Typha latifolia L. and Typha angustifolia L. 1. Humidity-induced pressurization and convective throughflow. Aquatic Botany 49:75-89. Bobbink, R; Hornung, RM; JGM Roelofs. 1998. The effects of air-borne nitrogen pollution on species diversity in natural and semi-natural European vegetation-a review. Journal of Ecology 86:717-738. 187 Braakhekke, WG; DAP Hooftman. 1999. The resource balance hypothesis of plant species diversity in grassland. Journal of Vegetation Science 10:187-200. Brewer, JS. 2003. Nitrogen addition does not reduce below-ground competition in a salt marsh clonal plant community in Mississippi (USA). Plant Ecology 168:93-106. Britto, DT; HJ Kronzucker. 2002. NH 4 + toxicity in higher plants: a critical review. Journal of Plant Physiology 159:567-584. Brix, H. 1997. Do macrophytes play a role in constructed treatment wetlands? Water Science and Technology 13:11-17. Brix, H; BK Sorrell. 1996. Oxygen stress in wetland plants: Comparison of de- oxygenated and reducing root environments. Functional Ecology 10:521-526. Brix, H; Sorrell, BK; H-H Schierup. 1996. Gas fluxes achieved by in situ convective flow in Phragmites australis. Aquatic Botany 54:151-163. Brix, H; Sorrell, MK; PT Orr. 1992. Internal pressurization and convective gas flow in some emergent freshwater macrophytes. Limnology and Oceanography 37:1420- 1433. Brix, H. 1989. Gas exchange through dead culms of reed, Phragmites australis (Cav.)Trin.ex. Steudel. Aquatic Botany 35:81-98. Camacho, A; Wurtsbaugh, WA; Miracle, MR; Armengol, X; E Vicente. 2003. Nitrogen limitation of phytoplankton in a Spanish karst lake with a deep chlorophyll maximum: a nutrient enrichment bioassay approach. Journal of Plankton Research 25:397-404. Cao, Y; Glass, ADM; NM Crawford. 1993. Ammonium inhibition of Arabidopsis root growth can be reversed by potassium and by auxin resistance mutations aux1, axr1, and axr2. Plant Physiology 102:983-989. Clarke, E; AH Baldwin. 2002. Response of wetland plants to ammonia and water level. Ecological Engineering 18:257-264. Clarke, E. 1999. Constructed wetlands for treating dairy wastewater: suitability of plant species and treatment effectiveness. M.S. Thesis, University of Maryland, College Park, Maryland, USA. Craft, C; Reader, J; Sacco, JN; SW Broome. 1999. Twenty-five years of ecosystem development of constructed Spartina alterniflora(Loisel) marshes. Ecological Applications 9:1405-1419. 188 Cronk, JK. 1996. Constructed wetlands to treat wastewater from dairy and swine operations, a review. Agriculture, Ecosystems and Environment 58:97-114. DeKroon, H; R Bobbink. 1997. Clonal plant dominance under elevated nitrogen deposition with species reference to Brachypodium pinnatum in chalk grassland. In: HD Koon and J van Groenendael, eds. The ecology and evolution of clonal plants. Backhuys, Leiden, The Netherlands. Dijk, E; H Olff. 1994. Effects of nitrogen, phosphorus and potassium fertilization on field performance of Dactylorhiza majalis. Acta Botanica Neederlandica. 43:383-392. Dijk E; N Eck. 1995. Ammonium toxicity and nitrate response of axenically grown Dactylorhiza incarnata seedlings. New Phytology 131:361-367. DiRico, EF. 2000. Treatment efficiency of an agricultural treatment wetland. M.S. Thesis. University of Maryland, College Park, Maryland, USA. DiTommaso, A; LW Aarssen. 1989. Resource manipulations in natural vegetation: a review. Vegetatio 84:9-29. Downing, JA; E McCauley. 1992. The nitrogen: phosphorus relationship in lakes. Limnology and Oceanography 37:936-945. Emery, NC; Ewanchuk, PJ; MD Bertness. 2001. Competition and salt-marsh plant zonation: stress tolerators may be dominant competitors. Ecology 82:2471-2485. Engelhardt, K; Ritchie, AM; ME Ritchie. 2001. Effects of macrophyte species richness of wetland functioning and services. Nature 411:687-689. Gale, PM; Reddy KR; DA Graetz. 1993. Nitrogen removal from reclaimed water applied to constructed and natural wetland microcosms. Water Environment Research 65:162-168. Glass, ADM; Britto, DT; Kaiser, BN; Kinghorn, JR; Kronzucker, HJ; Kumar, A; Okamoto, M; Rawat, S; Siddiqi, MY; Unkles, SE; JJ Vidmar. 2002. The regulation of nitrate and ammonium transport systems in plants. Journal of Experimental Botany 53:855-864. Goldberg, DE; AM Barton. 1992. Patterns and consequences of interspecific competition in natural communities: a review of experiments with plants. American Naturalist 139: 771-801. Grace, JB; RG Wetzel. 1981. Habitat partitioning and competitive displacement in cattail (Typha): experimental field studies. American Naturalist 118:463-474. 189 Green, EK; SM Galatowitsch. 2002. Effects of Phalaris arundinacea and nitrate-N addition on the establishment of wetland plant communities. Journal of Applied Ecology 39:134-144. Grime, JP. 1979. Evidence for the existence of three primary strategies in plants and its relevance to ecological and evolutionary theory. American Naturalist 111:1169- 1194. Gumbricht, T. 1993. Nutrient removal processes in freshwater submersed macrophyte systems. Ecological Engineering 2:1-30. Gusewell, S; Koerselman, W; JTA Verhoeven. 2003. Biomass N:P ratios as indicators of nutrient limitation for plant populations in wetlands. Ecological Applications 13:372-384. Gusewell, S; Koerselman, W; JTA Verhoeven. 2002. Time dependant effects of fertilization in Dutch floating fens. Journal of Vegetation Science 13:705-718. Hageman, RH. 1984. Ammonium versus nitrate nutrition in higher plants. In: Nitrogen in crop production. ASA-CSSA-SSSA, Madison, Wisconsin, USA, pp.67-82. Hammer, DA. 1992. Designing constructed wetlands to treat agricultural nonpoint source pollution. Ecological Engineering 1:49-82. Hammer, DA; RL Knight. 1994. Designing constructed wetlands for nitrogen removal. Water Science and Technology 29:15-27. Hill, DT; Payne, VWE; Rogers, JW; SR Kown. 1997. Ammonia effects on the biomass of five constructed wetland plant species. Bioresources Technology 62:109-113. Howes, BL; Howarth, RW; Teal, JM; I Valiela. 1981. Oxidation-reduction potentials in salt marshes: spatial patterns and interactions with primary productivity. Limnology and Oceanography 26:350-360. Humenik, FJ; Szogi, AA; Hunt, PG; Broome, S; M Rice. 1999. Wastewater utilization: a place for managed wetlands. Asian-Australiasian Journal of Animal Science 12:629-632. Hunter, RG; Combs, DL; DB George. 2000. Growth of softstem bullrush (Scirpus validus) in microcosms with different hydrologic regimes and media depths. Wetlands 20:15-22. Jansson, M; Bergstrom, AK; Drakare, S; P Blomqvist. 2001. Nutrient limitation of bacterioplankton and phytoplankton in humic lakes in northern Sweden. Freshwater Biology 46:653-666. 190 Jespersen, DN; Sorrell, BK; H Brix. 1998. Growth and root oxygen release by Typha latifolia and its effects on sediment methanogenesis. Aquatic Botany 61:165-180. Kadlec, RH; RL Knight. 1996. Treatment Wetlands. CRC Press, Boca Raton, Florida, USA. Karpiscak, MM; Frietas, RJ; Gerbas, CP; Sanchez, LR; E Shamir. 1999. Management of dairy waste in the Sonoran desert using constructed wetland technology. Water Science and Technology 40:57-65. Keddy, P; Gaudet, C; LH Fraser. Effect of low and high nutrients on competitive hierarchy of 26 shoreline plants. Journal of Ecology 88:413-423. Kent, R; Odum, HT; FN Scatena. 2000. Eutrophic overgrowth in the self-organization of tropical wetlands illustrated with a study of swine wastes in rainforest plots. Ecological Engineering 16:255-269. Kielland, K. 1994. Amino acid absorption by arctic plants: implications for plant nutrition and nitrogen cycling. Ecology 75:2373-2383. Knight, RL; Gu, BH; Clarke, RA; JM Newman. 2003. Long-term phosphorus removal in Florida aquatic systems dominated by submerged aquatic vegetation. Ecological Engineering 20:45-63. Knight, RL; Payne, VWE; Borer, RE; Clarke, RA; JH Pries. 2000. Constructed wetlands for livestock wastewater management. Ecological Engineering 15:41-55. Koerselman, W; AFN Meuleman. 1996. The vegetation N:P ratio: a new tool to detect the nature of nutrient limitation. Journal of Applied Ecology 33: 1441-1450. Konyha, KD; Shaw, DT; KW Weller. 1995. Hydrologic design of a wetland: advantages of continuous modeling. Ecological Engineering 4:99-116. Kronzucker, HJ; Britto, DT; Davenport, RJ; M Tester. 2001. Ammonium toxicity: the real cost of transport. Trends in Plant Science 6:335-337. Lammerts, EJ; Pegtel, DM; Grootjans, AP; A van der Veen. 1999. Nutrient limitation and vegetation changes in a coastal dune slack. Journal of Vegetation Science 10:111- 122. Lechowicz, MJ; GR Shaver. 1982. A multivariate approach to the analysis of factorial fertilization experiments in Alaskan arctic tundra. Ecology 63:1029-1038. Mamolos, AP; DS Veresoglou. 2000. Patterns of root activity and responses of species to nutrients in vegetation of fertile alluvial soil. Plant Ecology 148:245-253. 191 Mamolos, AP; Veresoglou, DS; N Barbayiannis. 1995. Plant species abundance and tissue concentrations of limiting nutrients in low nutrient grasslands: a test of competition theory. Journal of Ecology 83:485-495. Magalhaes, JR; Machado, AT; DM Huber. 1995. Similarities in response of maize genotypes to water logging and ammonium toxicity. Journal of Plant Nutrition 18:2339-2346. Mehrer, I; H Mohr. 1989. Ammonium toxicity: description of the syndrom in Sinapsis alba and the search for its causation. Physiologia Plantarum 77:545-554. McJannet, CL; Keddey, PA; FR Pick. 1995. Nitrogen and phosphorus tissue concentrations in 41 wetland plants: a comparison across habitats and functional groups. Functional Ecology 9:231-238. Mitsch, WJ; JG Gosselink. 2000. Wetlands, 3 rd ed. John Wiley and Sons, New York, New York, USA. Mittelbach, GG; Steiner, CF; Scheiner, SM; Gross, KL; Reynolds, HL; Waide, RB; Willig, MR; Dodson, SI; L Gough. 2001. What is the observed relationship between species richness and productivity? Ecology 82: 2381-2396. Ostertag, R; JH Verville. 2002. Fertilization with nitrogen and phosphorus increases abundance of non-native species in Hawaiian montane forests. Plant Ecology 162:77-90. Pegtel, DM; Bakker, JM; Verweij, GL; LFM Fresco. 1996. N, K, and P deficiency in chronosequential cut summer dry grasslands on gray podsol after succession of fertilizer application. Plant and Soil 178:121-131. Perez-Corona, NE; van der Klundert, J; JTA Verhoeven. 1996. Availability of organic and inorganic phosphorus compounds as phosphorus sources for Carex species. New Phytologist 133:225-231. Powell, CL. 1974. Effects of P fertilization on root morpholgy and P uptake in Carex coriacea. Plant and Soil 41:661-667. Reddy, KR; Partick, WH; CW Lindau. Nitrification-denitrification at the plant root- sediment interface in wetlands. Limnology and Oceanography 34:1004-1013. Rhee, GY. 1978. Effect of N:P atomic ratios and nitrate limitation on algal growth, cell composition, and nitrate uptake. Limnology and Oceanography 23:10-25. Roem, WJ; F Berendse. 2000. Soil acidity and nutrient supply ratio as possible factors determining changes in plant species diversity in grassland and heathland communities. Biological Conservation 92:151-161. 192 Rogers, KH; Breen, PF; AJ Chick. 1991. Nitrogen removal in experimental wetland treatment systems: evidence for the role of aquatic plants. Journal of the National Pollution Control Federation 63:934-941. Romero, JA; Brix, H; FA Comin. 1999. Interactive effects of N and P on growth, nutrient allocation and NH 4 uptake by Phragmites australis. Aquatic Botany 64:369-380. Rubio, G; Oesterheld, M; Alvarez, CR; RS Lavando. 1997. Mechanism for increase in phosphorus uptake of waterlogged plants: soil phosphorus availability, root morphology and uptake kinetics. Oecologia 112:150-155. Ryser, P; Verduyn, B; H Lambers. 1997. Phosphorus allocation and utilization in three grass species with contrasting response to N and P supply. New Phytologist 137:293-302. Schaafsma, JA; Baldwin, AH; CA Streb. 2000. An evaluation of a constructed wetland to treat wastewater from a dairy farm in Maryland, USA. Ecological Engineering 14:199-206. Shaver, GR; FS Chapin III. 1995. Long-term responses to a factorial NPK fertilizer by Alaskan tundra wet and moist tundra sedge species. Ecology 18: 259-175. Smith, FW; Mudge, SR; Rae, AL; D Glassop. 2003. Phosphate transport in plants. Plant and Soil 248:71-83. Smith, FW. 2002. The phosphate uptake mechanism. Plant and Soil 245:105-114. Smith, LK; Sartoris, JJ; Thullen, JS; DC Anderson. 2000. Investigation of denitrification rates in an ammonia-dominated constructed wastewater treatment wetland. Wetlands 20:684-696. Soil Conservation Service, USDA. 1991. Technical requirements for agricultural wastewater treatment. National Bulletin No. 210-1-17. P.O. Box 2890, Washington D.C. 20013. Surrency, D. 1993. Evaluation of aquatic plants for constructed wetlands. In: GA Moshiri, ed. Constructed wetlands for water quality improvements. Lewis Publishers, Boca Raton, Florida, USA. Svengsouk, LJ; WJ Mitsch. 2000. Dynamics of mixtures of Typha latifolia and Schoenoplectus tabernaemontani in nutrient-enrichment wetland experiments. American Midland Naturalist 145:309-324. Taiz, L; E Zeiger. 2002. Plant Physiology, 3 rd ed. Sinaver Associates, Sunderland, Massachusetts, USA. 193 Tanner, CC; RA Kadlec. 2003. Oxygen flux implications of observed nitrogen removal rates in subsurface-flow treatment wetlands. Water Science and Technology 48:191-198. Tanner, CC; Clayton, JS; MP Upsdell. 1995a. Effect of loading rate and planting on treatment of dairy farm wastewaters in constructed wetlands ? I. Removal of oxygen demand, suspended solids and fecal coliforms. Water Resources 29:17-26. Tanner, CC; Clayton, JS; MP Upsdell. 1995b. Effect of loading rate and planting on treatment of dairy farm wastewaters in constructed wetlands ? II. Removal of nitrogen and phosphorus. Water Resources 29:27-34. Tilman, EA; Tilman, D; MJ Crawley; AE Johnston. 1999. Biological Weed control via nutrient competition: potassium limitation of dandelions. Ecological Applications 9:103-111. Tilman, D. 1997. Mechanisms of plant competition. In: MJ Crawley, ed. Plant Ecology Second Edition. Blackwell Science, Oxford, UK. Tilman, D. 1987. On the meaning of competition and the mechanisms of competitive superiority. Functional Ecology 1:304-315. Tilman, D. 1985. The resource-ratio hypothesis of plant succession. American Naturalist 125:827-852. Tilman, D. 1982. Resource competition and community structure. Princeton University Press, Princeton, New Jersey, USA. Tornbjerg, T; Bendix, M; H Brix. 1994. Internal gas transport in Typha latifolia L. and Typha angustifolia L. 2. Convective throughflow pathways and ecological significance. Aquatic Botany 49:91-105. Tripler, CE; Canham, CD; Inouye, RS; JL Schnurr. 2002. Soil nitrogen availability, plant luxury consumption, and herbivory by white-tailed deer. Oecologia 133:517-524. Tu, C; LQ Ma. 2003. Effects of arsenate and phosphate on their accumulation by an arsenic-hyperaccumulator Pteris vittata L. Plant and Soil 249:373-382. van der Eergen, LJM. 1982. Toxicity of ammonia to plants. Agricultural Environment 7:223-235. Van Duren, IC; DM Pegtel. 2000. Nutrient limitations in wet, drained, and rewetted fen meadows: evaluation of method and results. Plant and Soil 220:35-47. 194 Verhoeven, JTA; Koerselman, W; AFM Meuleman. 1996. Nitrogen- or Phosphorus- limited growth in herbaceous, wet vegetation: relations with atmospheric inputs and management. Trends in Ecology and Evolution 11:494-497. Vitousek, PM; RW Howarth. 1991. Nitrogen limitations on land and in the sea: how can it occur? Biogeochemistry 13:87-115. Vojtiskova, L; Munzarova, E; Votrubova, O; Rihova, A; B Juvricova. 2004. Growth and biomass allocation of sweet flag (Acorus calamus L.) under different nutrient conditions. Hydrobiologia 18:9-22. Vought, LB-M; Pinay, G; C Ruffinoni. 1995. Structure and function of buffer strips from a water quality perspective in agricultural landscapes. Landscape and Urban Planning 31:323-331. Walch-Liu, P; Neumann, G; Bangerth, F; C. Engels. 2000. Rapid effects of nitrogen form on leaf morphogenesis in tobacco. Journal of Experimental Botany 31:227- 237. Walker, JRL; S. Evans. 1978. Effects of quaternary ammonium compounds on some aquatic plants. Marine Pollution Bulletin 9:136-137. Wang, W. 1991. Ammonia toxicity to macrophytes (common duckweed and rice) using static and renewal methods. Environmental Toxicology and Chemistry 10:1173- 1177. Wassen, MJ; Olde Venterink, HGM; EOAM De Swart. 1995. Nutrient concentrations in mire vegetation as a measure of nutrient limitation in more ecosystems. Journal of Vegetation Science 6:5-16. Wilson, SD; D Tilman. 1995. Competitive responses of eight old-field plant species in four environments. Ecology 76:1169-1180. 195